So. I submitted my first audiobook to ACX, aka Audible Creation Exchange (an Amazon company – basically KDP for audiobooks) last night. It took, like everything else in this fallen world, a lot more time and effort than it seemed like it should. But I love you (No, really, I do.) and I’m here to help you avoid some of the pitfalls, pratfalls and wrong turns I encountered. Herein, Unka Marc’s Guide To Audiobooketizing Your Prosebook.
Where to begin?
Well, there are neither beginnings nor endings to the turning of the Wheel of Indiepub. But this is a beginning.
I signed up with ACX (http://www.acx.com if you hadn’t guessed) which is free, and will link you directly into your KDP account if you sign up with the same email. This also automatically makes all your Author Page books available for instant integration as rights-available or DIY projects, which is VERY cool. If you do things in the right order, your audiobook should appear as a format choice on the book’s page when it goes live, which is SUPER cool. If you want to do it some other way, knock yourself out. You might still find some of the technical stuff below helpful, but obviously the ACX specifics, not so much.
So you sign up and you put in your information (bank stuff to get paid, tax information, legal status, stuff like that.) It’s very similar to signing up for KDP or any other self-publishing platform. Then you go through all the help pages and try to figure out what the Hell to do.
Note: ACX’s “How To” pages are monumentally badly organized. It’s not that hard to find any given thing you might want to know about, especially since Google has them thoroughly mapped. But trying to develop a workflow just from reading them is nigh-impossible. Even watching all the tutorial videos all the way through, in order, which you should do, will not help much with the nuts-and-bolts of the process. Read through this ALL THE WAY TO THE END and you should have a grip on the thing. A tenuous, three-finger grip on the side of some Godforsaken alp, but a grip never-the-less. Also here are some useful web pages:
The ACX tutorial videos: https://www.acx.com/help/video-lessons-resources/200672590 (You have to watch each one and then the next will appear for viewing.)
The ACX “Do It Yourself Author/Narrator” page: https://www.acx.com/help/authors-as-narrators/200626860
The ACX page that actually tells you what your submitted file has to look like: https://www.acx.com/help/acx-audio-submission-requirements/201456300 (Spoiler: it should be a 44Khz 192Kb constant sample rate mono/NOT joint stereo MP3. The specs are near the bottom.)
So once I was sure I was all set up, I had to decide which of my books to audiobooketize, and how I was going to do it. I started with one of my short stories, Dream a Little Dream, which is about 6700 words long, and which the rough math said should be about 45 minutes of audio. (It came in at 44 minutes. Damn, I’m good at rough math.) Please note that “being too short” is grounds for rejection, but “too short” is not defined. There are approved books online which are less than half an hour long: I don’t know if there’s a bright-line rule or not. I picked this one because it a) is reasonably short so I wouldn’t invest too much time if there was some problem with my books, B) is one of my tamer books (the protagonist is married to the only person he has sex with, everything that happens is arguably consensual,) and C) contains a realistic hypnotic induction scene. This last is a bit of a gamble: that might turn some people off, but OTOH people who are into hypno-erotica might be really into this. Time will tell.
Okay, I was all set to sell audiobooks. I just needed a, you know, audiobook. To sell.
I decided to record my book myself. I’m a trained hypnotherapist and I REALLY like to talk so I was confident that I could manage the recording. I have a reasonably acceptable voice, very clear diction and boundless ego, so I wasn’t worried it wouldn’t sound good enough. If you don’t have any of these properties, or just want a “real” narrator, you have a couple of options.
1) Put your book in as “OPEN FOR AUDITION.” This will let ACX show your proposed book to the narrators who are signed up to receive such offers, and let them make offers to record it. You can either pay a flat fee, or split the royalties with the narrator. The latter requires less money up front but of course if your book sells well means less money in your pocket. Which to choose is an individual business decision. I didn’t do this, so I can’t comment on it, but it seems simple enough.
