Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Apple Books digital narration

authors.apple.com

Post reply on HN
Print view View original

121–130 of 244 posts

Re: Apple Books digital narration

#121

Post by dmitriid »

qwerty456127 wrote: post #68
andy_ppp wrote: post #63

"Once your request is submitted, it takes one to two months to process the book and conduct quality checks." My guess is that these generated voices are far from perfect and someone has to go in and crank the algorithm to get a fair number of passages to not sound strange. Even in the example Helena there is a word at the end of a sentence that sounds like it should be in the middle and has a bit of weirdness to it.…

Why is that we still can't have a perfect or near-perfect text-to-speech given all the astonishing advances in ML taking place? Is TTS an area nobody is really interested in or is it harder than generating beautiful pictures and sophisticated writings? This thing by Apple already sounds way better than the best I heard previously (NextUp Ivona) but it is not an instant-result offline tool yet and that's sad.

> Why is that we still can't have a perfect or near-perfect text-to-speech

Define perfect ;) Two different people will read the same text slightly (or not slightly) differently.

A great example is this brilliant and funny rendition of "To be or not to be" by Tim Minchin, Benedict Cumberbatch, Judy Dench, David Tennant and others. Sorry for the Facebook link, but it's very hard to find this video anywhere: https://www.facebook.com/watch/?v=585252039999241

Re: Apple Books digital narration

#122

Post by ivansavz »

hiidrew wrote: post #118
ivansavz wrote: post #112

Or you could use the command `say` on the command line on any current mac to get good-enough text-to-speech. See full script here: https://gist.github.com/ivanistheone/de3ccb244224d101bb93320... and this doc explains how you can setup a keyboard shortcut to turn any text selection into an audio book https://docs.google.com/document/d/1mApa60zJA8rgEm6T6GF0yIem... Here is a sample if you want to hear what it sounds lik…

TIL. This is an interesting capability of the command line. Have any more fun ones? (at least fun to a CL noob)

Here is another script `getmp3.sh` that you can use to download .mp3 file from any youtube music video:

   #!/usr/bin/env bash
   echo "Downloading mp3 from $1"
   yt-dlp -x --audio-format mp3 "$1"
You'll need to install https://github.com/yt-dlp/yt-dlp#installation before you can use that. As you can see, the "script" is just so to add a options `-x` (extract audio) and `--audio-format mp3` to convert to mp3 in the end.

Re: Apple Books digital narration

#123

Post by drexlspivey »

vortegne wrote: post #92
drexlspivey wrote: post #89

Earlier quoted context omitted.

Voices have a race?

Often there are clear voice and accent differences. Are you saying there aren't? That would be a very strange statement.

Accent depends on where you were born/raised, not on your race.

Re: Apple Books digital narration

#124

Post by macintux »

mark_l_watson wrote: post #103

While TTS has broad application, I am skeptical about Apple’s process being able to compete with the best narrators. I have my biases. My wife and I have licenses to listen to about 500 Audible audio books and in the best of them I feel like I have a human to human relationship with the narrator that is similar to a relationship with the author. I have mostly worked on deep learning projects over the last eight years…

Best narrators? Agreed. But as someone who did some recording for a local radio station years ago, it’s an incredibly time-intensive project to record a book.

Re: Apple Books digital narration

#125

Post by drexlspivey »

tchalla wrote: post #97
drexlspivey wrote: post #89

Earlier quoted context omitted.

Voices have a race?

Are you saying that voices across all races, cultures etc sound the same?

No, it depends on the culture. Both whites and blacks raised on East London will have similar accents for example.

Since we are strawmanning, are you saying that cultures and races have a 1-to-1 relationship?

