In this podcast episode, we invited Cassie Smyth, the Audiobooks Manager at ECW Press to talk about AI integration into audiobooks, accessible audiobooks, supporting the Canadian audiobook ecosystem, and more.
(Scroll down for a transcript of the conversation.)
Want to make sure you never miss an episode of the podcast? You can subscribe for free on Spotify, iTunes, Pocket Casts, TuneIn, or SoundCloud.
Further reading/listening
Spotify is making it easier to release audiobooks narrated by AI
Upcoming Webinar: Making Audiobooks Accessible: Challenges and Opportunities
Transcript
Adaobi Nnaobi: Welcome to the BookNet Canada podcast. I’m Adaobi Nnaobi, the Marketing & Research Associate and the host of this month’s episode. Today, I have Cassie Smyth with me. Cassie is the Audiobooks Manager at ECW Press. In her role, she has managed the production of more than three hundred audiobooks, all by Canadian publishers and authors, and narrated by Canadian voice talent. She holds a BA in English, minor in Business from Carleton University, and a certificate in Creative Book Publishing from Humber College.
Hello, Cassie. Thanks for joining us.
Cassie Smyth: Hi, Adaobi. Thank you so much for having me.
Adaobi: Let's start with an overview of the audiobook industry in Canada. Tell us about what's currently happening with audiobooks in the Canadian book market.
Cassie: It's a really interesting time in the Canadian audiobook market. We're seeing a period of change. I'm not sure how many folks know, but from 2019 to 2024, the Canada Book Fund ran an initiative for accessible digital books. Now, this was both audio and ebook, and so there was this period of funding for Canadian publishers to apply to. And in that period, there was just this boom of audiobook production in Canada, and a lot of Canadian publishers came to ECW for their audiobook production work.
So, during this time, ECW was producing north of 120 audiobooks per year. We were very busy. It was very chaotic but very fun. And during this time, we created this ecosystem of these Canadian publishers that were doing audiobooks authored by Canadian authors, narrated by Canadian talent, recorded in Canadian studios, edited by Canadian editors, proofed by Canadian proofers. We had this huge ecosystem that we built up during these five years. And now the funding has gone away, unfortunately.
And so there's been this little lull of how do we react to the lack of that funding. ECW has continued with our audiobook production work for our own in-house titles, but we've noticed that other Canadian publishers have, kind of, backed off from the audiobook front or have found other methods to go about it, whether that's finding another producer or licencing to audiobook publishers down in the States, which we can talk about more later. But really we were doing 120 audiobooks per year like I mentioned and now we're down to forty or fifty. So, it's a big decrease in production, but we still think it's really important to get this content out for those who maybe have a print disability where they require the audio to enjoy books but also people who just love to listen to them for the thrill of a good story being told to you. And I think there's a lot of merit in that and a lot of merit in this form of books.
Adaobi: Yeah, and that leads us straight to our next question. According to our State of Publishing in Canada study, most publishers, 89%, found there were barriers to audiobook production. The biggest barrier by far was financial constraints at 73% agreement. How can publishers, especially smaller publishers, work through this hurdle?
Cassie: It's a really tough one. A general rule of thumb that we apply at ECW is that maybe one out of every 10 audiobooks you produce will be a success that will make all the expenses for the other nine books worthwhile. And that's a tough pill to swallow because you have to take a chance on 10 audiobooks, and that's a lot of money upfront. In audio, though, we do find that backlist is far more evergreen than in other mediums. So, if you do have a strong backlist that helps finance further audiobook production. For ECW, for example, Moon of the Crusted Snow is huge for us in all formats, but especially in audio, that's quite a boon. And so that really helps us create further front-list audiobook titles. But that is a big investment for publishers to swallow to get that initial backlist, kind of, under your belt.
