Not all authors are against this either. Some big names are all in supporting Spotify (and other new audiobook platforms) because they are sick of Amazon/Audible monopoly.
Personally though, Spotify's audiobook user experience is paltry compared to a dedicated app.
How does Spotify affect ‘number of books sold’ statistics? I am an author and I don’t much care about statistics (the win for me is how often my readers directly engage with me) but I have a couple of author friends who are obsessed with sales numbers.
This year I have been shifting content platforms, buying audio books from Libra.fm and buying eBooks from Kobo. That said I am still a loyal Amazon, Apple Books, and Google Play Books customer.
I don't think it matters too much. I know someone who is a little obsessed with gaming Amazon book rankings - sort of the same thing as NYT bestsellers list.
How far away are we from AI reading books so well that everybody who has the book in text form automatically has it in audio form, because they can just turn it into audio via a multitude of AI text to speech networks?
By the way, I would love a service where I could listen to HN. Play the first title and then wait for me to hit the space bar to play the next. Highlight the one I just heard, so when I'm interested int a submission, I have it right in front of me.
We're already there. I use Readwise Reader and often take advantage of the text-to-speech feature for listening to articles from the Economist. The Economist already provides an audio narration of the article by a human, but some of the narrators' British accents are a bit too thick and difficult for me to understand at 1.5x. Readwise Reader's text-to-speech does a much better job at 1.5x for my American English brain. AI-narration isn't perfect—here are some issues I have encountered:
- Section headers are read just like regular body text, which is confusing
- Not great with reading numbers. Narrating a year is very different from narrating a monetary value.
- Parenthetical statements are read just like regular body text
I also use Readwise Reader and the speech update they dropped a month ago made great strides towards AI reading.
It’s just missing some UX improvements now, mainly the ability to continue playing across multiple articles. It’s also not smart enough to skip boilerplate junk that appears on top and bottom of every newsletter, which is easy to skip over visually but a pain when read out loud.
A while ago, Amazon tried adding TTS to kindle and got immediately slammed by the publishers, so they withdrew it. However, with the current state of AI, I think users will start to demand it as a standard feature of anything they’re reading. Not only can it read at human-like capability, it can do so in any voice theoretically. Soon you’ll also be able to pause it and ask questions about what it just read or to summarize the rest of the article and so on.
Neat feature. I assume is AWS Polly under the hood. This caught my eye though:
> Authors can set a list price between $3.99 and $14.99 and will receive a 40% royalty
Do they mean you get 40% of the work you created and they ran their model on? Oh lord. Let me run this through a TTS myself and publish it like that. Not sure if they even allow that though.
That's the standard payment when publishing your audiobook using ACX (how independents get their books on Audible) - and then only if you go exclusive with Audible. If you're not exclusive, you get 25%.
Audible's market share is like 60%, so you loose out on a lot if you don't publish there.
I wrote https://github.com/aedocw/epub2tts and the quality is pretty phenomenal if you use XTTS for the voice generation. It's really excellent, but requires the book not have DRM.
You can use iOS’s built-in screen reader to narrate the Kindle app. I listened to an entire book that way. Audible is much better but you can do it this way too.
Hardware Kindles (and the app) used to support TTS for accessibility, but the functionality was removed at the request of publishers (who wanted to sell audiobooks).
I listened to an entire Kindle book this way many years ago. I imagine / hope Siri's screenreading only got better in that time.
I think you need to set the Kindle app to continuous scrolling (in which case it won't work with older books), then you trigger the iOS screenreading shortcut, and off it goes.
On the one hand, the technology is already there, if you're merely looking for a "mechanical" rendering that sounds like a person with no emotion or understanding of the text -- no meaningful emphases, no meaningful change in rhythm and pitch, etc. (Well and the occasional mispronunciation of a word, especially heteronyms).
But if you're looking for tech that can read it as well as a human can in terms of turning meaning into prosody -- all of the vocal inflections that we use and hear without even realizing it, the things that make it sound human and appropriate to the text -- that's hard to say. (Especially when you consider that most people are bad at audiobook narration, as it's a real skill.)