Re: Apple Books digital narration

#126

Post by macintux »

hiidrew wrote: post #118
ivansavz wrote: post #112

Or you could use the command `say` on the command line on any current mac to get good-enough text-to-speech. See full script here: https://gist.github.com/ivanistheone/de3ccb244224d101bb93320... and this doc explains how you can setup a keyboard shortcut to turn any text selection into an audio book https://docs.google.com/document/d/1mApa60zJA8rgEm6T6GF0yIem... Here is a sample if you want to hear what it sounds lik…

TIL. This is an interesting capability of the command line. Have any more fun ones? (at least fun to a CL noob)

I haven't figured out how to effectively search my HN favorites, else I'd probably be able to find a few more of these, but this was discussed recently:

https://git.herrbischoff.com/awesome-macos-command-line/abou...

Re: Apple Books digital narration

#128

Post by patentatt »

rickdeckard wrote: post #11
bongobingo1 wrote: post #6

Will we see this level of voice synthesis in the public domain? Maybe I am out of touch but I found those examples very impressive - more impressive than the jobs vs rogan demo a few months back. But I am also saddened at a future where all this is locked up in corporate hands - obviously there is money needed and (licensed) data needed too which Apple can get at. Honestly I would rather eschew the ethics of it and j…

My guess is that such "digital narration" is on the brink of becoming available as a service to authors and publishers, with Amazon and Apple trying to get ahead of it by selling a product that can only be published on their platforms then. They are surely able to undercut human narration, but even on digital narration they have an unfair advantage of earning a share on every sale as well. Will be interesting to see…

I'm curious what the economics are here, because getting someone to just read a book can't possibly be that expensive, right? I'm not talking celebrity voice-over work, but just getting anybody with a passable voice to sit down a spend a day or so reading a book into a microphone? Does that really cost that much more than having someone sit down a listen to the whole AI-generated audio book to do QA? And then there's all of the engineers who have to work on the project, the hardware to run all of it, etc. Seems like if it takes months to do and has to be QA'd anyways, it can't be that much more cost effective. Now, if it's completely computer-generated and can be a push-button feature? Sure, that makes sense. I wonder how close they are to that.

Re: Apple Books digital narration

#129

Post by RegnisGnaw »

mark_l_watson wrote: post #103

While TTS has broad application, I am skeptical about Apple’s process being able to compete with the best narrators. I have my biases. My wife and I have licenses to listen to about 500 Audible audio books and in the best of them I feel like I have a human to human relationship with the narrator that is similar to a relationship with the author. I have mostly worked on deep learning projects over the last eight years…

Not every book can afford the best narrators. A good one charges somewhere in the ballpark of $300-500 a finished hour. So for an average novel that's like $3000-5000. Not all writers can afford that, so this is an cheaper alternative.

Its like an Lexus vs Hyundai.

Re: Apple Books digital narration

#130

Post by falcolas »

ghaff wrote: post #105

I'm skeptical given the state of the art. There is way more good audio content out there than I have the time/interest to listen too and I can't believe I'm that atypical. And a book is a relatively big listening time commitment. I'll happily pay a few dollars more for a good human narrator.

A couple of comments from a narrator whose worked through ACX before.

First, the last few years have seen a race to the bottom for narrator rates, since during the pandemic it was recognized that it's a job that can be easily done from home, literally from anywhere in the world.

Accordingly, the up-front cost for an average quality 10 hour book is only about $1,500, and can be turned around in under two weeks from a human. If you get a really good and well known narrator, it's still only about $4,000 (and you'll probably get it quicker).

Also, they're going to be competing against revenue share models from Amazon/Audible, which basically means it costs the author nothing up front. Amazon's bite out of audiobook sales is absurdly high (60%), so other companies could (and are) definitely improve on that. It's mostly a fight against Audible's brand at this point.

But back to AI: AI narration is going to have to compete against humans willing to do a lot of work for very little pay. I'm honestly not sure the compute and QA costs will be competitive. And frankly, even if it is cheaper, it's not as if those savings will be passed back to the customer.

If you'd like to look at how little it can cost to get a human to do voiceover work, check out fiverr.com and look for voice actors and narrators.

Post reply on HN