Recently, Blaine wanted to create an AI-narrated audiobook using one of the voices from Eleven Labs. Halfway through the book, he decided to live stream his work session, so that anybody interested in what was involved could see exactly how it works.
If you haven't watched it already, we did demo how to use these new technologies a little over a month ago in Training #512, A.I. Narrated Audiobooks. That training covered Eleven Labs, Google Play, Play.ht, as well as Apple and Amazon's offerings. It also covered the differences and similarities between human and A.I. narrated audiobooks.
On this page, you can watch exactly what steps that Blaine took to create his audiobook, using a non-fiction book that he wrote. There were some technical difficulties (it doesn't appear in the recording but there was a massive echo for Blaine until he realized he was still logged into the webinar platform in a second window!) so the first 18 minutes are a summary of the types of issues that he found while creating the first half of his book.
Once the technical problem was solved, you can follow along with him while he finishes the second half of the book, and talks out loud about his thought process for each decision he made and action he took.
After the recording, there is summary of the lessons that were learned in this process.
Also to note, there are two transcripts and subtitles included; the first includes all of the text being read by the A.I. as well as Blaine's commentary as he went along, and the second only includes Blaine's commentary if you aren't worried about wading through the actual text of half of his book.
Click anywhere within this unedited transcript to jump directly to that part of the work session. This transcript includes both Blaine's narration as he is working on the project, and the results generated by the Eleven Labs synthetic voice.
Okay, so this afternoon and this evening I was working on editing the audio book made through 11 labs for a book. I just finished, writing and getting edited, so I thought I would share that with you because it's been an interesting process. Let me find the right tab so I can bring that up and share it with you. Okay, share screen. Okay, so now this book is about how to put on a road race and is my experiences in doing so. And so what I've been doing is there are, let me go back a little bit here. So we have, 28 chapters total. There's 24 chapters plus, introduction, final note from the author, my title and copyright, et cetera. And so I've been going through and I had gone in, chosen the, voice that I wanted to use and went and, did some testing, tested a few different voices, and now I'm in the process of proofing the whole thing and it does all need to be proofed. So, let's see, where did I leave off? I was on stage races, I believe is where I left off. So I'm just going to show you the process of what I've been doing and the changes that I've needed to make in order to get things to work. You do have to listen to the whole book. I'm almost halfway through it right here, and it's about 75 minutes total for the whole thing, but I will have to recreate some of it, and I won't necessarily stream the entire rest of this, but I just thought it might be interesting to show some of the changes that I have to make in some of the mistakes and thing, that it gets wrong and some of the things that it gets right.
📚 Technical Difficulties
Chapter 11, stage races. A stage race is a combination of a few different types of races. While technically a single race, it comprises multiple races that are usually run consecutively with a different leg each day. Each runner has a set amount of time to get from one point to the next each day, and their time is tracked and then added to subsequent times for each stage of the race. After runners finish for the day, they will usually have an overnight stay at a camp or other form of accommodation, which allows the runners to refuel rest and prepare for the next day's stage. Most stage races tend to be very long in duration.
All right, I'm going to actually pause this because I'm getting a pretty bad echo, so let me go in here and turn that off. Maybe that'll make a difference. Okay. Nope. Still getting an echo.
A stage race is a combination of a few different types of races while technically a single race. It comprises multiple races that are usually run consecutively with a different leg each day. Each runner has a set amount of time to get from one point to the next each day, and their time is tracked and then added to subsequent times for each stage of the race. After runners finish for the day, they will usually have an overnight stay at a camp or other form of accommodation, which allows the runners to refuel rest and prepare for the next day stage. Most stage races tend to be very long in duration. The self-transcendence run starts at six 9:00 AM and runners have until midnight to finish that day's. Miles. Runners attempt to cover 3,100 miles in 52 days. The marathon de sles is a six day stage race covering 251 kilometers, 156 miles through the Sahara Desert In Morocco. Stage races usually involve a team effort with one person being the runner, and the rest comprising that runner's crew. They make sure that the runner has the fuel and gear that they need at various checkpoints and aid stations along the way, and take care of any needs at night so that the runner can concentrate on recovering from the day's toils and prepare for the next day.
📚 Lessons Learned From the First Half
All right, so I do have a really bad echo and I can't seem to figure out how to get that to go away. So rather than letting you follow along, I'll just go back and show some of the changes that I've made and turn this off before I try proofing it again. 'cause I'm not gonna be able to do so effectively otherwise. So if we come in, some of the things that it's done really well, this particular one, I actually use this my paragraph as my sample paragraph, what I tried a bunch of different voices for. And it did a really good job on the first one with this voice. But then when I actually came in and redid the whole project after choosing the new voice that I wanted to use, it didn't sound nearly as good, but I was able to regenerate it and got that to come through pretty well. Let's see, some of the other changes, I had to split this paragraph up because it was having a lot of trouble with this one, this particular one, especially the second half of the paragraph. So one of the neat things about this is you don't need to have it matching exactly what's in your book. So you can go in and change a few things. So for example, I took out the beta breakers and then it had five kilometers in parentheses, and then after a run for passive river, that one's, it was, like two to 10 kilometers or something along those lines. And I also changed passive to a phonetic spelling of it to make it easier to pronounce it correctly. but it was having a lot of trouble with the individual, having the parentheses with the, the kilometers in there. One thing I did notice too is there is a few places where despite having, multiple people edit it, and I've personally gone through this book at least seven or eight times at this point, if not more. I am finding that there are some typos in there or some missing words. And, I've been going through and updating my, original manuscript alongside of it for when I find an actual error that I wanna fix in the print version and in the ebook version, I'm making notes of those and updating 'em as I go along. Let's see, what else did we find? Here was another one where I had to make some changes. for spacing wise. I did split this paragraph up into two paragraphs just to gi force it to give me a little bit more of a pause. And then originally this was one dash 99 and 200 to two nine to dash 299, et cetera. So I did change these one and at change the dashes to the word two, which made it a little bit clear exactly what I was going for. one thing I did like was, at one point I, when the first time I used RFID, it's a term related to the technology we're talking about in this book for this particular chapter. And it did have, a pretty nice, segue into defining it 'cause I just had it in parentheses right after. And that worked pretty well. Let's see, what did we have? Here's another spot where instead of 19 dash under, I changed it to 19 and under and added the twos in here instead of dashes. let's see. this one, I'm not sure if I'm gonna keep this paragraph. I may end up going back and changing this because I use U-S-A-T-F, which is the governing body multiple times in this paragraph, and then again in this next one. And it was kind of grading on me, so I'm probably gonna come back and make a change to this and I'll update the actual book for that, but I haven't done that one yet. here's another one where. I changed the punctuation. I added an ellipses instead of, a comma just to add a, put a little separation between these and that helped it figure out the right, way to, to say that particular line. and again, we have our dashes and I also added, punctuation, the ellipses on the end of this here. And I added this and in here on its own paragraph, just to provide that little bit of separation in here and to improve the way that it sounds. Now in the book, this is a bullet, bullet list. so you would see literal bullet points, and I don't even think I have a period or anything on the end of each line, but I wanted it to sound a little bit better with, for the person who's listening to this book. All right. Let's see. I know there was a few other items that I found that kind of tricked it a little bit. let's see. We're getting pretty close to where I had caught up. yeah, I think this one is probably a good one where I had to do that. So in this one. It's saying 1.0, I could change this to one, 0.0 and then type out point and type out zero or do zero. But 1.0 sounded, sounded fine to me. we did have, I think it might be the next one where I had to explicitly, oh, here's one, 5.72 miles. So you'll see this one. I, it was saying something like 5.72, and obviously that's not what I wanted, so I did have to explicitly spell this one out in order to get that to work. I also had a few spots where I had to change an actual word to something different just because it could not figure out how to, get it right. Let me see. I think it was in this one. I had, yeah. So the original sentence here, one way to accomplish this is to assign people to waves by, by their bib or something along those lines. here, I've got the book open. I can read off the exact one. We're in net time. Yes. One way to accomplish this is to assign people to waves by bib number, for example, by having a hundred people start at each minute. And I could just not, I mean, it says bibs right here, just fine. It was saying bib number just fine in this spot, but for this one, for whatever reason, it just could not pronounce the word bibs. So I had to change it, and just re reword it completely. I just gave up trying to get it to work. Let's see. We looked at that one. another one, it was actually doing an okay job. This was saying 18 colon, 30 and 20 colon 20, because these are times for people running their races. but for the most part, since there were other issues that I wanted to adjust inside of these paragraphs, I just took out the colons and left it as a space. I think it just tended to come out a little bit better. and the pacing for where it would say those words, it slowed down a tiny bit with a space instead of a colon. And I liked the sound of that better, and I didn't do that in every single one of them. You notice this one I never got around to changing. but I did do it on here. So it's basically on a case by case basis. Everything was converted up front, so I was able to come in and see how exactly it sounded and when I did need to make a change in that paragraph. Anyway, I just updated, I mean, I obviously missed this one when I was doing it, but it still came out sounding fine. So I didn't, didn't worry about it. And, This one down here did give me a little bit of an effort. I did wind up adding some random ellipses in here just to, used to be commas just to get this to, to speak it in a way that, that was worth listening to. another way you can actually, instead of creating new paragraphs, you can add these little dividers. but for the most part, I had more luck using actual para, changing it into new paragraphs and that seems to work just fine. Here's another one, 4.167. originally was just all nu, was an actual decimal, but I think that catches up to where we are and that's most of the, types of changes in editing that I've been doing. one thing that I did notice is the download. If we come in here, we can see our project versions and we'll see. It does take a long time to create that even after we've gone in and run our convert and done the whole, even if you do the whole thing at once, this was actually pretty quick to, to generate all of the text. But creating the MP three took a lot longer and it was, took me a little while to figure out how to get it downloaded, just 'cause I wanted to see how long the whole thing was as a project, but. You can come in here and download the whole project, or you can do it on a chapter by chapter basis by looking at the chapter versions and downloading that way. so I'll probably wind up doing both once I'm done, but most likely I'll be working with the individual chapters and depending on how those come out, I will wanna have a single MP three that's the entire book without having separate chapters, and I'll just cre create that myself. But I wanna, I may want to add a little bit of room tone in between the chapters or something along those lines. That'll be something I'll experiment with as I am going along a little bit later. But anyway, hopefully people found this, mildly interesting at least. Sorry I can't just go in and let you guys watch over my shoulder, but that echo that's coming in because I'm sharing the audio and I've got my headset on to listen to, was just a little too distracting and would take me way too long to actually finish doing the work.
📚 Technical Difficulties Solved!
So I'm going to log off of here and I will see y'all next week on Erin's training on Tuesday. Oh, I think I figured out why my. Room is echoing. I am actually logged into our room in two places. Find the other one. Here it is. Now I'm gonna just try this again and maybe you can watch over my shoulder.
A stage race is a combination of a few different types. That's much better races that are usually run in the, each runner has a set amount of time to get to one point to the next. Each day, their time is tracked and added to subsequent time. After runners finish for the day, they will usually have an overnight stay at a camp or other form of accommodation, which allows the runners to refuel rest and prepare for the next.
So I think I was just doing that again and I stopped sharing my screen so you guys aren't actually able to follow along with that. So let me put that back on. So sorry about that. So just jump, 18 minutes forward and I'll leave a note. I'll try to remember to leave a note. So if anyone wants to watch the actual process, skip the first 18 minutes of this.
📚 Workflow: Follow along while Editing an Audiobook in ElevenLabs
And since you guys, if you have been watching this, probably listen to this like three or four times now. I'll just skip that one and come back after I stop live streaming. Just to give you something a little different to listen to
Chapter 12, the fun run to top things off in this.
Yeah, I don't like the way that it said that. So I'm gonna just try this again. Just takes a moment once I chapter 12,
regenerate the fun run. It's a little better. Chapter 12, the fun run to top things off. In this section, let's talk about the most basic kind of race one where no one even needs to worry about scoring runners. The fun run can be anything from a runner going from point A to point B following a set course or just running through a field. There is some kind of race course and there may or may not be a clock at the finish line. The race may or may not allow people to wear watches, but the race organizer will not bother tracking who is coming over the line or when they do so. The point of the fun run is camaraderie and not competition. Most races for children are fun runs. They may be races for kids at their local elementary school or at a county fair, or they may be races held before or after an adult race to involve the whole family. Maybe each kid will get a ribbon or a medal at the end, or maybe not running clubs and training groups often organize fun runs either to practice for a big upcoming race or just to encourage fun and camaraderie.
Okay, so this is another spot where it didn't sound quite right. So what I'm gonna do is I'm gonna actually try bring a ellipses in here and then have it just regenerate all three of these sections and see if it sounds a little bit better.
Most races for children are fun runs. They may be races for kids at their local elementary school or at a county fair, or they may be races held before or after an adult race to involve the whole family. Maybe each kid will get a ribbon or a medal at the end, or maybe not running clubs and training groups often organize fun runs.
So one thing, hasn't come up here and we might see it as we're going through, but I've also found sometimes just little artifacts will come in and appear
randomly either to practice for a big upcoming race
or it was generating that as I went along. but one of the main reasons you do have to listen to the whole thing is so that you can hear those artifacts, when you're going through. 'cause you don't wanna leave those in the final, version. So here, why don't I zoom up a little bit just to make it easy for people to see if they wanna read along as I'm going. Alright, so now we have this all set. I think that delay seemed
just fine. Running clubs and training groups often organize fun runs, either to practice for a big upcoming race or just to encourage fun and camaraderie. They also encourage participation from runners who are new to the sport or might otherwise be intimidated by a more serious race. Fun runs can be large and well organized events as well, even without the competition. For example, color runs are five kilometer untimed events where participants start with a white shirt but then run through different colored powders at e kilometer, at the finish line. Organizers throw a large party. There are very few rules to a fun run, and it doesn't matter when folks cross the line. Race organizers still wanna make sure the course is safe and that everybody's needs are met, of course, but they don't have to worry about tracking everybody's time and providing results.
All right, so that was a pretty straightforward, chapter. We can move right on to the next one, and this starts a new part of the book. So what I did was instead of having a separate chapter for, the parts, different parts of the course, of the, of the book, 'cause the book is split into single race single runner. Part two is multi race single runner. Part three is a single race, but a team of runners. And then part four is just specialty formats that don't fit into any of those. I wound up setting these as all is heading one to just kind of give the, this the same gravitas as the chapter, declaration. And then, I figured it's only a couple of seconds for each one. So I just put those at the beginning of the sections and didn't worry about having their
own part two multi race single runner.
Alright, I don't know why it didn't keep going.
Chapter 13. There we go. Cumulative time or distance.
Just wanna make sure that's not a huge extra delay and it was just having to redo this for whatever reason.