2) Hire a narrator to record your book for you outside ACX. You could do this any way you’d find any other kind of talent, including Craigslist. If you do this, be sure to have a good solid work-for-hire agreement, vetted by a lawyer licensed in your jurisdiction, so that the rights you have and the obligations you have to the narrator are clear to all parties up front. If you do this, you’ll probably get a “finished audio” file from the narrator, and then you can proceed like I did with it as if you had recorded it yourself. GET A WAV FILE OR OTHER UNCOMPRESSED MASTER FILE. Do not just take an MP3. If you ever need to edit it you’ll be a lot better off with a real audio master file format. If you don’t want to do any audio editing, be sure they put the opening credits, closing credits, content file(s) and sample file in separate MP3 files (see the “Upload” section below.) But get BOTH the uncompressed masters and the MP3 conversion, or you’ll have to do it yourself or find somebody else to do it.
3) Have your dulcet-toned significant other, Cousin Pat, or bored friend record it for you. I’m not saying this is a bad idea per se, but see all the stuff in #2.
I already had a nice computer and was familiar with the basics of digital recording from experience as a systems administrator at a design studio, but it isn’t hard. I’ll hit the highlights of my preparation and setup.
1) I have a basement office where I do my writing and work at home for my day job. It already had a nice curtain backdrop set up for my video conferences. I just extended it a bit so that I was surrounded by free-hanging fabric on three sides, which dampened ambient echo very well. I put up a fuzzy blanket on the one large flat surface still present (the door) and I was all set.
2) I only recorded late at night or when nobody was home. A microphone good enough to get good audio will pick up a lot of background noise, even a cardioid or supercardioid. (Ask me about how the dog laid down on the floor above my head and chewed on her bone for forty-five minutes straight when I wanted to record that one time. I DARE YOU.) Also, I turned off all my external hard drives and made sure I had no fan noise from my computer picked up by the mike. I bought a small solid state external drive to record onto, they’re very fast and make no sound at all. That minimized the chance that my internal HD would spin up. It’s not very loud, but every little bit helps.
3) I got a nice USB microphone (a Blue Snowball iCE) from TigerDirect and an inexpensive pop filter (those embroidery-hoop-with-pantyhose looking things) from Guitar Center. The pop filter not only lowers the chances of plosives making your recording, well, pop, but it gives you a fixed reference point so when you do more than one session, you know how far to put your face from the mike.
4) I already had some nice headphones. You don’t need umpty-hundred dollar ones, but do get ones which are “closed back,” i.e. they completely seal off your ears. It makes reviewing a lot easier.
5) I downloaded the free Open Source program “Audacity” from the internet. I’d used it before, many years ago, and it’s gotten even better since. And you can’t beat the price or the availability of plugins for it, including a free MP3 converter. If you have a Mac, you could use GarageBand, which has a voice-recording channel, and there are lots of other audio programs out there. The fanciest consumer-grade one is probably Adobe Audition, and then you go up into the pro-level stuff. I know nothing about any of them. J
Here’s a link to download Audacity: http://audacity.sourceforge.net/
Here’s a link to the MP3 converter, which you have to install separately (it’s easy:) http://manual.audacityteam.org/o/man/faq_installation_and_plug_ins.html#lame
Here’s a link to a page for a compressor plugin that a lot of people find very useful for making audiobook MP3’s: http://theaudacitytopodcast.com/chriss-dynamic-compressor-plugin-for-audacity/
6) I got a nicer microphone stand than the one the mike came with, too, and a flexible gooseneck adapter for it, at Sam Ash. I also bought a few sheets of soft closed-cell craft foam at Joann’s, and set that on my desk and then the microphone stand on that. That was a lot cheaper than a shock mount, and it provides quite a bit of acoustical decoupling of the microphone from the desk I sat it on. I kinda want a shock mount (which is basically a web of bungee cords you suspend the microphone from to dampen external vibration) but a) my office isn’t that noisy and my equipment doesn’t vibrate much and b) I am a notorious gearhead and I have to draw lines somewhere. If you want one, especially if you have a noisier/vibrate-ier environment, by all means get one.
