Thread: Let's talk audiobooks!

Hey all,

New here, but I'm happy to field any questions you might have about audiobooks. I'm a narrator and producer so I'm happy to chat about anything audiobook related!

As a narrator, do you do work to specifically improve your voice? Like exercises, lots of reading out loud, and that kind of stuff?

Before I work, I do a couple of tongue twisters or do some other exercises to warm up. It's much like singers do to warm up.

but really the best things for anyone who uses their voice are rest and hydration. Drinking water is of critical importance. If your chords get dried out, not only does it affect your voice, it can also be damaging.

Ah, okay. Thanks.

What all goes into producing an audiobook? I wouldn't even know where to start, are there programs around that give you the howto's and if so, which ones do you recommend?

A few sort-of vague questions I guess: one's about staple effects. What kind of post processing is done to the recording exactly, besides the situational stuff? I was planning to look this up but never got to it. I mean besides the proper mic and a quiet-room, which I bet is pretty much just an absolute necessity, I guess, how is the recording modulated for clarity? And like I've noticed voices are given more presence somehow, also, and I'm pretty eager to hear what goes into it, like whether it is an interplay between narrator and equipment or ALL just genetic talent and conditioning when the proper mic is in place, or the acoustic properties of the sound room, whatever. A good example I guess would be Seth MacFarlene, flexible with voices it seems but they give him, his whole cast really on Family Guy, some kind of weird presence that appears hard to emulate without imbuing a kind of edit into the recording, different from interviews of him. Could possibly be this holistic illusion too... don't know, like this cognitive abstraction formed at the collection of media forms, like a bloated impression of presence which isn't there when the periphery aspects are removed... But if not, is it a matter of balancing bass... gain? Or what?

And how long, typically, is a take? Do you set a length relative to time or by chapter or paragraph or is this all situational and dependent? Is it difficult to splice it together and retain a sense of flow and even pitch? Or are slight variations in pitch even normally noticed post production, in your experience, when they are not intentional flourishes I mean? I ask this for a couple of reasons really. One's because I wanted to know if you set something akin to a "base pitch" which is most often returned to between dialogues. Or is this base pitch also just speaking voice and the rest is clear pronunciation and eloquence? But the other reason is that I was curious if you pre-form an emotional understanding of the section you are reading and base the narrative pitch changes on that. Not the dialogue I mean, but the narrative itself. Is reading tempo sped up and pitch increased in cases of intense action or emotional tension? Or should the dialogue's pitch changes stand in contrast to a baseline narrative? Or is this again a situational thing based on interpretation and the material?

Ugh, I had one more question, but it completely escapes me. But thanks a lot for your time, man. I appreciate it. Sort of serendipitous I happened to see this thread.

Edit:

Oh yeah! If one can hear saliva and whisper-y air in the recording, like besides the hissing that comes with background noise in a congested area with a cheap mic, is it because one is too close to the mic, is it the equipment itself, or is this just a property of voice and must be conditioned out?

Wow! So many questions! Love it!

Ok, I'm not sure I'm going to bullet point each answer to the questions, but hopefully the following information will answer most of them. If not, ask away.

Audiobook recording is really a unique genre. Think of it as a one-person play.

@Gumby - The most popular free software is Audacity. It will have everything you need, technically, to create an audiobook. But unless you have some serious acting chops, I don't recommend you make your own. I don't say this because I want your business. Not at all. I say this because you want your baby reflected in the most positive light. You didn't write your manuscript with crayons and then have your Jersey Longshoreman Uncle Billy edit it, did you? Then don't do the equivalent with an audiobook. If you can't afford to have one produced correctly, then don't have one produced. Wait until you can do it right. If you plan to distribute to Audible, a self-produced audiobook will get killed in the reviews and ratings. Your book could be the next coming of JD Salinger, but if the audio and narration quality is subpar, your book will get slammed. You don't want great writing to be hampered by mediocre narration.