Part two, multi race, single runner. Chapter 13, cumulative time or distance. Not all race scoring deals with individual races. Sometimes runners are tracked across a series of races with awards based on how they perform throughout those races and not just with a single event. The easiest way to score these races is by calculating the cumulative time or cumulative distance. If the races are for a set duration instead of a set distance, scoring by cumulative time is pretty simple. Simply add up a runner's finishing time for each race, and then sort based on that time, each runner has to run every one of the races to count or the whole thing falls apart. Alternatively, race organizers could count only a subset of the races if they don't wanna force every runner to run every race. In the series, I help organize the weekly back cove 5K in Portland, Maine. It is a series of races every summer that, depending on the summer, is between 14 and 19 weeks long. Once each week, the runners compete on the same course. The organizers track the cumulative time for all the races that each participant has run. Runners can then compare themselves against everybody that has run in the same number of races that they have, as not everybody runs every week for scoring. However, the organizers only add the six fastest races. Each participant has run as long as a runner has completed at least six races in the series. They can be considered for the leaderboard and prizes. The top male and female usually win free pairs of shoes and age group winners get gift certificates to the local running store. Alternatively, we could have chosen to score a runner's first six races or their last six races, but using the six best finishing times encourages participation and improvement throughout the season. Don't love that. It'll work. Effect requiring a certain number of races in a long series is that the leaderboard can and often does change each week, thanks to varying weekly participation. A speedy runner who has only been able to complete five races earlier in the summer can seem to jump out of nowhere when they tow the starting line at the final race of the season and move from prize ineligible to top of the leaderboard. The leaderboard can also be shaken up quite a bit by people who run a lot of the races because they may not actually treat all of them as races. Sometimes a runner might decide to run with another person or push their child in a stroller or use the motivation of the other runners to complete a speed workout. In other words, some runners may have some race times that are a lot slower than their actual race pace push.
Yeah, this one has quite a few problems, so I'm gonna split this into multiple paragraphs here so that we can get a good paragraph, a good reading on each one of those and not have to regenerate at all at once. 'cause you do have. Every time I regenerate, I am using some of my quota and I'm not gonna go anywhere close to this much. The whole book is about 76,000 characters total, about 13,000 words. so I obviously have plenty left, but I've already used, about 90,000 so far and I'm halfway through the book. So it's, probably will wind up with around a hundred to 110,000 characters left, is my guess.
The leaderboard can also be shaken up quite a bit by people who run a lot of the races because they may not actually treat all of them as races.
Alright, those are generated, let's play it
through. The leaderboard can also be shaken up quite a bit by people who run a lot of the races because they may not actually treat all of them as races. Sometimes a runner might decide to run with another person or push their child in a stroller or use the motivation of the other runners to complete a speed workout. In other words, some runners may have some race times that are a lot slower than their actual race pace. Pushing them temporarily lower on the leaderboard than you would otherwise expect towards the end. Alright?
Still not a huge fan of the way that it's saying speed workout over here. Let me try putting a dash in here and see if that helps a little bit. Another option is we can go into these foot voice settings and try things. Oh, it looks like it saved my style. Exaggeration. This should be much lower. I thought that was gonna be on a paragraph by paragraph basis, but apparently it's been using that since then. So that's a good thing to know. If you do change your paragraph settings, it seems to carry over for any future ones that you're updating. but let's see. Good thing I checked.
Sometimes a runner might decide to run with another person or push their child in a stroller or use the motivation of the other runners to complete a speed workout. In other words, some runners may have, yeah, that sounded much better. Some race times that are a lot slower than their actual race pace, pushing them temporarily lower on the leaderboard than you would otherwise expect towards the end of the season when they may take the races more seriously. The difference between how fast they ran that week and what is now their seventh fastest time can cause them to jump from a lower place on the leaderboard to a much higher position. This phenomenon is more likely to occur when the series includes significantly more race opportunities than it requires finishes for prize eligibility. It is also more likely to occur when the series fee covers the entire season or the per race fee is minimal. These possibilities make scoring the final race at the end of the season, especially fun if expedient awards are a priority. These possibilities also require a little organization. In the case of the series I mentioned above, I am responsible for figuring out who has the potential to jump onto the leaderboard. I need to account for multiple interactions with each runner's race performance, how much time is going to be dropped off for each runner that is behind first place. If they run a new top six time for the season, the slowest time each of those people can run in order to overtake first how much faster they'll have to run than the first place person in order to prevent first from maintaining their lead. Thankfully, we only give out first place prizes and don't have to worry about managing any other positions, which would make for exponential numbers of relationships to figure out. Giving age group awards did not increase the complication level thankfully. But it has increased the amount of work to figure everything out as a general rule.
Cumul. Alright. I think that I'm gonna regenerate this. I think it sounds okay, but it could sound better. I haven't tested this yet, but I'm gonna try seeing if I can give a little more emphasis by putting that there. I don't know if that will work, so let's give it a try and see.
Giving age group awards did not increase the complication level, thankfully, but it has now increased the amount of work to figure everything out. Yeah, that
definitely didn't sound good. So, let's get rid of that and you'll notice as a side note, it tells you what's been converted already. See the black lines here and then we can see when I make changes, then I need to still reconvert these particular pieces so you can see, where exactly you still need to get some work through. Let's try doing both of these again and see how well that works. Sometimes you just have to regenerate three or four times and you wind up with, with something that does catch it pretty well.
Thankfully, we only give out first place prizes and don't have to worry about managing any other positions, which would make for exponential numbers of relationships to figure out. That one's all
set now. Let's do this one again.
Giving age group awards did not increase the complication level, thankfully, but it has increased the amount of work to figure everything out.
Yeah, that sounds way better than the original version of this particular sentence, but thankfully was good the first time. Pretty bad the last time, but the emphasis on has is much better with this one. So
as a general rule, cumulative time works best when races are all the same distance, if not on the same course. That way every race is on an even playing field and one specific race give a certain type of runner more of an advantage in the overall scoring. There's a big difference between being beaten by one minute in a one mile race than somebody being beaten by one minute in a half marathon after all.
Alright. Not in love with this particular one, but it sounds a little awkward with the way I phrased it. So I'm gonna just read this out loud a little bit to myself and try and see what I was thinking when I wrote it this way. And I may reword this, which will involve me actually updating the book itself as well, not just the way I'm having it read in the audio. So as a general rule, cumulative time works best when races are all the same distance, if not on the same course, actually. Like the way that it, it read that. So hopefully it keeps that same intonation that it used that way. Every race is on an even playing field and one specific race give a certain type of runner. Yeah, I think give is the wrong word I'm trying to use here. Yeah. So no matter how many times you've edited something, you're always missing a word here. So I'm gonna try splitting these a little bit and regenerate these and I'll go in and update at the same time in my original manuscript.
As a general rule, cumulative time works best when races are all the same distance, if not on the same course. That way every race is on an even playing field and one specific race doesn't give a certain type of runner more of an advantage in the overall scoring. There's a big difference between being beaten by one minute in a one mile race than somebody being beaten by one minute in a half marathon. After all that becomes evident in alright, it did do this sentence really well.
I guess I should have split this up so I didn't have to regenerate it. We'll do that first, but yeah, it's let's put a dash there. See if there's a way I can get this so that it gets it right. Let's see. There's a big difference between being beaten by one minute and a one mile race and somebody being beaten by one minute and a half marathon after. Let's see if adding that comma fixes that. There is a bit of waiting in this process
that way every race is on an even playing field and one specific race doesn't give a certain type of runner more of an advantage in the overall scoring. There's a big difference between being beaten by one minute in a one mile race than somebody being beaten by one minute in a half marathon after all. Perfect. That's much that becomes evident in Trail. Monster Runnings, Bradbury Dirt Series in Pound, Maine. That summer series consists of three races of six, nine, and 12 miles respectively. While all are located in the same state park, they are each run on very different courses. A fast road runner can excel on the six mile course, but may struggle on the nine mile course, which is much more hilly as its summits or near summits the mountain, multiple times each loop, that same road runner may be positively lost in the 12 mile course where they have to navigate twisty, turny, Rocky Rudy, single track. However, a runner losing a minute per mile in the six mile race compared to the nine or 12 mile race will not be disadvantaged nearly as much as if they lost a minute per mile in the longer races. Where that slower pace increases their cumulative time by an extra 50% or 100% compared to the first race. One way around that is to score using the average pace rather than the cumulative time.
Alright, now this paragraph I'm trying to decide, it's some things it does really well. A few things, it's just kind of adequate. So I may revisit this and come back and split this into multiple paragraphs and do a few generations on it. not in love with it, but it's not as bad as some of the others have been. So for now, I think I'll just move along. So I do notice when I click on a new chapter, it leaves the position of whatever you're looking at, down at the bottom. So you do have to scroll back up, which is a little bit of an annoyance in the user interface, but it's not the end of the
world. Chapter 14, average pace.
Alright, that seemed really fast.
Chapter 14, average pace.
Yeah, I think I'm gonna regenerate this. It just seems to be saying it super fast and I don't really like that
chapter 14 average pace
there. That's much better. So, one thing you may have noticed, I pressed this button. This, if you have it turned off, it'll just read the paragraph you're on and if you turn it on, it'll let you go through and listen to each paragraph one after another and
keep going. Chapter 14, average pace a race series scored using average pace can be a little fairer when scoring races across different distances in these races, you figure out what the average per mile or per kilometer pace is for each race and then score the runners based on those times rather than on their finishing time. For example, a friend of mine directed a series of races that included a one mile race and a half marathon if he were to score using the average pace. Then
some. See it does pronounce things the same words differently. You'll notice that, on a chapter or two ago, I had to have a dash to put half a marathon together. 'cause otherwise it was sounding kind of funny. But this time it got it no problem. So that's good.
Somebody who ran a 5 31 mile and a 91 46 half marathon would have an average pace of five 30 per mile in the one mile race, and seven NAS per mile in the half marathon.
Alright, now you notice it had this seven knots, right here, which is obviously not correct. so I think what I'm going to do here is I. I'm gonna just add in the one mile and a 91 46 half marathon. That should be fine. would have an average pace of five and it's, and 30 seconds per mile in the one mile race and seven minutes per mile in the half marathon. So other than, a slight change, up here, it, the five minutes and 30 is just another way of saying that, so I'm phonetically putting it out. All right. That one should probably be fine there, and let's just try these again and see how it comes out.
For example, a friend of mine directed a series of races that included a one mile race and a half marathon.
I'm putting me in the, into the actual manuscript because I think I, that sounds a little bit
better. If were to score using the average pace than somebody who ran a five 30 in the one mile and a 91 46 half marathon would have an average pace of five minutes and 30 seconds per mile in the one mile race and seven minutes per mile in the half marathon. Yeah,
that sounds much, much better. So let's generate this one.
They could then be scored with an average pace of six 15 per mile, which is arrived at by adding the two times together and then dividing by the number of races. Alternatively, they could be scored by adding the average paces for each race for a cumulative total of 1230. Average pace suffers from a similar drawback to cumulative time. However, if Andrew has an average pace of six 15 per mile between the two races after running a five 30 mile and 91 46 half marathon, and Bob wants to beat him, then how would he best do so? If he runs 13 seconds faster in the mile, but 13 seconds slower in the half marathon, then his per mile pace would be five 17 in the mile and seven SUD or a one in the half marathon for an average pace. Pace of six.
So it's does not know how to deal with zeros in the times seven SUD one is a new one That's kind of interesting. Wasn't really in love with that anyway. So why don't we generate these 7 0 1 and a half marathon and we're just gonna preemptively assume this is gonna need fixing two. the rest of these should be good, so let's try it again.
Average pace suffers from a similar drawback to cumulative time. However, if Andrew has an average pace of six 15 per mile between the two races after running a five 30 mile and 91 point 46 half marathon, yep, that
just needs to be regenerated. I had it right the first time, but you can't keep the first half of it unless I wanted to export and edit it manually, which I don't wanna do.
Average pace suffers from a similar drawback to cumulative time. However, if Andrew has an average pace of six 15 per mile between the two races after running a five 30 mile and 91.46 half marathon, all
right, I'm gonna just get rid of that colon. I don't want it to be 91.46. I just want 91 46 half marathon. All right. Does not seem to wanna be cooperating. Here we go.
Average pace suffers from a similar drawback to cumulative time. However, if Andrew has an average pace of six 15 per mile between the two races after running a five 30 mile and 91 46 half marathon, and Bob wants to beat him, then how would he best do so? If he runs 13 seconds faster in the mile, but 13 seconds slower in the half marathon, then his per mile pace would be five 17 in the mile and 7 0 1 in the half marathon for an average pace of 6 0 9. He wins. However, if he were 13 seconds slower in the mile, but 13 seconds faster in the half marathon, then his per mile pace becomes 5 43 and 6 59 respectively for an average of 6 21. Now, Bob loses losing to another runner by 13 seconds and a half marathon isn't really that big of a deal. Both runners basically ran the same pace the whole way. However, that 13 seconds is a much different pace between the two runners in the one mile race, and that's reflected a bit more fairly in the final scoring. However, it gives an advantage to the sprinter over the long distance runner in these races. A fairer way to score the runners from race to race would be to take time out of the equation altogether. All right, that's a lot better. Chapter 15, positional scoring. Positional scoring takes the time out of the picture by awarding the person based on what position they finish the races often by averaging them out. Instead of keeping track of finishing times, the race would keep track of racers finishing positions add those positions together and the lower score wins. Racers could track what place runners finish each race for their specific category, such as age and gender, or using the overall position they finish. This gets around the problem of giving an unfair advantage to runners because of a discrepancy in the length of each race, but it does mean that there can be an unfair advantage based on who actually appears to compete in each of the races. If races choose to track the positions runners finish in, they should also create a category for runners who have competed in every event, or at least in the minimum number of events that are going to be used for scoring. That way you account for the finished position only among those that are eligible to score and aren't impacted by a more popular race, which could skew the results towards whichever has a larger field. For example, in the case of the one mile and half marathon, let's keep the numbers simple and say that three people run both races as part of the challenge, but there are a total of 10 finishers in the one mile race and a hundred finishers in the half marathon. Let's suppose that Andrew finished in fifth overall in the mile and 50th overall in the half marathon, and Bob finished in fourth and 70th, and Carol finished in eighth and 60th. If you consider their overall finishes, Andrew's score adds up to 55, Bob to 74, and Carol to 68. If you only consider the finishes amongst the three who ran both races, then Andrew has three, Bob has four, and Carol has five points. In both scenarios, the lower score would win and Andrew would come in first place for the two race challenge. However, if you are using the overall finish results for all runners and not just the ones that are competing in both races, then the smaller race has a much smaller impact and in this case wouldn't impact the finishing order at all as Carol would get second place. By limiting the finishing positions only amongst those taking part in both races, each race's performance plays an equal role in determining the final standings. In that scenario, Bob's first place and third place finishes would give him a slight edge over Carol's second place and third place finishes For my summer series, we don't score people based on their average finishing position, but we do report their average finishing position to them as an interesting data point.
All right. That chapter seemed okay on its own
chapter 16 point scoring. Point scoring is a great way to take the discrepancy of time or distance out of the picture and is a lot easier to manage the math at the end of each race in the series to see where each person is going to end up. How point scoring works is that you earn a certain number of points based on what place you come in during the race. There are a few decisions races need to make when,
okay, so one of the issues I found in my editing was, I use the word you multiple times to refer to different people, race directors, the races themselves, the runners. I. It wasn't very clear. So that was one of my big editing tasks was to go in and obviously I, missed a spot right here. So how point scoring works is that runner earn a certain number of points based on what place they come in during the race. so this will fix that particular paragraph. So while that is regenerating, I'm actually going to have to go in and fix my manuscript as well. So this is a nice editing process too if you do it before the book's been published, like I'm doing here.
