Spoken Press for Audiobooks ~ My Final Analysis

Since late April, I've played around with Spoken Press's audiobook creation. The company is still in its beta phase, but according to updates, it's set to go live soon. Thus, while it's still free to use, I wanted to fully test its capabilities.

I documented my initial impression here. This experiment involved uploading my entire manuscript for Running From Herself, then setting out to assign voices to each character for a multi-character audio rendition. The process was very tedious. I was vaguely cognizant of including several characters in my novel, but many only had a line or two. However, I was tasked by Spoken with assigning voices to each and every one of them. At some point, I ran out of distinct voices to choose from, so no doubt some of the characters sounded exactly the same. (I never listened to the entire playback.) In my case, this process took a couple of days of inputting character descriptions into the text box, then auditioning voices from Spoken's recommendations. 

To be clear, Spoken provided summary descriptions for the characters that featured prominently in the story, and I only needed to tweak a few of its assumptions to match a character's dialect and spoken attitude. For the incidental characters, however, I needed to input descriptions, which was silly, since they barely spoke at all. And there is a minimum word requirement, so I couldn't, for example, simply input, "This is a male." Thus, I had to detail his age, whether he spoke with a western or Midwestern twang, etc., then still being short of the minimum description, I found myself adding, "I have nothing more to say about this character." 

There were other frustrations that I detailed in my first analysis, so I won't reiterate them here. Suffice it to say that for a 115,000-word manuscript, the entire process was unwieldy. 

In the end, I only listened to a snippet before deciding that I hated the multi-voice format and switched it to single voice.

I chalked the whole thing up as time (not) well wasted, as Brad Paisley would say. I stuck the link to the "book" in the news section of my website and forgot about it.

Subsequently, I would receive email updates from Spoken from time to time, detailing its latest enhancements, and finally after getting a notice that the site was about to go live, I decided to give it one more chance while it was still free to use. This time, I determined to only convert the first four chapters of my story to audio, with the intent of sticking the link on my author site as a preview. Here is how this process began. 

Assuming that a multi-voice project would now be easier to work with, since the amount of text I was using was much smaller, I gave multi-voice another try. I assigned my voices, then took a preliminary listen. Immediately there were issues with sentences being broken in two and words with unnecessary spaces added between letters, which resulted in incoherent narration. I went back in and corrected Spoken's every weird transcription, but upon playback I found that I'd missed a few, so I needed to start the entire process again. 

And again and again and...

Every time I downloaded what I thought was the final version, upon playback I found new issues. Either a sentence's emphasis was wrong (I did use Spoken's editing tips to configure "speech" as I wanted it to be voiced) or one character shouted his line angrily instead of playfully, or the primary narrator stumbled over pronouncing basic words. I must have fixed this project a half dozen times or more, downloaded each iteration, began listening back and found another mistake. At one point I think Spoken became exasperated with my repeated fixing and stopped working all together. By the time I went to bed last night, it was still publishing my last attempt.

Well, I've now had it. 

My latest playback lasted maybe three minutes before I found a new problem. Mind you, the problems are not on my end. The latest was the Burt character saying, "Did you know you're our only guest?", then the narrator's passage would kick in with "he said". Except, no. Burt voiced, "Did you know you're our only guest, he said".

I can't do this anymore. The problems will never end. 

Finally (and truly finally), I went back to Spoken, switched everything to a single narrator, and washed my hands of it. I don't know if I even want to listen to this playback, because something will invariably be wrong (again). I've done FAR too much work for a project that isn't even important to me. Yes, it would be a nice bonus to include an audio snippet on my site, but not at this much of a psychic cost to me.

If Spoken was to ask for my feedback, I wouldn't even know how to advise them on fixing all their glitches. I'm hardly a programmer, but I assume they employ one or two. They really do need to ask for feedback, though, if they hope to ultimately compete with other audiobook services. 

As it stands now, an author would have to be a masochist to utilize Spoken's process to create an audiobook. 

I can certainly empathize with an entrepreneur trying to make a go of creating something brand new. I've created my share of "new" things (stories) and they've mostly been a bust (sales-wise, not quality-wise). Maybe I should have asked for feedback and things might have gone better for me. But my dollar outlay was, well, zero. I assume Spoken's owner(s) invested a good amount of money into their venture and want to see a return on that investment. Thus, they need to ask for beta user feedback, yet I've never once been asked.

There is definitely a market for a service like this, in which impoverished authors could produce their own audiobooks. Once Spoken's model goes live, it claims that creation costs will be between $40-$100. Even someone like me could afford that.

I've given Spoken more than a fair shot. It doesn't work for me; others may feel differently ~ I don't know how, but they might. While I haven't listened to any other authors' completed projects, perhaps they figured out how to tweak the system to produce an acceptable result. But it shouldn't have to involve guessing and constant retrying. And hour upon hour of manipulation.

If Spoken ever incorporates the necessary improvements, I'd be willing to give it another go, but I wouldn't lay out any money until I was sure I would get the result I wanted, and I'm pretty sure no company lets people try before they buy.

Rather than giving one overall letter grade, I'll rate Spoken's components individually and leave it up to the author to decide.

Ease of text upload: A

Voice options: B-

Absence of glitches: F

Publishing speed: A-

While uploading is easy and the publishing process is fast, the same can be said about virtually any service. However, the frustration factor (i.e., "absence of glitches") makes Spoken a no-go for me.