Besides that, I would just recommend we continue lobbying funding bodies for financial support. I do think that there was just such a robust ecosystem created in the audiobook sphere in Canada during those five years. And it's such a shame that that has kind of ... We've taken our foot off the pedal, let's say, and we're, kind of, coasting right now and, kind of, seeing what happens. But it'd be wonderful to continue building that out instead of licencing content down to the States. We do it too, we know that, but licencing down to American publishers to create audiobooks and at least get them out in some way, shape, or form. But it's wonderful that we have this ecosystem right here in Canada. It would be wonderful if we were able to use that to our advantage and continue to support local.
Adaobi: I, of course, agree with that. From our Canadian Leisure and Reading Study, we learned that 82% of audiobook listeners prefer listening to a human narrator over AI-narration, 65% have stopped listening to an audiobook because of the sound of the narrator's voice, and 68% search for specific narrators when looking for audiobooks to listen to.
Narration is very important to audiobook listeners. With AI-narrated audiobooks, where AI-generated voices read books, and authorized voice replicas, which use licensed samples of a specific human voice to create an AI-generated narration, how do you feel about the use of AI-narration for audiobooks in Canada?
Cassie: Narration quality is something that's very important to listeners. And I urge everyone to think about how important it is to casting directors like myself. Just a little bit of background, when we put out open calls for auditions, if we're not specific enough, we might sometimes receive more than 150 auditions. But it's part of my job to sort through these and determine the best of the best. And so when we finally select a narrator and put that forward to the public, of course, we hope that it is the best of the best and that they agree with the decision that we've made.
Now, AI. It's a tough one because at ECW, we believe in working with human narrators and supporting Canadian talent and all that. But I do see the argument being made on the other side that if the audiobook is not going to be made, period. If it's, let's say, a very backlist title, that it does not make any sense financially and that we want to put it out there for accessibility purposes so that it's just out there, period. I do think that that is a valid argument so that people who rely on the audiobook to read the book, have that available to them. Of course, I love to see the story be brought to life in a human narrator. We're just never going to have the nuance and the emphasis and the storytelling capability that a human narrator has in AI. At least I hope not because I would love to see humans, you know, rise up above the robots. But I understand the financial argument, and I understand the accessibility argument, and I appreciate those. But I will always be in the camp that I would love for a human being to be telling me a human story.
Adaobi: With AI narration seeming like the cheaper and more efficient option in some ways, what’s the case for human narration today?
Cassie: Like I said, for accessibility purposes, it's better than nothing, but it really can't convey the emotion or story like a human voice can. If you're coming up to a really emotional scene, the narrator knows to, kind of, modulate their tone and their voice to convey the gravitas of that to the listener. Or if there's a joke being told, let's say, in dialogue by one of the characters, AI, at least in its current form, doesn't know how to deliver a punchline. So, we really just need that human interpretation of the text and delivery in a way that AI cannot do at present time.
Adaobi: Okay, so just to be clear on what you're saying, while we all, in quotes, prefer human narration, if there's a case where we don't have the financial backing for that in a certain case, or an audiobook would not be made because of certain financial constraints, go ahead with AI?
Cassie: I think it comes down to the individual person. I, personally, I don't have to make those decisions. And so in my current state where I don't have to make those decisions, I would not personally opt for AI. And the belief of ECW is not for AI. I will mention totally upfront and transparent that there have been maybe one or two instances where we have used AI to record a bibliography, which is like “http://www dot” blah, blah, blah, blah, blah…. And that is very arduous work for a narrator, and there is not that same acting and emotional delivery and nuance and performance to it, as let's say a fiction or non-fiction or memoir or something, that the main body of the text requires. When it's, kind of, straightforward accessibility material like that, I feel like there is more of an argument for that and we have done that a couple times in the past. But when it comes to the main body of a text, I would not personally choose it, but I do understand the argument that it's important for books to be made accessible to anyone who wants to read them. And so if AI is the only option to make that happen, I think that there's a case for it.
Adaobi: Understandable. Now, there are different levels of production and narration for audiobooks like solo narrated audiobooks, author narrated audiobooks, audiobooks with a full cast accompanied by sound effects and music, and more. How do you choose the type of audiobook you produce with listener preferences and cost in mind?