How point scoring works is that runners earn a certain number of points based on what place they come in during the race. There are a few decisions races need to make when choosing how the scoring will work. However, first races need to decide if they are going to score low or score high. If the runners are going to score low, they earn a number of points for the position that they finish the race in and the low score wins. This is known as cross country style described later in this book in more detail under team scoring. If they are going to score high, then the race will assign a higher score for finishing further up in the field and the points runners earn go down the further back in the field that they finish. The race needs to decide who can score and how they count them. If they count everybody in the race, then a popular race can impact the series because there will be a lot of extra runners that don't run all or enough of the races to score in the series. Those extra runners might take some points that would otherwise go to somebody competing. On the flip side, if they only count runners who have run a specific number of the races, there's extra overhead to weed out those that shouldn't count. Finally, races need to decide how much weight they want to give each position. A runner may finish in, for example, do they get one point for each runner that person beats to the finish line, or does first place get 10 points and five points goes to second and one point to third? Are points going to be awarded to every runner in the race or is there a limit? One of the nice things about sounds, a little petulant being able
to earn points, I kinda like it though, is that
they all feel as though they are being included, but then run into the problem mentioned previously where a popular race could then unfairly skew the number of points awarded in that race. For example, let's say in a series of three races, there are 30, 40, and 70 finishers, and they award one point for each person, each runner beats. That means that if Andrew finishes in second, 10th and 14th, he would earn 28, 29, and 55 points for a total of 112 points. If Bob finishes in 10th, 20th and then wins the third race, he would earn 19 points for both of the first two races and 69 points for his third race, giving him 107 points. This method won't work quite as well if you aren't comparing people across all the same races. For example, if we scored people that weigh in my weekly series through the summer and only counted the points from their top six races, then somebody who finishes first in all six of his fastest weeks could lose to somebody who finishes in 10th in all six of his fastest weeks. If there are more runners those weeks and they are able to earn more points that
alright, this particular paragraph, the second half of it was fine, but this first sentence and the transition was kind of awkward. So I'm going to split this up a little and just do the, have it regenerate these,
this method won't work quite as well if you aren't comparing people across all the same races. That sounds fine.
Let's do this one.
For example, if we scored people that way in my weekly series through the summer and only counted the points from their top six races, then somebody who finishes first in all six of his fastest weeks could lose to somebody who finishes in 10th in all six of his fastest weeks. If there are more runners, those weeks
and, all right, this one, wherever it was up here, could have been a little bit of a glitch. And just in terms of putting it in, I'd wanna listen again, but it's definitely a little more awkward down over here, so there's no point in listening to the whole thing. We'll just regenerate this whole paragraph, whole sentence. It's a very long sentence.
For example, if we scored people that way in my weekly series through the summer and only counted the points from their top six races, then somebody who finishes first in all six of his fastest weeks, could lose to somebody who finishes in. Okay, I'm gonna
just rewind a little. You don't have to actually start at the beginning of the paragraph. You can use the scrubber bar if that wasn't obvious. I just wanna see if that's a problem with it. Just loading the second half of that a little bit later and that's where the lag is. I'm assuming that's the case, but we'll find out.
Fastest weeks could lose to somebody who finishes 10th in. Then somebody who finishes first in all six of his fastest weeks could lose to somebody who finishes in 10th in all six of his fastest weeks. If there are more runners those weeks and they are able to earn more points,
yeah, so that was just a glitch. So what happened if you were listening is it had basically generated the first half of the sentence and started playing it, and it hadn't quite finished the second half to add it on. So that little hiccup that we had in between the all and the six was just a little bit of lag while it finished generating and gave us that, second half of the sentence. So when we re-listened to it, it sounds fine the way that it is, and this generation's much better. So let's keep going on. Now,
for example, if we scored people that
way and oops, I'm ing the wrong button, this is just replaying the one that I was already on. What I actually wanted to do was hit the play button up here.
One way around that would be to assign a certain number of points for first place and then to drop down from there with the top runners earning more points. That might mean that points is awarded for first nine for second on down to one point for 10th.
that might mean that 10 points. Apparently I deleted the 10 somewhere in my editing process. So generate that and let's fix our manuscript. My manuscript
is correct. One way around that would be to assign a certain number of points for first place and then to drop down from there. With the top runners earning more points, that might mean that 10 points is awarded for first nine for second on down to one point for 10th place, or for a larger race or one that wants to be more inclusive. They could start at 30 points or a hundred or even 1000 points. Note that they don't have to have that many runners. First place is always worth the same number of points throughout the series. There could also be tiers.
Alright, so one thing that I noticed is that this missing 10 points, was actually in my manuscript. I don't know why it disappeared when I imported my epub could be a problem with, I was using vellum to create the eub. So maybe for some reason it just disappeared from there. So that's something I wanna keep a, let's see. actually, am I looking in the right spot? Nope. I was actually looking at this one here, so I was off the wrong paragraph. So nevermind. yes I am in fact missing the 10. So it was not a problem with 11 labs, it was just a problem with my looking in the wrong paragraph. Alright, so this is one of the nice things as you can get in a little bit extra editing and making sure you have things right and you're not putting in simple typos. one of the suggestions that we have in our self editing is to go in and actually read your book out loud or use, your computer to have it read to you. And this is basically just another way of having the computer read it to you except that we're getting a recording out of it as well.
So one way around that would be to assign a certain number of points for first place and then to drop down from there with the top runners earning more points. That might mean that 10 points is awarded for first nine for second on down to one point for 10th place. Or for a larger race or one that wants to be more inclusive, they could start at 30 points or a hundred or even 1000 points. Note that they don't have to have that many runners. First place is always worth the same number of points throughout the series. There could also be tiers
where, alright, so this is a situation where I think we need to have a little pause. So I'm gonna go in here and I'm gonna try seeing if I can get two pauses that way. Let's try and see how that
works. There could also be tears where at a certain point, one way around that would be to assign us the same number of points throughout the series. There could also be tears.
that didn't really work 'cause you'll notice that there was a little bit of an artifact glitch. Let's try regenerating and see how that sounds. Okay, that sounds better.
Let's try it again. One way or wrap, many runners first place is always worth the same number of points throughout the series. There could also be tiers pause, where at a certain point the runners would earn the same number of points as some other runners. For example, 10 points could be awarded for first, nine for second, eight for third, and so on until the two points for ninth place. Then award two points for anybody who finishes in positions nine through 19. And then one point for anybody who finishes in 20th or later. And as with single race awards, age and gender awards could be further broken down to only compare how people place versus others in their age group rather than how they finish overall when awarding points. The final results can then be used in multiple ways to figure the final standings. First points could be awarded for each person based on their overall finish with age group divisions determined by how each person raced in the overall field compared to others of their same age and gender. Alternatively, points could be awarded both overall and within age and gender divisions to provide individual sets of results for all runners and for each division. The overall results would take the entire field into consideration and the individual divisions would only consider others that belong in that same division, potentially with overall winners removed or ineligible to win a second award. Now finally, points could be awarded solely with,
alright, and now we notice there's a little artifact at the very beginning of this one.
Then each age group finally, points could be awarded solely within, oh, maybe it's at the end of this one. Alternatively, points could be ineligible to win a second award. Now finally, points could be award. Yep. So that
little glitch is at the end of this one. So even though this paragraph sounds pretty good, we're going to want to go in and regenerate it. Usually when it's a little artifact like that, then regenerating comes out just fine and there isn't a problem the second time, but you do have to check and figure out where exactly that
artifact is going. Alternatively, points could be awarded both overall and within age and gender divisions to provide individual sets of results for all runners and for each division. The overall results would take the entire field into consideration, and the individual divisions would only consider others that belong in that same division, potentially with overall winners removed or ineligible to win a second award. that's
sounded better the second time anyway, so that's good.
Finally, points could be awarded solely within each age group division, but then those points totals could be consolidated into an overall awards list. In this case, a runner's score is based on how well they competed against others in their same demographic, and that performance is then reflected in the overall field. This would allow a dominant runner in their twenties to compete fairly with a dominant runner in their eighties, despite the likely difference in how fast each of them can run their respective races. This final method could easily cause more ties in the results. However,
not a huge fan of the end of it, but the rest of that sounded fine, so I'm gonna leave it as is. Now we're in part three.
Part three, single race team, chapter 17, positional cross country scoring.
Alright, so I don't, sometimes I want the slash to be ignored. Sometimes I want it to be said. Sometimes I want it to be ignored and it just puts something random in there. So in this case, I want to pl explicitly say positional slash cross country scoring because it's more of a either or type of situation.
Chapter 17, positional slash cross country scoring. Positional scoring for teams is also known as cross country scoring. Basically, the top runners from each team have their finishing places added together and the low score wins. There are a few variations, but in almost every case, teams need to have a minimum number of runners in order to have their score counted. In high school and college cross country races, at least here in the United States, a team comprises seven runners, five of whom score a team is only considered in the results as long as at least five runners finish and up to seven can displace runners from other teams depending upon the size of the meet, which is what a cross country race is called. Other runners may be allowed to run, but any slower runners on the team do not displace other runners from other teams. When scoring in larger races, teams will often be limited to only being allowed to start seven runners, both to keep the size of the race at manageable levels, and also because when there are more than seven runners and one runner gets injured, then that team has an unfair advantage by allowing an eighth runner to move up to the seventh runner spot. Under those rules. The best score a team can achieve is 15 points by scoring runners in the top five overall in all five positions. This is considered a perfect score when using positional scoring for a team race, a runner's time does not matter. All that matters is how well he can run compared to how well members of other teams can run. It truly makes it a team effort because no matter how fast your first runner, yeah, I didn't
really like the first few, so I'm gonna redo this one.
When using positional scoring for a team race, a runner's time does not matter. All that matters is how well he can run compared to how well members of other teams can run.
Still not a huge fan. So let's regenerate that one and I'm gonna try and explicitly get a pause in there and see how that sounds.
When using positional scoring for a team race, when a runner's time does not matter, all that matters is how well he
added an extra word in there for some reason,
when using positional scoring for a team race in a runner's time does not matter all that ma, so
this.dot does not always work. Let's try Semicolon and see if that works a little bit better. I haven't had it just putting in a random word like it seems to be doing this time, but you know, the, other pieces of 11 Labs has a way where you can explicitly put a pause in, but that particular feature doesn't work in these projects where we're working on a whole audiobook at once.
So when using positional scoring for a team race, a runner's time does not matter. All that matters is how well he can run compared to how well members of other teams can run. It truly makes it a team effort because no matter how fast your first runner or first few runners can be. Those fourth and fifth place runners can make all the difference by reducing the number of points the team earns and the sixth and seventh place runners on the team can push another team score higher.
All right, so I definitely think that this came out. I'd still not a huge fan. In
adult races, you do not see cross country scoring nearly as often. Instead, cumulative time is typically used.
Yeah, so I'm still not a huge fan of this right here, but I feel like it got the end of the sentence right. So I'm gonna stick with what we've got now. Alright, we're getting close to the end. Chapter
18, cumulative time. Cumulative time in team scoring is used for pretty much every team race that I have run once I graduated from college, whether they be cross country or road races. Similar to cross country style, A team comprises a specific number of runners, but instead of adding up the position they finished in the race, you instead add up their finishing times U-S-A-T-F. The governing body for races in the United States usually requires between three and five runners, depending on the distance of the race. Anything up through a half marathon will usually require five runners and anything that is a marathon or longer will usually only score the top three runners. They must feel it will be harder to field a team for anybody that has to run over 26 miles. Teams in these races will usually be segregated by gender with both open and masters' divisions. Masters races are limited to runners that are 40 years or older, but those runners can also score on their teams, open teams as well, even in the same race. For other races that are not U-S-A-T-F championships, the number of runners can vary widely based on the size of the race and the whims of the race director. I have seen races that only required two runners per team, and I have seen races that required up to 10 or 12 runners per team. There usually will not be a cap for the number of runners on any given team, and whether they displace other runners will not matter because it's only the finishing time that makes a difference. That allows for a team to field any runners in their club, even though only the top three or five or however many runners will actually score. This also means that the final times for specific teams can be weighted heavily by a particularly slow or a particularly fast runner.
I don't think I'm gonna regenerate it. I wasn't a huge fan right in here with the way that it paced there, but, it's okay.
Chapter 19, team composition. In high school and college, it is very rare that girls and boys will race together. They will almost always have separate races segregated by gender, and may also have separate races for varsity. The top seven runners on the team and jv, everybody else in team races for adults, the rules are a lot looser. As mentioned previously, U-S-A-T-F championship races are segregated in the scoring by only considering men's teams and women's teams. But in other races you will find plenty of examples of mixed teams as well. Note that international rules require that men and women run in separate races. Even if that just means starting the elite women ahead of the men to be eligible for world records, that could mean that there are separate.
Alright, so I'm assuming that this is where that little artifact came in. 'cause it sounded like it was stuck on the vat.
That could mean that there are separate categories for
or not, looks like it is at the end of this one
in team rate to be eligible for world records that could, yep.
So we're gonna have to regenerate this whole thing 'cause of that, the very end.
In team races for adults, the rules are a lot looser. As mentioned previously, U-S-A-T-F championship races are segregated in the scoring by only considering men's team and women's teams. But another races, you will find plenty of examples of mixed teams as well. Note that international rules require that men and women run in separate races, even if that just means starting the elite women ahead of the men to be eligible for world records.
I, other than that little artifact at the end, I liked the first time that this came through, but the second time, I really like how it emphasized the word elite there. It gave me a, made me wanna chuckle a little bit.
That could mean that there are separate categories for men, women, and mixed genders. Or it could mean that there's only a team score and it doesn't matter what the breakdown is between the genders, gender will often impact the requirement.
All righty, there's another
little artifact that could mean that there are s events based on team results, but do not want to limit or dictate who can be on a team is to score. That could mean that doesn't matter what the breakdown is between the genders. Gender will often
impact. Yep. There is a little artifact on the end of that one, so we'll have to regenerate this one too. And this is why we have to go through and proof this. So, as I said, this was originally about 75 minutes long for the, that could mean
that there are separate categories for men, women, and mixed genders. Or it could mean that there's only a team score and it doesn't matter what the breakdown is between the genders,
but I'm probably spending good three hours at least on, on doing this when you count the first half and figuring out what I'm doing and everything.
Gender will often impact the requirements for having a valid team, but don't necessarily impact how the race is scored. So if a race allows mixed gender teams, for example, there may be a requirement for five runners and at least two people of each gender scoring will usually be done with either cumulative time or using positional scoring. A fair alternative for races who want to score their events based on team results, but do not want to limit or dictate who can be on a team, is to score based on the age and gender graded times compared to somebody of each person's age and gender's world record time. This puts every runner on an even playing field without requiring specific rules about who can run on each team.
this one sounded fine in terms of the audio, but this paragraph is a little bit awkward, so I think I'm gonna wanna rewrite, write it a feral turn of races. Score their events based on team results, but do not wanna limit or dictate who can be on a team is to score based on the age and gender grade times, discussed in chapter four. I just think that'll be a clearer way of saying that. It just felt a little awkward to me.