It might be very soon, if the kind of meaning embedded in LLM's turns out to be perfect for the job. Or it might be very far away, if it turns out that it requires building a kind of accurate model of human emotion and feeling and how that maps to vocal qualities, that has nothing to do with how LLM's work.
Yes, that's exactly the kind of "mechanical" version I was talking about. The article goes on to describe the challenges I was talking about:
> Still, computer-generated voices have long struggled to hold the attention of listeners for long periods of time, and to overcome the ‘uncanny valley’ effect of synthetic human speech. Human intonation and inflection are notoriously difficult to predict and replicate.
Some samples here: https://authors.apple.com/support/4519-digital-narration-aud.... It's good, but it would definitely get very monotonous to listen to for 10+ hours, and even in these short cherry-picked examples there are a couple kind of jarring tonal decisions. You'll also notice they avoided sampling any dialogue, presumably because these voices will read all characters in the exact same manner as the narration.
I listen to a lot of audiobooks and I definitely notice and get annoyed when I don't like the narrator. I think I could maybe get used to this for one book but I would avoid it if possible, and I would really hate it if every book I listened to used this exact same style.
I would say this is probably already good enough for short non-fiction articles, but not there yet for fiction. It is genuinely confusing (for me at least) when an audiobook narrator reads a phrase with the "wrong" emotion, or doesn't vocally distinguish between characters.
Exactly -- the voices all sound like they belong to somebody emotionally "lobotomized", for lack of a better term. Every sentence has the exact same rhythm patterns, and every sentence that ends with a period ends on the exact same pitch. Listening to a single sentence is fine; listening to just ten sentences is monotonous and just feels strange.
It's perfectly fine for Siri responses or reading back a dictated text message. Not for listening to a whole book though.
I've listened to a lot of Books on Tape from the 80s and 90s where narrating in a mostly flat monotone was the norm, but even they would inject emotion and vary their delivery where appropriate.
Good modern audiobooks are a true performance even when the narrator isn't doing a bunch of highly distinct voices. You can't replace that with AI any more than you can replace a great voice actor.
Spotify's push to include more shit that will further crowd the UI and require spamming us outside the OS's native notification protocol sparks annoyance among users.
Yes, I can't wait for another "app drawer" alongside podcasts that I'm forced to ignore. This is causing me start buying music more and more from bandcamp so I can have control over all of this again.
Paying, but not paying enough. Spotify's music streaming business is unprofitable, filling the inventory with cheaper content (and raising prices each year) is the only way.
They'll lose customers if they charge a market price, it's already as expensive as most streaming video options (many of those are also not profitable).
This is the Silicon Valley B2C model writ large: charge below-market prices with VC funds, hoover up millions of users, go public, then use outside financing to cover the operating losses and invest in new business areas. Finally, when financing becomes expensive, start raising prices.
* Only being paid when a book is finished is bogus. I have a few audio books that have listened to for only a portion of the book. The biggest example being Seveneve where I don't want to listen to part 3 again.
* The 15 hour monthly limit might begin to act as pressure to keep audio books length under 15 hours. Some in sci-fi can exceed that limit now.
It would be as if I would pay my bookstore only when I have finished my book. It sometimes take me years to go through a pile of books from a shopping spree. Edit: wrote library meant bookstore
I think a lot of people browsing audible look at the Total Playtime field and think "Wow, that's a lot of playtime for one credit" and it draws them into some novels over others. This can put a cap on the authors that choose to chase that total playtime value proposition.
Of course, the trilogy or series trick happens often too. I'm not sure how new that is, or how much that is influenced by audio books.
The other impact of this strategy is very very slow narration. Stretch each word out. I feel I need to listen to most books at 1.5 speed just to get back to a normal
cadence of speech.
The multipart trend has another harmful effect: it encourages authors to "test the waters", and abandon the planned trilogy if the first one doesn't sell well.
Since authors are just trying an idea, frequently the first part is a slog where the plot barely advances, then a final action chapter that ends with a cliffhanger. No wonder many never continue.
> The 15 hour monthly limit might begin to act as pressure to keep audio books length under 15 hours. Some in sci-fi can exceed that limit now.