Cassie: Generally, when we're starting out a production, our practice at ECW is to connect with the author and get their feedback on what kind of voice we're looking for or voices. So, if this is, let's say, a fiction book, maybe a mystery where we're dealing with two perspectives in alternating chapters, the author will bring that to our attention, if we don't already know it already, which if it's an in-house book, we do know this already. But if we're working with an external publisher's book, we reach out to the author. We get their feedback. And in this, we also ask, what are we looking for? Gender, age, language knowledge. Sometimes we have a book with Italian or German or something like that. And it comes much easier to have a narrator that already has these language skills rather than sourcing pronunciation guidance for them. So, we proceed in that way, according to that.
If it's a non-fiction book and it's very personal to the author and the author feels the desire to narrate, then we put forth an opportunity for them to audition for the project. And we do this for two reasons. We want them to have the opportunity to experience narration. And granted, this is just a two to three page sample. And so it ends up being about five minutes maybe of narration. But we want them to experience what this is going to be like, because recording an audiobook is not easy. You're in the studio for a four-hour session reading out loud straight for many days. It could be two days. It could be six days, depends on how long the book is. And that's a long, long time, and it's a lot of work and a lot of effort, especially to keep, kind of, a throughline and keep that endurance through the entire recording so that it all maintains the same quality. So, a lot of authors will try out this audition, and they'll come back to me and they'll say, "Oh, my goodness, that was so much work. And I don't think I actually want to do this after all. It's fine." Which is totally fine. We're totally cool with that. We're great to look for voice talent elsewhere.
But then the other purpose of this is for us to see the author's tone and quality. And so if an author has experience in ... Maybe they have a podcast or they've done a lot of radio or something like that or sometimes it just comes down to them having a really good tone of voice and a really solid delivery. We're happy to work with the author and choose them to narrate the audiobook. Each production is different and we treat it like its own entity. We're always dealing with a different book, right? Straight off the hop, a different author, a different narrator, different studio, different production team. It's always a case by case basis. And that's one of the things that I love about my job is that I get to work with so many different people and so many different walks of life and work on so many books of different topics. But it also keeps us on our toes.
Adaobi: I'm, kind of, curious about something. So, I was wondering, in the case of an author that would like to narrate their book and has gone through the audition and has decided that they're fine with the hours and the work that's needed, and that you also find has a good voice but doesn't have much experience, is there a training involved or ... ?
Cassie: So, often if we'd like to proceed with the author to narrate, we often engage a director. And so this director … We have a robust roster of directors that we do work with that work on our audiobooks. But a lot of them come to this job, for lack of a better word, because they have experience maybe in stage directing, like in theatre, or they've narrated a number of audiobooks themselves, or they find us through all these different mediums, or maybe radio dramas, or something like that, or CBC documentaries. It's so interesting the way that our roster of audiobook directors has found its way to us. And each one of them is so talented in their own individual way. And we try to pair them up with a given book based on their interests and their background and stuff like that so they have a real wealth of knowledge to bring to a given project.
And so, in the case if an author wants to narrate and we're happy to have the author narrate, we do pair them up with an audiobook director and they're able to provide tips and tricks for the studio, help them practise their delivery, answer any questions they might have, like even ... Oh my, this is a little bit gross, but it's audiobook narration. So, if somebody's mouth waters as they speak, and they have a lot of mouth noise, right, and they're recording in the studio, one of the tricks that a director has told me before is that you take a bite of a green apple, and instantly your mouth clears up and you're good to go. But it's such a weird trick that you wouldn't know otherwise, and no author would — well, I assume no author would know that. I didn't know that. But it's just these little tips and tricks, and they're able to, kind of, coach the narrator through, or the author narrator, I should say, through the recording process.
We also find sometimes that our authors have written books that are very close to their heart and very emotional. And so being in the studio, our engineers that we work with on the recording end are fantastic. But having a director there to, kind of, walk them through the emotional side of it, of having to recount this story that was so moving for them to put to paper and now to vocalize, it's a whole different aspect of reliving that life or that memory. And so having directors there, it's really good on the technical side, but it's also really good on the emotional end of it.