A fair alternative for races who wanna score their events based on team results, but do not wanna limit or dictate who can be on a team, is to score based on the age and gender graded times discussed in chapter four. This puts every runner on an even playing field without requiring specific rules about who can run on each team.
All right. I was just updating my actual manuscript with my new wording. Alright, we are down in team composition, so let's get to relays
Chapter 20 Relay. I didn't like that. Chapter 20 relay much better. Another way of scoring a team race is by the team's finishing time as a relay. Similar to how you can score a team by adding up their cumulative time across a specific course. Relays do the same thing except that each runner runs consecutively rather than at the same time. Each runner may run only one time, or they may leapfrog around the course and cover multiple legs each. There will usually be an exchange zone where runners pass off to the next runner on their team before the second runner is allowed to begin their leg of the course. This may involve both runners physically being in the exchange zone at the same time, or a hand slap between the runners or passing a physical item from one runner to another. A baton is the traditional item to pass between runners, but it doesn't have to be. It could be a slap bracelet or a bib or anything else that can be carried with you. The old Mount Desert Island, 4th of July Relay used to use Barbie R dolls.
Alright, I knew this one wasn't going to work, so I, let's see. I wanna make sure it doesn't put an extra pause in here, so I'm gonna try and connect all of these words and see if that works. And obviously we don't need that registered symbol. That's a super script. If you are looking at it in the actual book. It doesn't have any sub script or Super scripts. They all just get turned into regular text for the formatting. Actually, I think I might wanna put ellipses in between these as well, but we'll see how it, how it sounds after we do it. First,
a baton is the traditional item to pass between runners, but it doesn't have to be. It could be a slap bracelet or a bib or anything else that can be carried with you. That works just
fine. Now for the one where
the actual mistake happens, old Mount Desert Island, 4th of July, relay used to use Barbie dolls. That's a relay may be on a short course, such as a track where they each in loops or even partial loops be before handing off to the next runner. It could be on a multim mile loop and runners each hand handoff at the start to finish for each loop or out and back stretch. Finally.
Let's see. I think this needs a little bit. I don't really wanna slash but start and finish might work. Okay.
A relay may be on a short course such as a track where they each run loops or even partial loops before handing off to the next runner. It could be on a multim mile loop and runners each hand off at the start and finish for each loop or out and back stretch. Finally, the race could be on a point to point course where runners go from one place to another. Depending on the logistics required, the runners may be provided with transportation to the exchange zones if they aren't exchanging right at the start and finish area. Alternatively, they may be required to provide their own transportation, which is often the case for point to point races. Where runners run multiple legs, often each team of runners will use a van to transport each other, dropping off a runner at each xang zone and then waiting, and sometimes
it just completely mispronounces a word. I could probably just regenerate this and it would be fine, but I'm gonna add that dash just to explicitly see if we can fix it 'cause. No such thing as an exchange gun or whatever it said.
Depending on the logistics required, the runners may be provided with transportation to the exchange zones. If they aren't exchanging right at the start and finish area, that sentence
sounded fine. Hopefully
the rest is good. Alternatively, they may be required to provide their own transportation, which is often the case for point-to-point races where runners run multiple legs, often each team of runners will use a van to transport each other, dropping off a runner at each x ch zone and then waiting for the previous runner to finish leapfrogging the one they dropped off until they reach the next exchange zone for longer.
Alright, so I think what we're gonna need to do is separate these into two separate ones and I am gonna go in and, you won't be able to see it 'cause I'm only sharing the tab, but I'm doing a search for the phonetic spelling of, exchange. Let's try this on Google because DuckDuckGo isn't giving me anything worthwhile. Phonetic spelling of exchange. Let's see if this works a little bit better because it does not know how to say that. I wouldn't have thought to do exchange, but let's try it and find out how it works,
depending on the logistics required. All right,
I'm gonna let that work on the next one and then we'll come back here and have it all ready to go so we can listen without having to wait again.
Often each team of runners will use a van to, depending on the logistics required, the runners may be provided with transportation to the exchange zones. If they aren't exchanging, they got right here at the start and finish area. Alternatively, they may be required to provide their own transportation, which is often the case for a point-to-point races where runners run multiple legs. this one sounds fine. That's good. Often each team of runners will use a van to transport each other, dropping off a runner at each ouch keys change zone, and then waiting for the previous runner to finish leapfrogging the one they dropped off until they reach the next UX change zone. Yeah,
that obviously did not work. But hey, we got half the paragraph done. So try exchange is two words and see if that works. Okay. Sometimes it's a little hit or miss, and if it doesn't work this time, I may just reword this for the audio book to not use the word exchange. In this particular part of the half of the paragraph,
often each team of runners will use a van to transport each other, dropping off a runner at each exchange zone, and then waiting for the previous runner to finish leapfrogging the one they dropped off until they reach the next exchange zone for longer races, especially ones that are alright. So that's
not horrible. Xchange. I probably should have just kept the space out, but it got leapfrogging much better this time than it had on one of those earlier ones. So I'm gonna stick
with that for longer races, especially ones that are a hundred, 200 miles or more. Oops.
Here's another spot where I missed turning a dash into the word two. I mentioned that was one of the things I had to do. Now you can see why
for longer races, especially ones that are 100 to 200 miles or more and may have relay teams of 10 or 12 runners. There are two general strategies used for how the race can be started. The first method is to go off of a team's gun time with the first runner of each team starting at the same time as all of the rest. The second method is to use a reverse seated start where the teams that are expected to run slower and take longer to complete the race will begin first. These races often take two or even three days for a team to complete. The advantage of using a reverse seated start is that they minimize the time that support needs to be provided by spacing out when teams will reach certain aid stations. For example, the early aid stations will still open close to the start of the race and then can close once the faster teams have moved through, which will be around the same amount of time that they would've been open, had everybody started at the same time. Aid stations later in the race, however, don't need to open up as early as the faster teams that take less time to arrive won't get to those stations until close to the time that the slower teams arrive. Meaning that the full duration of time that those stations need to be open will be a lot less later in the race. The runners will compress and finish closer to the same time.
Oh, that was a really long paragraph, so I'm glad it didn't have any glitches in it, and I think it did a pretty good job on that. Alright, we are in the home stretch. There's just a few sections left, a few chapters left.
Part four, specialty formats, chapter 21, non determinate courses. Some races don't have a predetermined route that all the runners need to follow. Orienteering is a race where participants need to find various checkpoints along the course using a map and compass. The competitors will know where the race starts, but they may or may not know where the race will end and will have to find their way to various points before they can reach the end or else they will be disqualified or otherwise penalized. Part of the challenge of competing in these races is how well they can recognize landmarks and their spatial awareness as it relates to the physical world and a map. Hash runs are very similar to orienteering because participants only know where they are going to start and then have to follow a map to various checkpoints and the eventual finish. The map, however, is not one they can carry. And if they are following the official hash house Harriers rules, then they are not allowed to use or even carry any technology that can help them navigate. The hash run starts with one runner known as the hare getting headstart. The hare will mark the course as they go. Usually that will entail drawing symbols in the trail, leaving sticks behind in certain configurations, or usually by dropping flour, chalk, sawdust, or some other easily carried and bio.
Okay, so here's another one where usually twice doesn't work, so I'm gonna have to redo this and I'm just gonna get rid of that first usually and regenerate and I'll update my manuscript.
The hash run starts with one runner known as the hare, getting a head start. The hare will mark the course as they go. That will entail drawing symbols in the trail, leaving sticks behind in certain configurations, or usually by dropping flower, chalk, sawdust, or some other easily carried in biodegradable substance to create the markings. Different types of markings are left along the trail indicating that the runners have a choice to make, or that they are going the right way or that they need to look for the trail. The hair will intentionally leave false trails that lead to dead ends. The runners have various commands they can yell out to inform others if they are on the right path or if they found a false trail. There may or may not be a beer check along the way often, but not always with some sort of treat left behind, such as beer or snacks. And eventually the running will continue until the end. A third type of non determinate courses, a search and find in which runners need to find checkpoints along the way. Usually they can run any route they want as long as they hit every checkpoint, but sometimes there will be a predetermined route from checkpoint to checkpoint. I have run a few of these centered on holidays, especially Halloween, which lends itself to this format. It could be on a loop course or an out and back trail. There may be some sort of challenge at each of the checkpoints, such as reaching into a mystery box of some sort for some sort of token, or having somebody in costume jump out for a cheap scare. A popular race in Tennessee called the Barclay Marathons could be considered a search and find in this race. The runners are given a short notice before the start of the race and have to make a copy on their own maps of what the route is for the loop. The loops are usually declared to be 20 miles long, though they are usually closer to 25 or more miles. And there are books hidden along the route to show that you ran the full loop. Each runner pulls out a page from the book corresponding to their bib number on that loop is proof.
Alright, so one interesting note I'll make about this one is usually I like to say route instead of root. And I spent a long time on one of the early chapters trying to get it to say route, and I eventually got it to say root, and I'm just like, all right, fine. It'll doesn't sound quite as good and isn't the way that I would've read that sentence, but it worked. and every other time it's shown up, it's always said it as root, except for this time it said route. I just found that interesting and wanted to
share Chapter 22 Obstacle courses. Another format of race that was especially popular in the late two thousands and early 20 teens were the various forms of obstacle courses. They began with mud races where pits were dug along the course and the participants had to basically swim through the mud before handing off to a relay runner or before the finish line, and then began adding other obstacles such as balance beams. Eventually, these style races became big business and evolved into the Tough Mudder and Spartan Race series. These series added different kinds of obstacles, such as balance beams, walls that needed to be climbed and pits that needed to be jumped over. They also added additional challenges such as physical fitness stations where you need to do pushups or burpees or contend with electrical currents or carry heavy objects. There were, of course, other
comp and again, we have another artifact, so let's try just regenerating and hope it all comes out in one go.
Eventually, these style races became big business and evolved into the Tough Mutter and Spartan race series. These series added different kinds of obstacles, such as balance beams, walls that needed to be climbed and pits that needed to be jumped over. They also added additional challenges such as physical fitness stations where you need to do pushups or burpees or contend with electrical currents or carry heavy objects. There were of course, other companies putting on these races, but they are very expensive to ensure because of the injury rates that occur at them. Many of the smaller races were acquired by these two largest brands. Tough Mutter eventually declared Chapter 11 bankruptcy in 2020 and was acquired by Spartan Races and continues to put the events on as of the time of writing. Chapter 23, Fox and Hound. Fox and Hound Races are a group of fun variations that involve runners, specifically chasing other runners. One style is a variation of the last man standing format in Last man standing. Everybody starts at the same time and runs the same predetermined course, but nobody knows how many times you will have to run that course or how far you will have to go in a fox and hound race based on last man standing, the foxes are given a headstart either in time or along the course. All of the runners will then run in a loop with the foxes trying to stay ahead of the hounds in one variation of the race. The foxes win if they catch up to the hounds and the hounds win if they catch up to the foxes. Another popular variation of this type of race is the zombie race, which works the same way except the runners are referred to as humans and zombies. If a zombie catches up to a human, then the human joins the zombie team and chases the rest of the humans. If a human catches a zombie, then the zombie is eliminated from the race. The run continues until every human has been caught or until the humans catch up to all of the zombies. It's not uncommon for people to dress up in costume for this style of race. There are also some races where a specific runner begins at the back of the pack and tries to catch as many runners as they can, whether they are at the back of the field or the rest of the field is given a headstart. There is usually some sort of celebrity or fundraising component, often with a pledged donation based upon the number of people that they catch throughout the course of the race. A race can use that same format without requiring an actual person to chase somebody down. For example, at the beat Beethoven 5K put on by the Fairbanks Symphony Orchestra runners have to complete the 5K course before Beethoven's fifth has finished being performed. Or in about 31 or 32 minutes, they give a voucher for a free concert ticket to any of that season's performances for anybody who finishes before the cutoff off Chapter 24, virtual races. With the changes in the global landscape
wrought. Now, that one sounds horrible
with the changes in the global landscape wrought by the, a virtual race.
Yep. We're just gonna regenerate that. That's completely unusable. haven't quite had that before. On, on 11 labs.
With the changes in the global landscape wrought by the COVID-19 coronavirus, virtual racing became the only method for competing in some or most areas of the world during the height of the pandemic. While they are waning a little in popularity, as people have become fatigued and as in person racing has resumed, they remain a viable option that is worth considering for most races, for at least the foreseeable future. A virtual race differs from an in-person race in that it's not relegated to a specific time or place. And the rules can vary for how they are scored and how participants can take part in them. There are a few decisions for a race to make when deciding to put on a virtual event. First and foremost, is the virtual event a primary or secondary event? During the pandemic with in-person races being canceled left and right. Most virtual events were not only the primary event, but were the only option available if runners still wanted to take part. Even before the pandemic. There were some events where the virtual option was always the primary focus and it's certainly more common now that P are used to the concept. If the virtual event is part of a series of events, are the runners allowed to run in person and virtually traditionally, a runner can't take part in, for example, both a marathon and a half marathon when the events happen at the same time. But if there is now an option to run a virtual event as well, virtual, should people be allowed to run both? If so, can they sign up for multiple virtual options if those are available? If they sign up for multiple virtual options, could they count the first half of their marathon as a half marathon and basically double dip? There is no right or wrong answer, but it is something that needs to be addressed ahead of time in a local Grand Prix.
That last sentence there just did not do it for me, so I'm gonna have to regenerate this paragraph. So let's split this up so that if we get a good one, we don't have to keep redoing everything. It is too bad too, 'cause it was good up until that last sentence.
Traditionally, a runner can't take part in, for example, both a marathon and a half marathon when the events happen at the same time. But if there is now an option to run a virtual event as well, should people be allowed to run both? If so, can they sign up for multiple virtual options if those are available? If they sign up for multiple virtual options, could they count the first half of their marathon as a half marathon and basically double dip? There is no right or wrong answer, but it is something that needs to be addressed ahead of time. That sounds much better. In a local Grand Prix series put on by the main track club during the first pandemic year, there was a clear leader in the series that had signed up and run every virtual option that was available to score extra points and which was completely within the rules. Is the extra participation more important, or should runners be limited to a single event that could not traditionally be run in tandem? If you want my opinion, I would lean towards allowing people to register for multiple virtual events, but not to count any overlapping segments in a single run for more than a single virtual event. But again, there is no right or wrong answer as long as it's stated in the rules for the event ahead of time. Where are the runners allowed to run? Most virtual versions of races allow the runner to run wherever they want, as long as they cover the appropriate distance or duration for the race. Further, rules and stipulations can be added. For example, in the virtual version of my Midwinter 10 Mile Classic, we require that virtual runners complete their 10 miles outdoors and don't allow them to count a treadmill or indoor track run. There are some races that require,
yep. There's another artifact on that one. Let me just make sure it's on this segment. Further
rules. And there are some races that require that you run on an official course, but let the runner further rules and stick. Oh, listening to the wrong one. There are some races that require that you run on.