When I read this I instantly went to review my collection. To my surprise, approximately half, were under 15 hours. I would have thought for sure none I've listened to were under 15 hours, but like I said, half were.
In my opinion, Spotify audiobooks will fail because of limits like this. When my listening was at its peak, I would crush multiple books a month (at 1x even). If I were a customer, the first time I hit that limit would be the last day I was a customer.
First the big-budget move into podcasts and now audiobooks; Spotify appears to be admitting they'll never make any money from streaming music. Maybe in a few years they'll go Netflix and the app will just have 20,000 store-brand bands all signed to Spotify Records.
I don’t understand why Spotify keeps diluting its focus. I moved to Qobuz once a bunch of podcast started to pollute the UI and now it is only getting worse.
Music has very low margins. The record industry eats the most of the revenue pie. By extending into other audio areas they leverage their existing infrastructure around audio encoding and delivery to provide a service with higher marginal revenue.
Spotify is constantly finding new ways to irritate me when opening their app.
I just want to listen to music, but their app wants me to look at modal popups for promos I don't care about, and recommends me music genres that neither I nor anyone I know wants.
I'm worried that ZIRP and the demand to constantly grow will cause Spotify to enshittify itself into irrelevance.
The push into audiobooks feels like the start of the ehshittification of Spotify Premium. With an artificial limit of 15 hours of listening, can we expect to have to start paying for specific albums or releases?
Yep, the golden age of streaming seems to be coming to an end on all platforms. Get ready to have to endure ads and other harassment, even as a paying customer. Install the app! Got the app already? We saw you were enjoying Beethoven so on a related note have you considered buying Taylor Swift concert tickets?
What's fundamentally different between music and audiobooks?
And more generally, what's so different between literature and other media?
For movies and music, we have streaming services where we have access to everything for a monthly fee, or we can buy one off physical or digital copies. For (audio) books, there is no equivalent to the streaming model, subscriptions are just giving you access to a set number of works per month. That's especially ironic since libraries have been "all you can read" since before computers even existed.
I have seen online libraries use ridiculous DRM schemes to emulate the fact that they have a limited number of copies to lend, why?
Video games are also not sold through "all you can play" platforms, but this is more understandable since they are more difficult to standardize due to their nature, and the monetization schemes are varied, and yet we are getting there.
62 comments
[ 3.6 ms ] story [ 119 ms ] threadPersonally though, Spotify's audiobook user experience is paltry compared to a dedicated app.
This year I have been shifting content platforms, buying audio books from Libra.fm and buying eBooks from Kobo. That said I am still a loyal Amazon, Apple Books, and Google Play Books customer.
[0] https://www.esquire.com/entertainment/books/a42189320/the-ne...
By the way, I would love a service where I could listen to HN. Play the first title and then wait for me to hit the space bar to play the next. Highlight the one I just heard, so when I'm interested int a submission, I have it right in front of me.
- Section headers are read just like regular body text, which is confusing
- Not great with reading numbers. Narrating a year is very different from narrating a monetary value.
- Parenthetical statements are read just like regular body text
It’s just missing some UX improvements now, mainly the ability to continue playing across multiple articles. It’s also not smart enough to skip boilerplate junk that appears on top and bottom of every newsletter, which is easy to skip over visually but a pain when read out loud.
A while ago, Amazon tried adding TTS to kindle and got immediately slammed by the publishers, so they withdrew it. However, with the current state of AI, I think users will start to demand it as a standard feature of anything they’re reading. Not only can it read at human-like capability, it can do so in any voice theoretically. Soon you’ll also be able to pause it and ask questions about what it just read or to summarize the rest of the article and so on.
1: https://unrealspeech.com/pricing
https://www.kdpcommunity.com/s/article/Invite-Only-KDP-Beta-...
> Authors can set a list price between $3.99 and $14.99 and will receive a 40% royalty
Do they mean you get 40% of the work you created and they ran their model on? Oh lord. Let me run this through a TTS myself and publish it like that. Not sure if they even allow that though.
Audible's market share is like 60%, so you loose out on a lot if you don't publish there.