Adaobi: That's very interesting and nice to know some of those tips. Readers get their audiobooks from a number of different sources today, with popular subscription services like Audible and Spotify recently getting into the audiobook game, is there something in particular that has worked well for ECW to stay competitive in the market?
Cassie: As I mentioned, we continue producing audiobooks in spite of the lack of funding or the financial challenges. And by creating so many audiobooks, we have a great backlist. And so part of the things that we do with that great backlist is we regularly submit to promotions that are run by these audiobook vendors. And so we put forward our backlist titles, like say there's a mystery and thriller sale going on or true crime or any kind of holiday that's going on. We like to bring back our backlist and put it forward for these promotions to keep them in the public's eye.
We also maintain strong relationships with a number of audiobook vendors like Hoopla, Overdrive, Audible, Spotify. And we meet with them every season before the season starts to first of all see how we might further our collaborations and what opportunities there are for us to grow and continue to work together in different ways. But we're also letting them know what we have coming so they know what might pique their interest and what they can merchandise, however they can do things on their end. We really just like to keep open lines of communication with the audiobook vendors so that any time anything comes up, we can reach out to them or they can let us know if they have a new promotion or a new idea coming up the pipeline just so we can continue to get word about our audiobooks out in any way, shape, and form.
Adaobi: Let's talk a little bit about accessible audiobooks. What are the standards or best practices that ECW implements to ensure your audiobooks are accessible for all readers?
Cassie: So, this is something that I feel like each production house has its own set of best practices. And over the last year, my coworker Jessica Albert and I and a number of other audiobook professionals in Canada have been lucky to work alongside CELA on an Accessible Commercial Audiobooks research project. We were the advisory committee overseeing the project. And through this research, they did a lot of outreach to audiobook listeners in Canada, trying to discern how audiobook listeners themselves felt about accessible content in audiobooks and if there's a way for us to create a standard for accessible content going forward so that everybody understands what they should do with alt text, what they should do with endnotes, footnotes, bibliography, etcetera. Because right now, we're all, kind of, handling it on a case by case basis.
ECW's practice is to include the accessible material and so this does include the types of material that I just mentioned, but it's, kind of, a case by case decision of whether that makes it into the final commercial audiobook that you would find on Audible or something like that. If we have a book that has a lot of images and complicated figures and stuff like that, we, kind of, have to make a judgement call of whether we think that an Audible or a Spotify or etcetera like just a regular average Joe, commercial listener would appreciate having that alt text in there. And sometimes the answer is a difficult no. And so we create two versions of the audiobook. We have a commercial audiobook and an accessible audiobook. The accessible audiobook is made available to CELA and NNELS and any library that requests it or any audiobook company or distributor body that requests it, but the commercial audiobook is the one that's sent out broadly.
But regardless, we do record all of the accessible material for the audiobook. So, we do that image alt text, and we make sure that that's up to standards. We record endnotes and footnotes, so that would be recording footnote or endnote number markers in place in the text. And then all of the endnotes, we make a decision whether those go at the end of a chapter or if we put them at the end of the audiobook. Footnotes, there's sometimes a case if they're more conversational that they're just integrated straight into the text. And so they're just not interrupting the flow of anything by having a footnote marker. Also, bibliography and works cited, like I said, that's tough because some bibliographies and works cited are very robust and several pages north of twenty, thirty, forty pages. And that is very challenging for a narrator to read, and it's very time consuming in the studio. And so I think when we talk about accessible audiobooks, and them being a difficult financial decision for some publishers to swallow, if we're talking about non-fiction, for example, and these big beefy bibliographies and works cited, and we want to make them accessible because we want all of our products to be accessible, there are some difficult conversations that need to be made around the bibliography and works cited. And so these are the instances that I mentioned earlier, where we have opted for an AI narrated bibliography when it's forty, fifty pages, and this is going to be days, not even hours but days in the studio, maybe north of 12 hours in the studio. And the cost of that is for the studio time, for the editing time, for the narrator is just very, very large, we'll say. And so that I find is a case where we can think about AI integration into audiobooks, where there's no huge benefit to it being narrated by a human narrator, if there's something cut and dry like that.