Yeah. Sometimes I get a little confused. So one thing to be aware of is this one will play on wherever you're clicked on and regenerate if you've made any changes. This one is just gonna play the same one that you've been working with and doesn't matter where you've
clicked. Further rules and stipulations can be added. For example, in the virtual version of my mid-Winter 10 Mile Classic, we require that virtual runners complete their 10 miles outdoors and don't allow them to count a treadmill or indoor track run. There are some races that require that you run on an official course, but let the runners come and run whenever they want. At the start of the pandemic millennium running in New Hampshire created the anytime 5K, which ran from May through August. In 2020, they sent everybody a chipped bib and kept the start and finish line setup in the parking lot of the local running store with the clock running for the entire summer, runners could come and run the course with an official timed result at any time or day they wanted as long as they were careful about how and when they crossed the start and finish line. The last consideration is a combination of two separate questions that impact one another and are difficult to answer one without considering the other. How do you plan on scoring the results and how will those results be reported? Most of the methods we've discussed throughout this book,
I don't know what that was, but it did sound kind of funny.
Most of the methods
we, yeah, we are regenerating that. We don't need that laugh at the end of it, so if anyone was wondering what an artifact was, we've certainly had our share of them. While you've been watching me work tonight,
the last consideration is a combination of two separate questions that impact one another and are difficult to answer one without considering the other. How do you plan on scoring the results and how will those results be reported? Most of the methods we've discussed throughout this book can be used to score a virtual event, a straight up net time that it took from start to finish of the prescribed distance and age and gendered time, throw times out, and just provide points for participation with a,
all right, that last sentence really did not work. Let's try this again. one advantage of it taking four years chapter, most of the methods we discussed
throughout this book chapter can be used for a virtual event, a straight up net time that it took from start to finish of the prescribed distance and age and gendered time, throw times out, and just provide points for participation with a little creativity. Most methods can adapt to a virtual format. Will virtual runners appear in the same results as in-person runners? If there is a series, is the virtual run worth the same number of points or make the same impact as running in person? How will runners report their results? If runners get a chipped bib and they are running a specific court
and we've got another artifact,
how will runners report their results?
Yeah, I didn't like that one. Let's try it again.
How will runners report their results? Much better. If runners get a chipped bib and they are running a specific course that has time balance, then it's no different from any other net time race other than individual runners could start any time throughout the span of a week or month or whatever the rules state. There are also racing apps which force the runner to carry their phone with them from the start to the finish, even if they can run anywhere that they want. The advantage of either of these methods is that while it's not impossible to cheat, it is a lot more difficult. Most virtual races, however, are going to rely on self-reporting the results that could be on the honor system, and the runner just reports their time or could require the runner to upload the GPS track generated by their watch or link to the evidence on their training log that shows a map of their run.
All right, let's try putting in, let's see here and see if that comes out a little better.
Most virtual races, however, are going to rely on self-reporting the results that could be on the honor system, and the runner just reports their time or could require the runner to upload the GPS track generated by their watch. Or link to the evidence on their training log that shows a map of their run that works. 'cause of the nature of virtual racing, there can be a lot of variety and mixing and matching for how to put a race together. For example, trail Monster running puts on a three race series of in-person races at Bradbury Mountain in Panel Maine every year. If a runner completes all three races, they earn a badass hoodie at the end of the series. Not everybody can make it to all three races, however, and those sweatshirts are coveted and collected by the people that know about them. There is one other way to earn one of those hoodies and that is to run all three race courses in order back to back on the same day. It's called the Ultra Extreme Bradbury Badass, and there are a few rules. First, the runner needs to run all three courses between the start of the first race in July and the end of the last race in September. They need to be run in the correct order and they have to run them on the same calendar day. The state park is closed overnight, so it isn't legally possible to run them across two calendar days in a single 24 hour span. They're allowed to double dip with the actual races, so a runner could get to the park exceptionally early, for example, and warm up with the first two race courses and finish up with the final race on race day. However, they will need to provide proof. They manage to follow the entire route correctly, such as running with one of the organizers or providing a GPS track of the route that they ran. Virtual races have been around for a long time and are certainly here to stay. Race organizers can be very creative with how they utilize them, but do need to give them the attention they deserve and to make sure that they are still providing a safe event and have any appropriate insurance before organizing. Just as with an in-person event, they are also a great compliment to an in-person event for people who schedule won't allow them to appear in person and to help fundraise for an event. Final note from the author. I hope that you have enjoyed exploring these couple of dozen different race formats and scoring methods with me, and that you've taken some inspiration on a future race to either direct or find to run yourself. Adding in a little variety can be a lot of fun for both the race directors and the runners. Yeah,
this is a little bit awkward. I'm gonna change this a little bit. I think that's a little clearer.
I hope that you have enjoyed exploring these couple of dozen different race formats and scoring methods with me, and that you've taken some inspiration on a future race to either direct or to find and run yourself. Adding in a little variety can be a lot of fun for both the race directors and the runners. Sometimes it can be a challenge to find the appropriate method for scoring races, especially when you begin moving into team or multi-event formats. Finding a balance between the fairest and the most efficient method for scoring your race can make a big difference in a runner's experience. I love hearing about new and unique formats for races and for race scoring. If you do find something new, whether it is something missing from the pages of this book or just a new format you haven't had a chance to try before, please feel free to reach out to me and let me know how it goes.
Alright, I think that sounds pretty good. I should probably add something to the end of this. and we can just add something here, which doesn't have to be part of the original book here. I dunno why I did that. Oh, it deleted me for some reason. Weird. this is not something you wanna do a lot of original writing in, apparently. Let's see how that sounds.
Thank you for spending some time and geeking out over race formats with me. Blaine. Moore,
eh, don't need my name on that. Some of the, Some of the audio book places, you do have to explicitly let someone know they have reached the end of the book.
Thank you for spending some time and geeking out over race formats with me.
Don't really like the way that's been done, so let's try regenerating.
Thank you for spending some time and geeking out over race formats with me.
Eh, that's good enough. Alrighty. So now that we've gone through,
📚 Finding Unconverted Text
I've actually been updating everything as we go along so we can come in, let me zoom out so I get all my buttons back. This is the normal one. So we're gonna click the convert button and if we look at this, we hit convert and it looks like if we do the whole project, there's a total of 75,808 characters. And as I've been converting, as I go along, 75,635 are taken care of and there's 173 left. So I need to figure out which one is missing. 'cause I'm gonna wanna listen and proof that. Let's see. So I have my normal text and production copyright on here. I'm also explicitly letting people know that I'm using a synthesized voice. you know, you don't necessarily have to, depending on the retailer that you're gonna sell these through, but I do recommend, don't hide it. Alright, I'm just gonna go through real quick and see if I can figure out what is missing, because it looks like, and what I'm doing here is I'm looking at these black lines to see what is done and what isn't. And so far everything is done. I could try converting these one at a time until I find the one that has it. In fact, that's probably quicker. What's, what am I on? Chapter eight. So let's see. Zero. Zero. Ah yes. Here we go. The last man standing. So that was a little quicker than what I had been doing. Oh yeah, I was having some trouble with this one. Let's try regenerating this again. Now what I could have done is just gone in and converted the whole project and it would've just done this one line and then it would've been all done, but. I wanted to hear actually how it says and make sure that it's sounds good.
So that's why I was looking for it. One of the early races to use the last man standing format was the 2011 Bigs Backyard Ultra in Bell Buckle, Tennessee. The winner completed 18 laps or just over 75 miles these days. A competitive last man standing format race can continue to 70 or 80 laps. In 2021, the winner ran 85 laps over 354 miles, starting a new loop every hour for three and a half days. All right,
so if I come in here, I've now got my two convert to zero, and I could have just had it go in and take care of that for me, but I didn't want to 'cause I wanted to hear how it sounded. And I'm gonna just listen to this one again, just because I couldn't do it when I was getting all of that, feedback at the beginning of this. So I expect we're gonna have trouble over here.
Chapter 11, stage Races. A Stage race is a combination of a few different types of races. While technically a single race, it comprises multiple races that are usually run consecutively with a different leg each day. Each runner has a set amount of time to get from one point to the next each day, and their time is tracked and then added to subsequent times for each stage of the race. After runners finish for the day, they will usually have an overnight stay at a camp or other form of accommodation, which allows the runners to refuel rest and prepare for the next day's stage. Most stage races tend to be very long in duration. The self-transcendence runs starts at six 9:00 AM and runners have until midnight to finish that day's. Miles Runners attempt to cover 3,100 miles in 52 days. The marathon de sles is a six day stage race covering 251 kilometers, 156 miles through the Sahara Desert in Morocco. If you do lose
your, play here, you can minimize it just so you have a little extra room. All right, so we have a few things that we need to fix on here. First, I'm gonna explicitly say six o'clock 'cause this isn't a running time. It's a time of day. And, I'm gonna try and see a phonetic spelling of, Very helpful. Try that and see how it works. I'm gonna give 'em separate ones just to make it a little easier, so I'll have to redo it 'cause I have a feeling I'll need to do it a bunch of times. So if you're going through and you're not doing everything exactly in order, do keep notes of what you need to go back and revisit. 'cause otherwise we might be back.
Most stage races tend to be very long in duration. The self-transcendence runs starts at 6:00 AM and runners have until midnight to finish that day's. Miles Runners attempt to cover 3,100 miles in 52 days.
Eh, it wasn't bad, but did have a little hiccup in there, so I'm gonna regenerate that
one again. Most stage races tend to be very long in duration. The self-transcendence runs starts at 6:00 AM and runners have until midnight to finish that day's. Miles Runners attempt to cover 3,100 miles in 52 days. The marathon dele is a six day stage race covering 251 kilometers, 156 miles through the Sahara Desert in Morocco. Stage races usually,
I should have this saying, or 156 miles, but it actually worked pretty well and I don't want to mess that up, but this one does keep messing up. It was fine on the original, but the last couple haven't been good, so I explicitly typing it out and seeing how that works. Getting late past my bedtime.
Most stage races tend to be very long in duration. The self-transcendence runs starts at 6:00 AM and runners have until midnight to finish that day's. Miles Runners attempt to cover 3,100 miles in 52 days. The marathon dele is a six day stage race covering 251 kilometers, 156 miles through the Sahara Desert in Morocco. Stage races usually involve a team effort with one person being the runner and the rest comprising that runner's crew. They make sure that the runner has the fuel and gear that they need at various checkpoints and aid stations along the way, and take care of any needs at night so that the runner can concentrate on recovering from the day's toils and prepare for the next day.
📚 Generating and Downloading Final Audio Files
Alright, I think I am in good shape. So let's go in. Now what we're gonna do is we want to do a convert because what's happening is we have all of these individual, files. So if we were gonna
hit this, a stage race is
a combination of a, you'll see we've got, this is a 33 second clip and I could download that individual 33 second clip, but I want to have full MP threes of each chapter. And so what I'm gonna do is come in here, click on convert, and there's nothing actually it needs to do on it. It just needs to go in and put everything into downloadable audio for each. So we'll click on the convert button and this will take a few moments we can see as it goes through. And if we look it's kind of doing, you can see which one it's working on at any given time. Alright. Our project has been converted and can now be downloaded, so there's a few ways to get at it. I haven't. It was a little frustrating when I was doing my test earlier. You can come up here and you can get the project download or individual chapter download. You can see the different versions. So you can see I've actually produced it a few times. I was working on it this afternoon and early evening and now it's kinda late so we can try downloading from here. I have a feeling, oh, that was way faster this time. So let's download that now. It was, it did take a long time and one thing I found is you don't wanna close this window. If you do, your download is going to stop. So the first time it was almost 70 megabytes. So it's gonna take a little while 'cause it's very slow. And if I was pre-recording any of this and not just letting you watch over my shoulder as I did some work, then this is something where I would speed it up or cut it outta the recording altogether. 'cause who really wants to watch that. But I think that anyone who's actually watching this now, you're interested in what the process looks like. This is obviously not a polished training, it's a, here's some actual work we're doing on an actual project and seeing exactly how it works and getting a good idea for that. So you got see half the book 'cause you know, obviously I'd done half of it ahead of time, but since we've been here for probably about two hours, I think, let's see. We've been streaming for two hours now. Oh, looks like there's some, chat. someone is from Facebook, says, Hey, Blaine was just searching for an audio in my book and was looking at 11 Labs. It's my first book. First time doing audio. Can you suggest a good audio system to me as a beginner please? I can. In fact, I'll probably go in and do that in the, oh, I was showing the download. You probably couldn't see that, I'm guessing 'cause I'm only sharing the tab and not the rest of the browser window. But, it's now up to 30 megabytes. So let me go to the Apex Authors site, which you won't be able to see, but we just did a training recently, so we'll go to dedicated trainings, AI narrated audio books. And that was from January 17th. You know, not even a month old at this point. And I'll come into the Facebook group and see if I can find the comment directly there since you won't probably see it since I'm the only person in our actual room. Ah, it was Dan that had said that. So let me reply. Oops. Pause that. All right, while I'm waiting for the download, here's a link to a full training we did about a month ago on that very topic. Also, if anybody is watching this and is curious about what the best one to use, I said in that training, things change really fast. And literally this afternoon around 4:00 PM maybe a little bit earlier, I saw that find a Way Voices is making a change to their terms of service. So let me grab that link as well. I posted about it in the group, but it's probably something people that are interested in doing AI audio books should be aware of. I don't think we need to get off of, find a Way right away, but, you Sure to check this Facebook, what's happening is they're basically doing a rights grab and way overreaching. I'm hoping that they're gonna get a lot of bad feedback and their attorneys are going to walk back, and change this. It doesn't take effect until March 15th, so there's no reason to stress right away. But it is definitely something that we do not want to have to, oh, canceled. Dang Abbot. Let me try downloading this again. Maybe it just, sometimes doesn't like going through, so we're starting that over again. All right, so now I'm at three megabytes. So downloading is a little painful when you're using these 11 lab projects, but once we get it, then that'll be good. but as I was saying, I don't think you need to stress about the quite yet. We'll see what happens over the next few weeks and we will definitely talk more about it at our next hot seats, which will be, two weeks from today, I believe. Let me bring up the calendar and see we've got the next couple of months of trainings all ready to go. let's see. Today is the 15th, so yes, two weeks from yesterday. The 28th will be our next hot seats, so that'll give us, two weeks, a little over two weeks before the new terms of service go in. So my guess is that there's going to be some more updates to this whole Find a Way Voices Saga before we. Have our next hot seat session. So we'll have plenty of time to discuss what, our actual recommendations will be and see whether they backtrack or update these changes that they're making. But yeah, it's a little aggravating if they're going to stick to their guns on this particular change that they're proposing to their terms of service. Alright, and Danya is saying thank you and you are welcome. Doesn't look like there's any more comments in there. So I'm going to go back to here and wait for my download. This is the exciting part of the job. If I wasn't live streaming, I'd probably have run upstairs and got some water or brushed my teeth or something since it's past my bedtime. But I don't have any more projects to do in the near future. So I wanted to get this done today because we actually rebuilt tomorrow and, I'm wasting a lot of characters. So this project, as we mentioned a little earlier today was, about 76,000 total characters for the whole book and the way 11 Labs does there. pricing is they, every time you compute something they just go based on the number of characters. And I'm on a hundred thousand characters per month plan right now. but they did have an option where you could fill out a survey, and the first a hundred people that did that got an extra a hundred thousand, characters, which I did do. So hopefully, even though after my rebuild goes through, I'll still get to keep that extra a hundred thousand 'cause I didn't actually need them. But it, 'cause I had I think about 215,000, quotas remaining when I started this project earlier this afternoon. And, just in terms of timing, I didn't keep track of exactly how much time I've spent on it. So it's been two hours we've been together and I probably spent about an hour working on it before that, plus 30 or 40 minutes just, experimenting with voices and deciding which voice I was gonna use. Erin and our daughter also had a little bit of input on that. So what's that come out to? About three and a half hours to create a full audiobook. All right. We are done 67.5 megabytes. Let me open this up and. We take a look. It's 73 minutes and 41 seconds for the entire book. Great. So I can close this now and we can come back and go to our chapter versions and you can see the various versions we have for each of the individual chapters. So the one that I created this afternoon, and then the new one that I have here. So I'm not gonna do it tonight, but I am gonna want to go through and individually download every one of these chapters. Let's try that and just see how it works. I don't know what the file format's going to be. This is my first complete project. Let's see, 11 labs and the time, title and copyright. So on the plus side, because it's time and date stamped. actually 17, let's see, 2 16, 3 17. Yep. So that is time and date stamped to when we actually push the button. So if I hit save there and I come here, download the chapter. Downloads are much faster, which is good. So this one is three, so it's using UTC time. So because it's got the timestamp in there, I'll be able to get them all in the correct order. And then I'll just go in and rename these afterwards. And it's got the title that comes in. So I'm gonna actually go in and just download all of these to make sure that I have them. But as I was saying, let's see, 215,000. So it took about 110,000 characters total to, generate this book. And it was a 76,000 characters. So that's about 25 to 30% extra characters over the number that was in the book that I was getting done. And that's about what we estimated it was going to be back when, we did the training about a month ago and we were showing how 11 labs works. All right, I'm on chapter four. Oops, wrong button. Let's see. A little bit annoying with this coming up just with where everything is, but so be it. let's see. I'm gonna delete this older version 'cause I don't want that one. This is full. And then copyright intro about the author. What called it? Oh, interesting. Doesn't have the full title. So that's chapter 1, 2, 3, 4. All right. Just making sure I didn't miss one. Yep. This is the exciting part of doing the work. Bit of rote, but just wanna make sure I have everything before I turn in for the night, and then I don't have to worry about it in the morning. Now, I don't know if there's very many people watching or not. I haven't actually checked, but if anybody is interested in doing, walkthrough and showing the, nitty gritty of how the work gets done, I can certainly do this for other projects in the future. Just let me know if you found this worthwhile or. Presumably you won't tell me if you didn't find it worthwhile, because why would you be still listening to me talk two hours later if you didn't enjoy it or didn't find it useful at all? And for anyone that has been watching, I hope you enjoyed geeking out with me over talking about racing. I didn't show you that particular one, but in the, about the author section, I talk about how I put on, I organize Erin hates, when I say put on, I organize between, 14 and 20 races a year, depending on the year. usually now towards the 14 end for 15 races. But, all righty, I'm gonna have to count these afterwards to make sure I've got 'em all. 'cause with the various popups that are coming up, it is getting a little confusing, but almost to the bottom.