Once I had all my stuff, I set up my chair and the stand so I was on-axis (my mouth was lined up with the mike) and I could stop and start recording with a minimum of noise. One nice feature of Audacity is that Shift-R will append to your existing track, so if you get into the whole stop-and-start thing, it’s really simple to do. I did some sampling of “room tone,” which is just the noise when nothing is apparently making noise, and found it was pretty quiet, so I was all set. Protip: if you want to insert pauses in your recording, copy/paste room tone. DON’T insert silence with the program’s Generate Silence feature. Dead silence sounds different from room tone and is disconcerting for listeners.
Then, I actually prepared the prose book for recording. I did this by turning it into a script I could scroll through. Lots of people like to read from paper but I live in the 21st Century and besides, my desk is cluttered enough, so I scrolled through it on my second monitor while watching the recording on the first monitor. (Oh, did I mention I have two monitors? This probably wouldn’t be so handy if you only had one and the paper thing would probably be much more efficient at that point.) I put in some stage direction (e.g., when the protagonist snaps his fingers, I actually snapped mine, and I put a big SNAP in the script where the snap was to happen) and I also added a few “he thought,” “a little voice in his head whispered,” etc. I write in third person limited omniscient and I often reveal a character’s thoughts. These are always in italics in my books, so it’s easy to see when it’s happening, but in an audiobook you have to make sure the context is clear. You don’t have to overdo it, but if it’s not instantly apparent who’s talking, identify them. I also added the required opening and closing credits (they’re listed on the submission requirements page I linked above) and a disclaimer about hypnosis because paranoid hypnotherapist lawyer.
Once I was happy with the script, it was time to *gulp* record.
This took several sessions: the rule of thumb is that you will spend five to ten minutes recording for each minute of recorded audio. I found I did a little better than this but it was still more than I could do in one sitting. Before each recording session, I drank a lot of water or some nice tea, starting a while before I was ready to record. (Then I went potty so I wouldn’t have to stop while I was in the groove.) I turned off my hard drives, got everything set, and spent a few minutes reading lines or just talking randomly into the mike while watching the signal meter to get warmed up and my voice profile back where it had been previously. Then I started recording.
I did this via a method called “roll and punch.” Google it for specifics, but basically, you hit record, wait a beat, and start reading. You read for several sentences or until you screw up, then stop the recording. If you screwed up, delete the screwup and then start recording again with the screwed-up line. In either event, go back and look at what you did. If anything looks iffy, rerecord it. You can record onto a new track by hitting R instead of Shift-R, and then copy-paste repaired phrases. You can also spot amplify words which are too soft or too loud (if you put a negative number in the amplification box, it will make the selection quieter.) And if any pauses are too long or too short, delete some interval or paste in a little room tone until it flows.
If you’ve set up your space right, the only ambient noise should be noises YOU make – breathing, clothing rustle, keyclicks, mouth noises, the occasional keen of existential despair. It’s easier if you edit that stuff out as you go, and try not to get it in the first place if you can. Adjust your clothing between rolls. :) As far as breathing, you can take that out (either with editing or by deleting it after the fact) or leave it in. There is a lot of debate about this. I usually didn’t have it in – I took my breaths between rolls. If I did, I didn’t worry about it unless it was obvious and distracting. Even if you take out the breaths, DON’T make the reading “breathless.” People expect pauses in speech. Making the whole book one long run-on sentence will not contribute to a quality listening experience. Leave yourself little gaps between sentences (or linked sentences) and you can edit them to fit until you get your tempo just right.
For another example, I have a habit of moistening my lips when I start talking, which is fine – it reduces lip-smacking noise. But it makes a very distinctive poppy-click sort of sound which I soon learned to recognize on the histogram. When I saw it, I deleted it. They only last about .07 seconds so it didn’t affect my inter-word or inter-sentences pauses or anything. If you have a similar habit of some kind, edit out the noise as you go, so you don’t have to take out hundreds of them when you review your mastering file. The mike will probably pick up your clicking “stop,” so get those as you go too.