We also have in-text citations that we integrate into our audiobooks to make them accessible and make sure those are all there. Really we want everything that is in the print book to appear in the audiobook so that anyone who is listening can have a similar experience too if they were to pick up the book from the shelf. I think one of our best examples of this was our audiobook for Denison Avenue that was on Canada Reads last year by Christina Wong and Daniel Innes, where the first — well, I shouldn't say the first half because you, kind of, flip it and you can start it all over again. It's really interesting. I'm really not describing it well, but it's such an interesting production of that physical book. But half of it is the written story, and then you can flip it, I guess, horizontally, and you can open it again. And starting from the back of the book, you have all of Daniel Innes' illustrations. And so to put that in an audio format when basically half of the book was a graphic novel was really interesting and really great to work with Christina and Daniel personally on that and figure out, "Okay, how are we going to describe these scenes that are in the graphic novel in a way that will appeal to an audience that they will actually glean interesting information and be enthralled by that as much as they are enthralled by the first half by Christina Wong of the novel. And so we really worked with that and integrated that. And I think we have a beautiful audiobook that really is a completely separate experience from reading the physical book, because we also have Christina and Daniel's production notes of why they chose this scene, why did they draw inspiration from this. It's really, kind of a behind the scenes look, and I think there's maybe something there to explore with accessibility that why are certain images included, why is certain information there, to really draw the average commercial listener into the fold and create, kind of, a unique behind the scenes experience that they wouldn't necessarily get from just reading the book off the shelf.
Adaobi: Thank you for telling us more about that book and the process to making it. Now, what is your vision for the future of audiobooks in Canada?
Cassie: I would love to see a strengthening of Canadian audiobook love pouring out. I think there's so much happening right now, given the political climate, of course, of love for Canadian products that you can get at the grocery store or at the drugstore or something like that. And, of course, we see it pouring out from Indigo and stuff like that for print books, but I'd love to see more focus on the audiobook side of things pour out in the next few months or however long we're in this state of time, we'll say. As I mentioned earlier, our audiobooks are narrated by Canadian talent, directed by Canadian directors, recording Canadian studios, Canadian editors, and so on. We have this wonderful ecosystem of Canadian talent in audiobooks that we've fostered over the last decade or so. And I’d love consumers to know that, and for there to be some way to demarcate to them that the audiobook that they're listening to is Canadian, born and raised, basically. I think it's just such an interesting thing, because of course, in Canada, we get so much American content, whether that's in our movies and TV shows and so on. And of course, in our books as well, but it's just wonderful I think for us to have this ecosystem in Canada of audiobooks. And I'd love for people to know that and support that in the same way that Canadians have been coming together as of late to support all Canadian products.
Adaobi: Thank you for that, Cassie, and thank you for joining us.
Cassie: Thank you so much for having me.
Adaobi: Before I go, I’d like to acknowledge that BookNet Canada's operations are remote and our colleagues contribute their work from the traditional territories of the Mississaugas of the Credit, the Anishinaabe, the Haudenosaunee, the Wyandot, the Mi’kmaq, the Ojibwa of Fort William First Nation, the Three Fires Confederacy of First Nations (which includes the Ojibwa, the Odawa, and the Potawatomie), and the Métis, the original nations and peoples of the lands we now call Beeton, Brampton, Guelph, Halifax, Thunder Bay, Toronto, Vaughan, and Windsor. We encourage you to visit the native-land.ca website to learn more about the peoples whose land you are listening from today. Moreover, BookNet Canada endorses the Calls to Action from the Truth and Reconciliation Commission of Canada and supports an ongoing shift from gatekeeping to spacemaking in the book industry.
We'd also like to acknowledge the Government of Canada for their financial support through the Canada Book Fund. And thanks to you for listening.
In this podcast episode, we talk to Cassie Smyth about AI integration into audiobooks, accessible audiobooks, supporting the Canadian audiobook ecosystem, and more.