Okay.
All righty. So we've got 24 chapters plus the final note. That's 25, 26, 27, 28, and then the full download is 29 files. So if I click select all 29 items, so that should be right now, let me just make sure I didn't do two in a row. Race scoring title, copyright intro. Actually, if I just do a list, I can bring this in and I can share that window with you. So we're looking through at all of our files here. We've got race scoring title and copyright. Introduction about the Author Place, net time. This is actually gun time place if there's a slash it doesn't get anything before the slash net time, age and gender awards. Gender grading. Reverse seated start criterion distance. I'm gonna bring this up on the side. intro out the other please. Handicapped gender grading. Reverse seated start. I'm scrolling in the background, not sharing that. So you probably can't see it, but, oh, I've got the 6, 7, 8, 9, 10, 11, 12. That makes it easier. 13, 14, 15, 16. Cross country scoring would be 17. Yep. 18, 19, 20, 21, 22, 23. Looks like I did it right. So one thing to know, if you put a slash in your chapter title, it does mess with the, naming convention a little bit, but that's okay. It's got them all in the right order. I've got 'em all. And if we come in and look at our columns, then we can see that we are at 73 minutes and 41 seconds. So thank you for joining me for the night and, it looks like there were some more comments, but they were all from me, except one person did say, that's worthwhile. Please do more. So, all right, maybe we'll do that again. I do plan on, creating a. A version of this book using Affinity Publisher. So I may live stream that we'll be doing a training on it as well, but having the live stream could be useful just to see all the, minutia rather than talking about it actually being able to see that. So I'm glad, at least one person found it useful. And in the meantime, have a good night and I'll, either link to this, Facebook Live or maybe even put it on a bonus page if anybody does want to take a look at it in the future. All right, have a good night everybody.
Click anywhere within this unedited transcript to jump directly to that part of the work session. This transcript includes only Blaine's narration as he is working on the project, and does not include any of the content from his book that is spoken by the Eleven Labs synthetic voice.
Okay, so this afternoon and this evening I was working on editing the audio book made through 11 labs for a book. I just finished, writing and getting edited, so I thought I would share that with you because it's been an interesting process. Let me find the right tab so I can bring that up and share it with you. Okay, share screen. Okay, so now this book is about how to put on a road race and is my experiences in doing so. And so what I've been doing is there are, let me go back a little bit here. So we have, 28 chapters total. There's 24 chapters plus, introduction, final note from the author, my title and copyright, et cetera. And so I've been going through and I had gone in, chosen the, voice that I wanted to use and went and, did some testing, tested a few different voices, and now I'm in the process of proofing the whole thing and it does all need to be proofed. So, let's see, where did I leave off? I was on stage races, I believe is where I left off. So I'm just going to show you the process of what I've been doing and the changes that I've needed to make in order to get things to work. You do have to listen to the whole book. I'm almost halfway through it right here, and it's about 75 minutes total for the whole thing, but I will have to recreate some of it, and I won't necessarily stream the entire rest of this, but I just thought it might be interesting to show some of the changes that I have to make in some of the mistakes and thing, that it gets wrong and some of the things that it gets right.
📚 Technical Difficulties
All right, I'm going to actually pause this because I'm getting a pretty bad echo, so let me go in here and turn that off. Maybe that'll make a difference. Okay. Nope. Still getting an echo.
📚 Lessons Learned From the First Half
All right, so I do have a really bad echo and I can't seem to figure out how to get that to go away. So rather than letting you follow along, I'll just go back and show some of the changes that I've made and turn this off before I try proofing it again. 'cause I'm not gonna be able to do so effectively otherwise. So if we come in, some of the things that it's done really well, this particular one, I actually use this my paragraph as my sample paragraph, what I tried a bunch of different voices for. And it did a really good job on the first one with this voice. But then when I actually came in and redid the whole project after choosing the new voice that I wanted to use, it didn't sound nearly as good, but I was able to regenerate it and got that to come through pretty well. Let's see, some of the other changes, I had to split this paragraph up because it was having a lot of trouble with this one, this particular one, especially the second half of the paragraph. So one of the neat things about this is you don't need to have it matching exactly what's in your book. So you can go in and change a few things. So for example, I took out the beta breakers and then it had five kilometers in parentheses, and then after a run for passive river, that one's, it was, like two to 10 kilometers or something along those lines. And I also changed passive to a phonetic spelling of it to make it easier to pronounce it correctly. but it was having a lot of trouble with the individual, having the parentheses with the, the kilometers in there. One thing I did notice too is there is a few places where despite having, multiple people edit it, and I've personally gone through this book at least seven or eight times at this point, if not more. I am finding that there are some typos in there or some missing words. And, I've been going through and updating my, original manuscript alongside of it for when I find an actual error that I wanna fix in the print version and in the ebook version, I'm making notes of those and updating 'em as I go along. Let's see, what else did we find? Here was another one where I had to make some changes. for spacing wise. I did split this paragraph up into two paragraphs just to gi force it to give me a little bit more of a pause. And then originally this was one dash 99 and 200 to two nine to dash 299, et cetera. So I did change these one and at change the dashes to the word two, which made it a little bit clear exactly what I was going for. one thing I did like was, at one point I, when the first time I used RFID, it's a term related to the technology we're talking about in this book for this particular chapter. And it did have, a pretty nice, segue into defining it 'cause I just had it in parentheses right after. And that worked pretty well. Let's see, what did we have? Here's another spot where instead of 19 dash under, I changed it to 19 and under and added the twos in here instead of dashes. let's see. this one, I'm not sure if I'm gonna keep this paragraph. I may end up going back and changing this because I use U-S-A-T-F, which is the governing body multiple times in this paragraph, and then again in this next one. And it was kind of grading on me, so I'm probably gonna come back and make a change to this and I'll update the actual book for that, but I haven't done that one yet. here's another one where. I changed the punctuation. I added an ellipses instead of, a comma just to add a, put a little separation between these and that helped it figure out the right, way to, to say that particular line. and again, we have our dashes and I also added, punctuation, the ellipses on the end of this here. And I added this and in here on its own paragraph, just to provide that little bit of separation in here and to improve the way that it sounds. Now in the book, this is a bullet, bullet list. so you would see literal bullet points, and I don't even think I have a period or anything on the end of each line, but I wanted it to sound a little bit better with, for the person who's listening to this book. All right. Let's see. I know there was a few other items that I found that kind of tricked it a little bit. let's see. We're getting pretty close to where I had caught up. yeah, I think this one is probably a good one where I had to do that. So in this one. It's saying 1.0, I could change this to one, 0.0 and then type out point and type out zero or do zero. But 1.0 sounded, sounded fine to me. we did have, I think it might be the next one where I had to explicitly, oh, here's one, 5.72 miles. So you'll see this one. I, it was saying something like 5.72, and obviously that's not what I wanted, so I did have to explicitly spell this one out in order to get that to work. I also had a few spots where I had to change an actual word to something different just because it could not figure out how to, get it right. Let me see. I think it was in this one. I had, yeah. So the original sentence here, one way to accomplish this is to assign people to waves by, by their bib or something along those lines. here, I've got the book open. I can read off the exact one. We're in net time. Yes. One way to accomplish this is to assign people to waves by bib number, for example, by having a hundred people start at each minute. And I could just not, I mean, it says bibs right here, just fine. It was saying bib number just fine in this spot, but for this one, for whatever reason, it just could not pronounce the word bibs. So I had to change it, and just re reword it completely. I just gave up trying to get it to work. Let's see. We looked at that one. another one, it was actually doing an okay job. This was saying 18 colon, 30 and 20 colon 20, because these are times for people running their races. but for the most part, since there were other issues that I wanted to adjust inside of these paragraphs, I just took out the colons and left it as a space. I think it just tended to come out a little bit better. and the pacing for where it would say those words, it slowed down a tiny bit with a space instead of a colon. And I liked the sound of that better, and I didn't do that in every single one of them. You notice this one I never got around to changing. but I did do it on here. So it's basically on a case by case basis. Everything was converted up front, so I was able to come in and see how exactly it sounded and when I did need to make a change in that paragraph. Anyway, I just updated, I mean, I obviously missed this one when I was doing it, but it still came out sounding fine. So I didn't, didn't worry about it. And, This one down here did give me a little bit of an effort. I did wind up adding some random ellipses in here just to, used to be commas just to get this to, to speak it in a way that, that was worth listening to. another way you can actually, instead of creating new paragraphs, you can add these little dividers. but for the most part, I had more luck using actual para, changing it into new paragraphs and that seems to work just fine. Here's another one, 4.167. originally was just all nu, was an actual decimal, but I think that catches up to where we are and that's most of the, types of changes in editing that I've been doing. one thing that I did notice is the download. If we come in here, we can see our project versions and we'll see. It does take a long time to create that even after we've gone in and run our convert and done the whole, even if you do the whole thing at once, this was actually pretty quick to, to generate all of the text. But creating the MP three took a lot longer and it was, took me a little while to figure out how to get it downloaded, just 'cause I wanted to see how long the whole thing was as a project, but. You can come in here and download the whole project, or you can do it on a chapter by chapter basis by looking at the chapter versions and downloading that way. so I'll probably wind up doing both once I'm done, but most likely I'll be working with the individual chapters and depending on how those come out, I will wanna have a single MP three that's the entire book without having separate chapters, and I'll just cre create that myself. But I wanna, I may want to add a little bit of room tone in between the chapters or something along those lines. That'll be something I'll experiment with as I am going along a little bit later. But anyway, hopefully people found this, mildly interesting at least. Sorry I can't just go in and let you guys watch over my shoulder, but that echo that's coming in because I'm sharing the audio and I've got my headset on to listen to, was just a little too distracting and would take me way too long to actually finish doing the work.
📚 Technical Difficulties Solved!
So I'm going to log off of here and I will see y'all next week on Erin's training on Tuesday. Oh, I think I figured out why my. Room is echoing. I am actually logged into our room in two places. Find the other one. Here it is. Now I'm gonna just try this again and maybe you can watch over my shoulder. So I think I was just doing that again and I stopped sharing my screen so you guys aren't actually able to follow along with that. So let me put that back on. So sorry about that. So just jump, 18 minutes forward and I'll leave a note. I'll try to remember to leave a note. So if anyone wants to watch the actual process, skip the first 18 minutes of this.