Note also that you will probably want to slow down your speech a bit from conversational tempo: although it doesn’t really mean much, the usual figure one sees is to read “about 10%” slower than you talk. ACX says the optimum speed is about 9600 words per hour. Don’t feel like you have to check a stopwatch, but give the listener time to enjoy the story and don’t rush. Stop your roll often, rinse your mouth with a little water, take a few breaths, and keep your speed and tone nice and even when you start recording again.
Every time I had to stop a session, I marked the script with where I left off, and my warmup the next time was to read the paragraph or so before that a few times. I hit save on the script only when I made a change or stopped for the day, but I hit save on my Audacity project every other punch or so. I mean, constantly. It took four sessions (some longer than others, one cut very short by the aforementioned dog going for the North American Bone-Chewing Noise Record) to finish the book, roughly four to five hours in all.
Once I was done, I saved the file, then saved a copy of it for mastering. NEVER MUCK WITH YOUR RAW AUDIO FILE. Save it and back it up, preferably immediately. I have Time Machine so my file got backed up automatically, but if you don’t have a similar automatic backup, copy it onto another disk or something every so often. I went through the mastering file and decided that I wasn’t going to use a compressor/gate/etc on it because it looked fairly clean, and I wanted to see what ACX thought of it. Had I been going to use such a filter, I would have done it to the mastering file.
I waited a day and then did a final quality pass through the mastering file. I found one session join which I hadn’t fixed (embarrassing) and some pauses I decided in hindsight were too long. I also did a little (I never went above +2dB or below -2db relative amplification) work in amplifying words I thought were still a bit faint or a smidge too loud. Then I got ready to upload that mamma-jamma.
I knew that ACX requires you to submit a separate sample file between one and five minutes in length: I selected a portion of the recording which I thought captured the flavor of it without providing any spoilers (and wasn’t graphic) and copy-pasted it into a separate Audacity project file. After a little tweaking and making sure I had 1.5 seconds of room tone on either end of the sample, I converted it to an MP3 file with the correct settings (see above) and listened to it with iTunes. It sounded fine so I went back to the main project file, did the conversion, and spot checked it as well. Again, fine. To the ACX upload page, me hearties!
I hadn’t actually created the project yet, so I did a New Project and it showed me a list of my KDP books. I selected the right one, and it filled in the blanks on the book information page. I did some tweaking, and then went on to the “upload data” portion. The part before uploading your data is very similar to uploading KDP – it asks about rights, territories and so forth, nothing complicated. It does NOT ask about pricing. (Why in a minute.)
Upon arrival to the upload page, it wanted the following:
1) Cover art. Which must be SQUARE. Not book-shaped. Square. They will make you a text-on-colored background cover if you want, but I had a nice cover from the book, so I just made it square in Photoshop.
2) Opening credits file.
*record scratch noise*
Say what, now? The opening credits are in the project file, like I thought I was supposeta!
Uh-uh. They go in a separate file. As do the closing credits. Fortunately, I had left a nice space between the credits and the book content, so I just copy/pasted both the opening credits and the closing credits into their own separate projects, converted, and resaved the master without them. To be honest, I’ll probably do this EVERY time, as I put the credits in the script anyway. Leave a nice pause, lead in and lead out, and you can do this just like I did it here. (Do NOT leave them in the master. That’s just tacky and would probably get you rejected for unprofessional conduct or roughing the kicker or something.)
3) Closing credits file (see #2.)
4) Content file(s.) No file can be larger than 170 MB, which at the settings they prefer is about two hours of recording más-o-ménos - my 45 minute book came in at about 63 MB. You can break the files up any way you want if your chapters are shorter than 2 hours: you should break each file into chapters, as that makes it easier for the listener to skip to where they want to go if they’re not starting at the beginning. My book is only one chapter (it has a few scene breaks, but I just left a slightly longer pause between them as they’re pretty clear from context.) So I uploaded it in a single file.