Before the acquisition, sonatics tech looked pretty cool. https://techcrunch.com/2022/06/13/spotify-is-acquiring-sonan...
that being said-- NOT A FAN. i want books read by human people who are good at reading to me. I dont want teddy ruxpin, i want my dad!
A colleague was saying to me that they would like to be able to have an ebook to read, but also to have it read out loud when driving or walking.
I think you need to set the Kindle app to continuous scrolling (in which case it won't work with older books), then you trigger the iOS screenreading shortcut, and off it goes.
But if you're looking for tech that can read it as well as a human can in terms of turning meaning into prosody -- all of the vocal inflections that we use and hear without even realizing it, the things that make it sound human and appropriate to the text -- that's hard to say. (Especially when you consider that most people are bad at audiobook narration, as it's a real skill.)
It might be very soon, if the kind of meaning embedded in LLM's turns out to be perfect for the job. Or it might be very far away, if it turns out that it requires building a kind of accurate model of human emotion and feeling and how that maps to vocal qualities, that has nothing to do with how LLM's work.
> Still, computer-generated voices have long struggled to hold the attention of listeners for long periods of time, and to overcome the ‘uncanny valley’ effect of synthetic human speech. Human intonation and inflection are notoriously difficult to predict and replicate.
I listen to a lot of audiobooks and I definitely notice and get annoyed when I don't like the narrator. I think I could maybe get used to this for one book but I would avoid it if possible, and I would really hate it if every book I listened to used this exact same style.
I would say this is probably already good enough for short non-fiction articles, but not there yet for fiction. It is genuinely confusing (for me at least) when an audiobook narrator reads a phrase with the "wrong" emotion, or doesn't vocally distinguish between characters.
It's perfectly fine for Siri responses or reading back a dictated text message. Not for listening to a whole book though.
Good modern audiobooks are a true performance even when the narrator isn't doing a bunch of highly distinct voices. You can't replace that with AI any more than you can replace a great voice actor.
This is the Silicon Valley B2C model writ large: charge below-market prices with VC funds, hoover up millions of users, go public, then use outside financing to cover the operating losses and invest in new business areas. Finally, when financing becomes expensive, start raising prices.
* Only being paid when a book is finished is bogus. I have a few audio books that have listened to for only a portion of the book. The biggest example being Seveneve where I don't want to listen to part 3 again.
* The 15 hour monthly limit might begin to act as pressure to keep audio books length under 15 hours. Some in sci-fi can exceed that limit now.
Why sell one book when you can sell a three part series.
This means the 15 hour limit is likely to only affect older novels before publishers caught onto this strategy.
Of course, the trilogy or series trick happens often too. I'm not sure how new that is, or how much that is influenced by audio books.
I think the stretched out audio hits harder on podcasts, but it's in audio books too.
Since authors are just trying an idea, frequently the first part is a slog where the plot barely advances, then a final action chapter that ends with a cliffhanger. No wonder many never continue.
When I read this I instantly went to review my collection. To my surprise, approximately half, were under 15 hours. I would have thought for sure none I've listened to were under 15 hours, but like I said, half were.
In my opinion, Spotify audiobooks will fail because of limits like this. When my listening was at its peak, I would crush multiple books a month (at 1x even). If I were a customer, the first time I hit that limit would be the last day I was a customer.
I just want to listen to music, but their app wants me to look at modal popups for promos I don't care about, and recommends me music genres that neither I nor anyone I know wants.
I'm worried that ZIRP and the demand to constantly grow will cause Spotify to enshittify itself into irrelevance.
(I work for Spotify now, but have no idea what the audiobook licensing is there, and couldn't tell you if I did know)
And more generally, what's so different between literature and other media?
For movies and music, we have streaming services where we have access to everything for a monthly fee, or we can buy one off physical or digital copies. For (audio) books, there is no equivalent to the streaming model, subscriptions are just giving you access to a set number of works per month. That's especially ironic since libraries have been "all you can read" since before computers even existed.
I have seen online libraries use ridiculous DRM schemes to emulate the fact that they have a limited number of copies to lend, why?
Video games are also not sold through "all you can play" platforms, but this is more understandable since they are more difficult to standardize due to their nature, and the monetization schemes are varied, and yet we are getting there.