📚 Workflow: Follow along while Editing an Audiobook in ElevenLabs
And since you guys, if you have been watching this, probably listen to this like three or four times now. I'll just skip that one and come back after I stop live streaming. Just to give you something a little different to listen to Yeah, I don't like the way that it said that. So I'm gonna just try this again. Just takes a moment once I chapter 12, Okay, so this is another spot where it didn't sound quite right. So what I'm gonna do is I'm gonna actually try bring a ellipses in here and then have it just regenerate all three of these sections and see if it sounds a little bit better. So one thing, hasn't come up here and we might see it as we're going through, but I've also found sometimes just little artifacts will come in and appear or it was generating that as I went along. but one of the main reasons you do have to listen to the whole thing is so that you can hear those artifacts, when you're going through. 'cause you don't wanna leave those in the final, version. So here, why don't I zoom up a little bit just to make it easy for people to see if they wanna read along as I'm going. Alright, so now we have this all set. I think that delay seemed All right, so that was a pretty straightforward, chapter. We can move right on to the next one, and this starts a new part of the book. So what I did was instead of having a separate chapter for, the parts, different parts of the course, of the, of the book, 'cause the book is split into single race single runner. Part two is multi race single runner. Part three is a single race, but a team of runners. And then part four is just specialty formats that don't fit into any of those. I wound up setting these as all is heading one to just kind of give the, this the same gravitas as the chapter, declaration. And then, I figured it's only a couple of seconds for each one. So I just put those at the beginning of the sections and didn't worry about having their Alright, I don't know why it didn't keep going. Just wanna make sure that's not a huge extra delay and it was just having to redo this for whatever reason. Yeah, this one has quite a few problems, so I'm gonna split this into multiple paragraphs here so that we can get a good paragraph, a good reading on each one of those and not have to regenerate at all at once. 'cause you do have. Every time I regenerate, I am using some of my quota and I'm not gonna go anywhere close to this much. The whole book is about 76,000 characters total, about 13,000 words. so I obviously have plenty left, but I've already used, about 90,000 so far and I'm halfway through the book. So it's, probably will wind up with around a hundred to 110,000 characters left, is my guess. Alright, those are generated, let's play it Still not a huge fan of the way that it's saying speed workout over here. Let me try putting a dash in here and see if that helps a little bit. Another option is we can go into these foot voice settings and try things. Oh, it looks like it saved my style. Exaggeration. This should be much lower. I thought that was gonna be on a paragraph by paragraph basis, but apparently it's been using that since then. So that's a good thing to know. If you do change your paragraph settings, it seems to carry over for any future ones that you're updating. but let's see. Good thing I checked. Cumul. Alright. I think that I'm gonna regenerate this. I think it sounds okay, but it could sound better. I haven't tested this yet, but I'm gonna try seeing if I can give a little more emphasis by putting that there. I don't know if that will work, so let's give it a try and see. definitely didn't sound good. So, let's get rid of that and you'll notice as a side note, it tells you what's been converted already. See the black lines here and then we can see when I make changes, then I need to still reconvert these particular pieces so you can see, where exactly you still need to get some work through. Let's try doing both of these again and see how well that works. Sometimes you just have to regenerate three or four times and you wind up with, with something that does catch it pretty well. set now. Let's do this one again. Yeah, that sounds way better than the original version of this particular sentence, but thankfully was good the first time. Pretty bad the last time, but the emphasis on has is much better with this one. So Alright. Not in love with this particular one, but it sounds a little awkward with the way I phrased it. So I'm gonna just read this out loud a little bit to myself and try and see what I was thinking when I wrote it this way. And I may reword this, which will involve me actually updating the book itself as well, not just the way I'm having it read in the audio. So as a general rule, cumulative time works best when races are all the same distance, if not on the same course, actually. Like the way that it, it read that. So hopefully it keeps that same intonation that it used that way. Every race is on an even playing field and one specific race give a certain type of runner. Yeah, I think give is the wrong word I'm trying to use here. Yeah. So no matter how many times you've edited something, you're always missing a word here. So I'm gonna try splitting these a little bit and regenerate these and I'll go in and update at the same time in my original manuscript. I guess I should have split this up so I didn't have to regenerate it. We'll do that first, but yeah, it's let's put a dash there. See if there's a way I can get this so that it gets it right. Let's see. There's a big difference between being beaten by one minute and a one mile race and somebody being beaten by one minute and a half marathon after. Let's see if adding that comma fixes that. There is a bit of waiting in this process Alright, now this paragraph I'm trying to decide, it's some things it does really well. A few things, it's just kind of adequate. So I may revisit this and come back and split this into multiple paragraphs and do a few generations on it. not in love with it, but it's not as bad as some of the others have been. So for now, I think I'll just move along. So I do notice when I click on a new chapter, it leaves the position of whatever you're looking at, down at the bottom. So you do have to scroll back up, which is a little bit of an annoyance in the user interface, but it's not the end of the Alright, that seemed really fast. Yeah, I think I'm gonna regenerate this. It just seems to be saying it super fast and I don't really like that there. That's much better. So, one thing you may have noticed, I pressed this button. This, if you have it turned off, it'll just read the paragraph you're on and if you turn it on, it'll let you go through and listen to each paragraph one after another and some. See it does pronounce things the same words differently. You'll notice that, on a chapter or two ago, I had to have a dash to put half a marathon together. 'cause otherwise it was sounding kind of funny. But this time it got it no problem. So that's good. Alright, now you notice it had this seven knots, right here, which is obviously not correct. so I think what I'm going to do here is I. I'm gonna just add in the one mile and a 91 46 half marathon. That should be fine. would have an average pace of five and it's, and 30 seconds per mile in the one mile race and seven minutes per mile in the half marathon. So other than, a slight change, up here, it, the five minutes and 30 is just another way of saying that, so I'm phonetically putting it out. All right. That one should probably be fine there, and let's just try these again and see how it comes out. I'm putting me in the, into the actual manuscript because I think I, that sounds a little bit that sounds much, much better. So let's generate this one. So it's does not know how to deal with zeros in the times seven SUD one is a new one That's kind of interesting. Wasn't really in love with that anyway. So why don't we generate these 7 0 1 and a half marathon and we're just gonna preemptively assume this is gonna need fixing two. the rest of these should be good, so let's try it again. just needs to be regenerated. I had it right the first time, but you can't keep the first half of it unless I wanted to export and edit it manually, which I don't wanna do. right, I'm gonna just get rid of that colon. I don't want it to be 91.46. I just want 91 46 half marathon. All right. Does not seem to wanna be cooperating. Here we go. All right. That chapter seemed okay on its own okay, so one of the issues I found in my editing was, I use the word you multiple times to refer to different people, race directors, the races themselves, the runners. I. It wasn't very clear. So that was one of my big editing tasks was to go in and obviously I, missed a spot right here. So how point scoring works is that runner earn a certain number of points based on what place they come in during the race. so this will fix that particular paragraph. So while that is regenerating, I'm actually going to have to go in and fix my manuscript as well. So this is a nice editing process too if you do it before the book's been published, like I'm doing here. to earn points, I kinda like it though, is that alright, this particular paragraph, the second half of it was fine, but this first sentence and the transition was kind of awkward. So I'm going to split this up a little and just do the, have it regenerate these, Let's do this one. and, all right, this one, wherever it was up here, could have been a little bit of a glitch. And just in terms of putting it in, I'd wanna listen again, but it's definitely a little more awkward down over here, so there's no point in listening to the whole thing. We'll just regenerate this whole paragraph, whole sentence. It's a very long sentence. just rewind a little. You don't have to actually start at the beginning of the paragraph. You can use the scrubber bar if that wasn't obvious. I just wanna see if that's a problem with it. Just loading the second half of that a little bit later and that's where the lag is. I'm assuming that's the case, but we'll find out. yeah, so that was just a glitch. So what happened if you were listening is it had basically generated the first half of the sentence and started playing it, and it hadn't quite finished the second half to add it on. So that little hiccup that we had in between the all and the six was just a little bit of lag while it finished generating and gave us that, second half of the sentence. So when we re-listened to it, it sounds fine the way that it is, and this generation's much better. So let's keep going on. Now, way and oops, I'm ing the wrong button, this is just replaying the one that I was already on. What I actually wanted to do was hit the play button up here. that might mean that 10 points. Apparently I deleted the 10 somewhere in my editing process. So generate that and let's fix our manuscript. My manuscript Alright, so one thing that I noticed is that this missing 10 points, was actually in my manuscript. I don't know why it disappeared when I imported my epub could be a problem with, I was using vellum to create the eub. So maybe for some reason it just disappeared from there. So that's something I wanna keep a, let's see. actually, am I looking in the right spot? Nope. I was actually looking at this one here, so I was off the wrong paragraph. So nevermind. yes I am in fact missing the 10. So it was not a problem with 11 labs, it was just a problem with my looking in the wrong paragraph. Alright, so this is one of the nice things as you can get in a little bit extra editing and making sure you have things right and you're not putting in simple typos. one of the suggestions that we have in our self editing is to go in and actually read your book out loud or use, your computer to have it read to you. And this is basically just another way of having the computer read it to you except that we're getting a recording out of it as well. where, alright, so this is a situation where I think we need to have a little pause. So I'm gonna go in here and I'm gonna try seeing if I can get two pauses that way. Let's try and see how that that didn't really work 'cause you'll notice that there was a little bit of an artifact glitch. Let's try regenerating and see how that sounds. Okay, that sounds better. alright, and now we notice there's a little artifact at the very beginning of this one. little glitch is at the end of this one. So even though this paragraph sounds pretty good, we're going to want to go in and regenerate it. Usually when it's a little artifact like that, then regenerating comes out just fine and there isn't a problem the second time, but you do have to check and figure out where exactly that sounded better the second time anyway, so that's good. not a huge fan of the end of it, but the rest of that sounded fine, so I'm gonna leave it as is. Now we're in part three. Alright, so I don't, sometimes I want the slash to be ignored. Sometimes I want it to be said. Sometimes I want it to be ignored and it just puts something random in there. So in this case, I want to pl explicitly say positional slash cross country scoring because it's more of a either or type of situation. really like the first few, so I'm gonna redo this one. Still not a huge fan. So let's regenerate that one and I'm gonna try and explicitly get a pause in there and see how that sounds. added an extra word in there for some reason, this.dot does not always work. Let's try Semicolon and see if that works a little bit better. I haven't had it just putting in a random word like it seems to be doing this time, but you know, the, other pieces of 11 Labs has a way where you can explicitly put a pause in, but that particular feature doesn't work in these projects where we're working on a whole audiobook at once. All right, so I definitely think that this came out. I'd still not a huge fan. In Yeah, so I'm still not a huge fan of this right here, but I feel like it got the end of the sentence right. So I'm gonna stick with what we've got now. Alright, we're getting close to the end. Chapter I don't think I'm gonna regenerate it. I wasn't a huge fan right in here with the way that it paced there, but, it's okay. Alright, so I'm assuming that this is where that little artifact came in. 'cause it sounded like it was stuck on the vat. or not, looks like it is at the end of this one So we're gonna have to regenerate this whole thing 'cause of that, the very end. I, other than that little artifact at the end, I liked the first time that this came through, but the second time, I really like how it emphasized the word elite there. It gave me a, made me wanna chuckle a little bit. All righty, there's another impact. Yep. There is a little artifact on the end of that one, so we'll have to regenerate this one too. And this is why we have to go through and proof this. So, as I said, this was originally about 75 minutes long for the, that could mean but I'm probably spending good three hours at least on, on doing this when you count the first half and figuring out what I'm doing and everything. this one sounded fine in terms of the audio, but this paragraph is a little bit awkward, so I think I'm gonna wanna rewrite, write it a feral turn of races. Score their events based on team results, but do not wanna limit or dictate who can be on a team is to score based on the age and gender grade times, discussed in chapter four. I just think that'll be a clearer way of saying that. It just felt a little awkward to me. All right. I was just updating my actual manuscript with my new wording. Alright, we are down in team composition, so let's get to relays Alright, I knew this one wasn't going to work, so I, let's see. I wanna make sure it doesn't put an extra pause in here, so I'm gonna try and connect all of these words and see if that works. And obviously we don't need that registered symbol. That's a super script. If you are looking at it in the actual book. It doesn't have any sub script or Super scripts. They all just get turned into regular text for the formatting. Actually, I think I might wanna put ellipses in between these as well, but we'll see how it, how it sounds after we do it. First, fine. Now for the one where Let's see. I think this needs a little bit. I don't really wanna slash but start and finish might work. Okay. it just completely mispronounces a word. I could probably just regenerate this and it would be fine, but I'm gonna add that dash just to explicitly see if we can fix it 'cause. No such thing as an exchange gun or whatever it said. sounded fine. Hopefully Alright, so I think what we're gonna need to do is separate these into two separate ones and I am gonna go in and, you won't be able to see it 'cause I'm only sharing the tab, but I'm doing a search for the phonetic spelling of, exchange. Let's try this on Google because DuckDuckGo isn't giving me anything worthwhile. Phonetic spelling of exchange. Let's see if this works a little bit better because it does not know how to say that. I wouldn't have thought to do exchange, but let's try it and find out how it works, I'm gonna let that work on the next one and then we'll come back here and have it all ready to go so we can listen without having to wait again. that obviously did not work. But hey, we got half the paragraph done. So try exchange is two words and see if that works. Okay. Sometimes it's a little hit or miss, and if it doesn't work this time, I may just reword this for the audio book to not use the word exchange. In this particular part of the half of the paragraph, not horrible. Xchange. I probably should have just kept the space out, but it got leapfrogging much better this time than it had on one of those earlier ones. So I'm gonna stick Here's another spot where I missed turning a dash into the word two. I mentioned that was one of the things I had to do. Now you can see why Oh, that was a really long paragraph, so I'm glad it didn't have any glitches in it, and I think it did a pretty good job on that. Alright, we are in the home stretch. There's just a few sections left, a few chapters left. Okay, so here's another one where usually twice doesn't work, so I'm gonna have to redo this and I'm just gonna get rid of that first usually and regenerate and I'll update my manuscript. Alright, so one interesting note I'll make about this one is usually I like to say route instead of root. And I spent a long time on one of the early chapters trying to get it to say route, and I eventually got it to say root, and I'm just like, all right, fine. It'll doesn't sound quite as good and isn't the way that I would've read that sentence, but it worked. and every other time it's shown up, it's always said it as root, except for this time it said route. I just found that interesting and wanted to comp and again, we have another artifact, so let's try just regenerating and hope it all comes out in one go. wrought. Now, that one sounds horrible Yep. We're just gonna regenerate that. That's completely unusable. haven't quite had that before. On, on 11 labs. That last sentence there just did not do it for me, so I'm gonna have to regenerate this paragraph. So let's split this up so that if we get a good one, we don't have to keep redoing everything. It is too bad too, 'cause it was good up until that last sentence. yep. There's another artifact on that one. Let me just make sure it's on this segment. Further Yeah. Sometimes I get a little confused. So one thing to be aware of is this one will play on wherever you're clicked on and regenerate if you've made any changes. This one is just gonna play the same one that you've been working with and doesn't matter where you've I don't know what that was, but it did sound kind of funny. we, yeah, we are regenerating that. We don't need that laugh at the end of it, so if anyone was wondering what an artifact was, we've certainly had our share of them. While you've been watching me work tonight, all right, that last sentence really did not work. Let's try this again. one advantage of it taking four years chapter, most of the methods we discussed and we've got another artifact, Yeah, I didn't like that one. Let's try it again. All right, let's try putting in, let's see here and see if that comes out a little better. this is a little bit awkward. I'm gonna change this a little bit. I think that's a little clearer. Alright, I think that sounds pretty good. I should probably add something to the end of this. and we can just add something here, which doesn't have to be part of the original book here. I dunno why I did that. Oh, it deleted me for some reason. Weird. this is not something you wanna do a lot of original writing in, apparently. Let's see how that sounds. eh, don't need my name on that. Some of the, Some of the audio book places, you do have to explicitly let someone know they have reached the end of the book. Don't really like the way that's been done, so let's try regenerating. Eh, that's good enough. Alrighty. So now that we've gone through,
📚 Finding Unconverted Text
I've actually been updating everything as we go along so we can come in, let me zoom out so I get all my buttons back. This is the normal one. So we're gonna click the convert button and if we look at this, we hit convert and it looks like if we do the whole project, there's a total of 75,808 characters. And as I've been converting, as I go along, 75,635 are taken care of and there's 173 left. So I need to figure out which one is missing. 'cause I'm gonna wanna listen and proof that. Let's see. So I have my normal text and production copyright on here. I'm also explicitly letting people know that I'm using a synthesized voice. you know, you don't necessarily have to, depending on the retailer that you're gonna sell these through, but I do recommend, don't hide it. Alright, I'm just gonna go through real quick and see if I can figure out what is missing, because it looks like, and what I'm doing here is I'm looking at these black lines to see what is done and what isn't. And so far everything is done. I could try converting these one at a time until I find the one that has it. In fact, that's probably quicker. What's, what am I on? Chapter eight. So let's see. Zero. Zero. Ah yes. Here we go. The last man standing. So that was a little quicker than what I had been doing. Oh yeah, I was having some trouble with this one. Let's try regenerating this again. Now what I could have done is just gone in and converted the whole project and it would've just done this one line and then it would've been all done, but. I wanted to hear actually how it says and make sure that it's sounds good. so if I come in here, I've now got my two convert to zero, and I could have just had it go in and take care of that for me, but I didn't want to 'cause I wanted to hear how it sounded. And I'm gonna just listen to this one again, just because I couldn't do it when I was getting all of that, feedback at the beginning of this. So I expect we're gonna have trouble over here. your, play here, you can minimize it just so you have a little extra room. All right, so we have a few things that we need to fix on here. First, I'm gonna explicitly say six o'clock 'cause this isn't a running time. It's a time of day. And, I'm gonna try and see a phonetic spelling of, Very helpful. Try that and see how it works. I'm gonna give 'em separate ones just to make it a little easier, so I'll have to redo it 'cause I have a feeling I'll need to do it a bunch of times. So if you're going through and you're not doing everything exactly in order, do keep notes of what you need to go back and revisit. 'cause otherwise we might be back. Eh, it wasn't bad, but did have a little hiccup in there, so I'm gonna regenerate that I should have this saying, or 156 miles, but it actually worked pretty well and I don't want to mess that up, but this one does keep messing up. It was fine on the original, but the last couple haven't been good, so I explicitly typing it out and seeing how that works. Getting late past my bedtime.