5) Sample file – 1 to five minutes, should not include credits, SHOULD be a mostly complete scene or at least a complete sequence of event or dialogue which interests the reader. I picked the part where my protagonist comes up with his Evil Plan. Be sure to leave a little room tone on each end so it doesn’t pop in. POPS ARE BAD. Also, note that iTunes (of COURSE you’ll be in iTunes, this isn’t the STONE AGE) just uses the first five minutes of the main content as the sample, so make sure that if you have any disclaimers or anything, they’re in the main content file, and that somewhere in the first five minutes there’s at least a sentence or two that is actually, you know, entertaining.
So I uploaded all that, after going through the rigmarole described in #1 and #2. When everything was uploaded, the “I’m Done” button on the upper right corner of the screen lit up. I pressed it, and got an “Are you sure?” dialog. I clicked on “You’re Damn Skippy,” and was put back out into the “Projects” page, with “Dream a Little Dream” now showing as a completed project with a status of “Awaiting Audio Review.”
Now, if you recall, I said I’d talk about pricing in a minute. You probably don’t read that fast, you slacker, but here we are anyway. Here is how pricing for ACX audiobooks works:
1) Audible decides what the book’s price will be.
2) Audible sells the book at that price.
Yeah. I was surprised too. But that’s how it works. There is a little list of “typical prices” on one of the help pages that ties price roughly to length, but the bottom line is, Audible decides how much to charge for the book, and if you don’t like it, door’s to your left. You get a royalty rate based on whether the audiobook is exclusive to Audible and whether you hired a narrator (ACX calls narrators “producers,” but that’s who they mean) and agreed to give them part of the take. The most you will get is 40% for an Audible exclusive with no shared royalty. Note, however, that “Audible Exclusive” means that no other audiobook publisher can sell the book – it’ll be available on Amazon, iTunes, and Audible.com, at the very least, even if you go “exclusive.”
Anyway, once it’s uploaded, you just wait for review and go-live, just like a KDP e-book. So at that point it’s out of your hands. I read a lot of forum posts about problems trying to learn what not to do and I found that ACX’s initial rejection message, if you get one, can be sketchy, but they will provide specifics (“It’s too quiet,” “The volume level is too inconsistent,” “There’s too much background noise,” “You sound like somebody trying to simulate English speech using only tones generated by gently caressing amorous bullfrogs,” etc.) if you ask.
Probably THE single most common complaint I heard others say they’d gotten was uneven volume. You MUST keep your volume levels consistent: if the listener has to keep adjusting the volume to hear you/avoid being deafened, they will not enjoy the experience, sayeth ACX. (And I think they have a point.)
So you want to watch for things like your P’s and T’s being really loud (those are what are called “plosives,”) your S’s being too drawn out, or the beginning of your sentences being a lot louder than the ends. (You could be running out of breath, you could be overemphasizing the first words, you could be moving toward/away from the mike as you speak.) Watch the histogram (that funky graph that looks like something out of an earthquake-based disaster movie.) It should stay relatively consistent. There shouldn’t be huge spikes (unless at some point you yelled, on purpose, and even then don’t get carried away some of us have sensitive hearing you scene-chewing hack.) There shouldn’t be places where there’s speech but the graph is really small (unless you whispered, and even then don’t make people strain to understand what was said.) If you have issues like that, look into that Craig’s Compressor program I linked to way up above yonder. It can help smooth stuff out if you don’t want to re-record it.
So there you have it. I was going to end this blog post here, but while I was writing it, the status of my audiobook went from “Awaiting Audio Review” to “Headed for Retail.”
HOLY CRAP YOU GUYS IT WENT THROUGH IN ONE PASS. ONE. PASS.
It took just over fourteen hours to go from upload to “Headed for Retail.” Wow, that was fast. Granted, it’s short, but I’ve had KDP books take more time than that! So now I guess I just wait for it to hit the (virtual) shelves. Once it does, or if there are more steps in between, I’ll blog about that too.
Because, like I said, I love you.
Thanks for reading: if you have questions, please post them in the comments! Keep coming back, because I’ll have free copies to give away when the audiobook goes live!