📚 Generating and Downloading Final Audio Files
Alright, I think I am in good shape. So let's go in. Now what we're gonna do is we want to do a convert because what's happening is we have all of these individual, files. So if we were gonna a combination of a, you'll see we've got, this is a 33 second clip and I could download that individual 33 second clip, but I want to have full MP threes of each chapter. And so what I'm gonna do is come in here, click on convert, and there's nothing actually it needs to do on it. It just needs to go in and put everything into downloadable audio for each. So we'll click on the convert button and this will take a few moments we can see as it goes through. And if we look it's kind of doing, you can see which one it's working on at any given time. Alright. Our project has been converted and can now be downloaded, so there's a few ways to get at it. I haven't. It was a little frustrating when I was doing my test earlier. You can come up here and you can get the project download or individual chapter download. You can see the different versions. So you can see I've actually produced it a few times. I was working on it this afternoon and early evening and now it's kinda late so we can try downloading from here. I have a feeling, oh, that was way faster this time. So let's download that now. It was, it did take a long time and one thing I found is you don't wanna close this window. If you do, your download is going to stop. So the first time it was almost 70 megabytes. So it's gonna take a little while 'cause it's very slow. And if I was pre-recording any of this and not just letting you watch over my shoulder as I did some work, then this is something where I would speed it up or cut it outta the recording altogether. 'cause who really wants to watch that. But I think that anyone who's actually watching this now, you're interested in what the process looks like. This is obviously not a polished training, it's a, here's some actual work we're doing on an actual project and seeing exactly how it works and getting a good idea for that. So you got see half the book 'cause you know, obviously I'd done half of it ahead of time, but since we've been here for probably about two hours, I think, let's see. We've been streaming for two hours now. Oh, looks like there's some, chat. someone is from Facebook, says, Hey, Blaine was just searching for an audio in my book and was looking at 11 Labs. It's my first book. First time doing audio. Can you suggest a good audio system to me as a beginner please? I can. In fact, I'll probably go in and do that in the, oh, I was showing the download. You probably couldn't see that, I'm guessing 'cause I'm only sharing the tab and not the rest of the browser window. But, it's now up to 30 megabytes. So let me go to the Apex Authors site, which you won't be able to see, but we just did a training recently, so we'll go to dedicated trainings, AI narrated audio books. And that was from January 17th. You know, not even a month old at this point. And I'll come into the Facebook group and see if I can find the comment directly there since you won't probably see it since I'm the only person in our actual room. Ah, it was Dan that had said that. So let me reply. Oops. Pause that. All right, while I'm waiting for the download, here's a link to a full training we did about a month ago on that very topic. Also, if anybody is watching this and is curious about what the best one to use, I said in that training, things change really fast. And literally this afternoon around 4:00 PM maybe a little bit earlier, I saw that find a Way Voices is making a change to their terms of service. So let me grab that link as well. I posted about it in the group, but it's probably something people that are interested in doing AI audio books should be aware of. I don't think we need to get off of, find a Way right away, but, you Sure to check this Facebook, what's happening is they're basically doing a rights grab and way overreaching. I'm hoping that they're gonna get a lot of bad feedback and their attorneys are going to walk back, and change this. It doesn't take effect until March 15th, so there's no reason to stress right away. But it is definitely something that we do not want to have to, oh, canceled. Dang Abbot. Let me try downloading this again. Maybe it just, sometimes doesn't like going through, so we're starting that over again. All right, so now I'm at three megabytes. So downloading is a little painful when you're using these 11 lab projects, but once we get it, then that'll be good. but as I was saying, I don't think you need to stress about the quite yet. We'll see what happens over the next few weeks and we will definitely talk more about it at our next hot seats, which will be, two weeks from today, I believe. Let me bring up the calendar and see we've got the next couple of months of trainings all ready to go. let's see. Today is the 15th, so yes, two weeks from yesterday. The 28th will be our next hot seats, so that'll give us, two weeks, a little over two weeks before the new terms of service go in. So my guess is that there's going to be some more updates to this whole Find a Way Voices Saga before we. Have our next hot seat session. So we'll have plenty of time to discuss what, our actual recommendations will be and see whether they backtrack or update these changes that they're making. But yeah, it's a little aggravating if they're going to stick to their guns on this particular change that they're proposing to their terms of service. Alright, and Danya is saying thank you and you are welcome. Doesn't look like there's any more comments in there. So I'm going to go back to here and wait for my download. This is the exciting part of the job. If I wasn't live streaming, I'd probably have run upstairs and got some water or brushed my teeth or something since it's past my bedtime. But I don't have any more projects to do in the near future. So I wanted to get this done today because we actually rebuilt tomorrow and, I'm wasting a lot of characters. So this project, as we mentioned a little earlier today was, about 76,000 total characters for the whole book and the way 11 Labs does there. pricing is they, every time you compute something they just go based on the number of characters. And I'm on a hundred thousand characters per month plan right now. but they did have an option where you could fill out a survey, and the first a hundred people that did that got an extra a hundred thousand, characters, which I did do. So hopefully, even though after my rebuild goes through, I'll still get to keep that extra a hundred thousand 'cause I didn't actually need them. But it, 'cause I had I think about 215,000, quotas remaining when I started this project earlier this afternoon. And, just in terms of timing, I didn't keep track of exactly how much time I've spent on it. So it's been two hours we've been together and I probably spent about an hour working on it before that, plus 30 or 40 minutes just, experimenting with voices and deciding which voice I was gonna use. Erin and our daughter also had a little bit of input on that. So what's that come out to? About three and a half hours to create a full audiobook. All right. We are done 67.5 megabytes. Let me open this up and. We take a look. It's 73 minutes and 41 seconds for the entire book. Great. So I can close this now and we can come back and go to our chapter versions and you can see the various versions we have for each of the individual chapters. So the one that I created this afternoon, and then the new one that I have here. So I'm not gonna do it tonight, but I am gonna want to go through and individually download every one of these chapters. Let's try that and just see how it works. I don't know what the file format's going to be. This is my first complete project. Let's see, 11 labs and the time, title and copyright. So on the plus side, because it's time and date stamped. actually 17, let's see, 2 16, 3 17. Yep. So that is time and date stamped to when we actually push the button. So if I hit save there and I come here, download the chapter. Downloads are much faster, which is good. So this one is three, so it's using UTC time. So because it's got the timestamp in there, I'll be able to get them all in the correct order. And then I'll just go in and rename these afterwards. And it's got the title that comes in. So I'm gonna actually go in and just download all of these to make sure that I have them. But as I was saying, let's see, 215,000. So it took about 110,000 characters total to, generate this book. And it was a 76,000 characters. So that's about 25 to 30% extra characters over the number that was in the book that I was getting done. And that's about what we estimated it was going to be back when, we did the training about a month ago and we were showing how 11 labs works. All right, I'm on chapter four. Oops, wrong button. Let's see. A little bit annoying with this coming up just with where everything is, but so be it. let's see. I'm gonna delete this older version 'cause I don't want that one. This is full. And then copyright intro about the author. What called it? Oh, interesting. Doesn't have the full title. So that's chapter 1, 2, 3, 4. All right. Just making sure I didn't miss one. Yep. This is the exciting part of doing the work. Bit of rote, but just wanna make sure I have everything before I turn in for the night, and then I don't have to worry about it in the morning. Now, I don't know if there's very many people watching or not. I haven't actually checked, but if anybody is interested in doing, walkthrough and showing the, nitty gritty of how the work gets done, I can certainly do this for other projects in the future. Just let me know if you found this worthwhile or. Presumably you won't tell me if you didn't find it worthwhile, because why would you be still listening to me talk two hours later if you didn't enjoy it or didn't find it useful at all? And for anyone that has been watching, I hope you enjoyed geeking out with me over talking about racing. I didn't show you that particular one, but in the, about the author section, I talk about how I put on, I organize Erin hates, when I say put on, I organize between, 14 and 20 races a year, depending on the year. usually now towards the 14 end for 15 races. But, all righty, I'm gonna have to count these afterwards to make sure I've got 'em all. 'cause with the various popups that are coming up, it is getting a little confusing, but almost to the bottom. Okay. All righty. So we've got 24 chapters plus the final note. That's 25, 26, 27, 28, and then the full download is 29 files. So if I click select all 29 items, so that should be right now, let me just make sure I didn't do two in a row. Race scoring title, copyright intro. Actually, if I just do a list, I can bring this in and I can share that window with you. So we're looking through at all of our files here. We've got race scoring title and copyright. Introduction about the Author Place, net time. This is actually gun time place if there's a slash it doesn't get anything before the slash net time, age and gender awards. Gender grading. Reverse seated start criterion distance. I'm gonna bring this up on the side. intro out the other please. Handicapped gender grading. Reverse seated start. I'm scrolling in the background, not sharing that. So you probably can't see it, but, oh, I've got the 6, 7, 8, 9, 10, 11, 12. That makes it easier. 13, 14, 15, 16. Cross country scoring would be 17. Yep. 18, 19, 20, 21, 22, 23. Looks like I did it right. So one thing to know, if you put a slash in your chapter title, it does mess with the, naming convention a little bit, but that's okay. It's got them all in the right order. I've got 'em all. And if we come in and look at our columns, then we can see that we are at 73 minutes and 41 seconds. So thank you for joining me for the night and, it looks like there were some more comments, but they were all from me, except one person did say, that's worthwhile. Please do more. So, all right, maybe we'll do that again. I do plan on, creating a. A version of this book using Affinity Publisher. So I may live stream that we'll be doing a training on it as well, but having the live stream could be useful just to see all the, minutia rather than talking about it actually being able to see that. So I'm glad, at least one person found it useful. And in the meantime, have a good night and I'll, either link to this, Facebook Live or maybe even put it on a bonus page if anybody does want to take a look at it in the future. All right, have a good night everybody.
Blaine Moore: She went through a demo for Push Button Book Formatting, Vellum, Atticus, and Reedsy.
Blaine Moore: I can probably do a more detailed walkthrough when I create the actual ebook for this one (we used a public domain book with illustrations for the demo).
Blaine Moore: I was using Vellum to create the epub for Eleven Labs, but it didn't actually create it exactly how I wanted so I might wind up using our Push Button formatter with a custom CSS file for the final version.
Choosing a voice should come before creating your project. Use the voice marketplace to find some voices that you think might work, and then use the normal speech synthesis page to test out some sample representative text from your book using the different voices. I thought I knew which voice I'd want to use based on the marketplace samples, but when actually reading my own words it didn't quite work as well I'd hoped and I found a different one that I liked better. I had to recreate the whole project from scratch to get it to default to the new voice.
Projects are a pretty effective way to create your audiobook; certainly much easier than the normal (and somewhat limited) speech synthesis interface. You can quickly generate an entire audiobook at once while still maintaining a separate between chapters, and you aren't limited by the character count.
Once you have a voice and your preferred settings ready to go, convert the whole book at once when initially creating it.
You will have to proof the entire book; you can't rely on it getting it right the first time. Things to listen for include:
Mispronunciation of words
Artifacts (often at the end of a sentence) which are random noises
The voice going completely off the rails (happened once in the entire book, but the paragraph sounded all scratchy/staticky/robotic)
Improper inflection in the reading. Often just didn't like the way it was read, and needed to regenerate the paragraph
Generate a process for proofing and fixing. Here's some of the tips I found:
Work through the book from start to finish.
Keep notes of anything you want to revisit; you won't remember otherwise.
Regenerating a paragraph will usually be enough to fix an issue.
For long paragraphs, you can split the paragraph into smaller paragraphs (such as one per sentence) so that if it will take multiple regenerations, you only have to continually recreate the one problematic paragraph once you have the others ready to go.
Sometimes you need to change a word to spell it phonetically so that the AI can pronounce it correctly.
Sometimes, no matter what you do, the AI won't be able to pronounce a word. (Even a word it pronounces just fine in other places.) You can continue to try regenerating or changing the phonetic spelling, but eventually it may be worth just rephrasing to eliminate the problematic word.
You can edit the text in your audiobook; it doesn't have to match your ebook exactly.
Keep your original manuscript open in the background; this is a great self-editing process and you can update typos, grammar issues, or places where your writing wasn't clear and you just want to rephrase something.
ElevenLabs projects include a black bar next to any sections that have been generated, so you can quickly and easily scroll through a chapter to see what has and hasn't been generated so far.
If you change voice settings in the sidebar, it isn't just for that paragraph but will continue to be the norm for all future regenerations. So if you do want it just for that paragraph, remember to change the settings back immediately after.
Expect it to cost at least 25-30% more characters than are actually in your book due to the editing process and having to regenerate passages.
Once you've proofed the entire book, use the convert all button to check if anything has been missed and needs reconverting, but don't actually run the conversion!
To find what still needs to be converted, it's faster to press the “Convert” button on each individual chapter to see if that includes unconverted characters than to scroll through the individual chapters looking for sections that need conversion.
Manually regenerate any sections that still need converting rather than using the automatic button during this phase.
Once you've completely regenerated anything that needed it, use the Convert button for the entire book. Since 0 characters still need to be converted, the process should be very fast.
It may take some time for the full audio files to be ready to download. It may be worth stepping away for some time before attempting to download the files to avoid frustration.
When downloading individual chapter's audio files:
The filename will include a timestamp for when you requested the file, so download them in order to keep the filenames in the right sequence.
The filename will also include the chapter number and name appended to the end of the file.
The chapter name will also cut off anything before a forward slash in the chapter name, including the chapter number.
You will need a minimum of the “Creator” account in order to use the projects feature.
“High” quality (192kpbs bitrate) output costs 20% more characters off of the top to produce than “standard” quality (128kbps).
ElevenLabs will forfeit any existing characters when your account rebills, even if you got bonus characters during a promotion part way through the month. Try to use as many as you can before you rebill. (If you run out, you can purchase additional characters while you wait for the rebill; it's a setting in your account that you will need to toggle on before it will allow you to.)
Discuss This Work Session
Eleven Labs
Training #512 – A.I. Narrated Audiobooks