Everyone has their own way of producing, but there are some general things we all do. First, you can expect a 5:1 ratio of work to product. So for every hour of finished audio, it will take you 5 hours to make it. So your average 11 hour audiobook will take 55 hours to make. Working full time on it, about a week and half to 2 weeks. For a book that long, I usually tell my clients 3 weeks. This gives me room for other day-to-day projects that come in, or god forbid, I get sick.

In my case, I try to take the technicals out of it. Because like any voice acting it isn't how your voice sounds, it's how you tell the story. So I sit and read the book. I do it straight through a chapter at a time, changing my voice for each character and narrator in real time. No it isn't always easy, especially if there are lots of characters. Again, think of it as a one-person play. After a chapter is recorded, I edit. This is where the real magic comes into play. Editing an audiobook is very nuanced and requires a good ear to get the right flow, especially the transitions between narrator and characters.

As far as noise and breaths: You'll want to edit out mouth noises, but for the most part, leave breaths, unless you sound like you're gasping for air or sound like Darth Vader. I generally tone down the volume of breaths at the beginning of sentences or replace them with room tone. Leave breaths in sentences because taking them out is unnatural. When someone is talking you expect to hear breaths. It's natural. Leave them in.

And if you do need to add space or eliminate some noise, never ever mute the space. Replace it with room tone. I keep a half second file of just room tone. no noise, just the sound of the room. I keep it in the clipboard when I edit. So if I need to add some space or whatever, I paste in room tone. It keeps the audio from sounding edited and improves the flow. It sounds like one long take, which is what you want.

And avoid Levelator if you can. If your room is so noisy, that you need to gate it with levelator, stop. Adjust your room, get a better mic or hire someone. Levelator simply drops the spots between sentences giving a weird, edited sound, almost robotic. Remember what I said about room tone? yeah, levelator does just the opposite. blech.

Once the whole thing is done, I mix and master it based on the distributor. Audible has a crazy long list of things they need in the file and what levels they like it mastered with. I don't compress or EQ. I do a bit of limiting to cut some of the louder spikes. But other than that, I like a dry sound. For most audiobooks, less is more.

Phew! I think I hit all the major points. If not, ask away!

You are right about Audacity. It has everything needed for basic audio editing. Plus, the pushbuttons look just like the buttons on an Ampex 350 tape deck, so I feel right at home.

I have an agricultural radio show that I produce using a handheld digital recorder and Audacity. Many of the interviews are with farmers or field extension agents, so a lot of editing is needed. The trick is to take out all the stumbles and repeats and uhs without disturbing the rhythm of the person's speech. I deliberately hunt for locations with the right kind of background noise - pigs, cows, chickens, sheep, a distant tractor, - so the listener understands this is no studio production.

You mention limiting to get rid of spikes. Do you go through the file and reduce the amplitude of the spikes individually, or do you use a limiter? I do them by hand, one at a time, but my programme is only 14 minutes, so it dosen't take very long.

When I mentioned spikes, I meant places where the volume is too loud. Like if the character is yelling, that place might be louder than other places. But you don't want to actually shout loudly into the users earbuds. So I just run a limiter filter across everything at -6db. In other words, anything louder than -6db is cut down automagically.

So you're not dealing with individual voice spikes.

Look a few posts down and you'll see a post called 'Digging in the Dirt' with a link to my one and only youtube contribution. It has been said that every total idiot has a blog and is on youtube. I don't have a blog so I don't qualify as a total idiot. Give me time.

The audio was recorded and edited in my home studio in Consejo Village.

Sure, sometimes, I have an individual voice spike. I lean in a little too close for half a second. But running a limiter pass on everything removes those, too.

I confess I have never thought about the technical process behind the creation of an audiobook, or that it would take so long to create, although of course now that you've said that it makes perfect sense. If you don't mind I'm going to add to your question list.

You mentioned to think of it as a "one person play" have you ever been involved in an audiobook which had a number of people recording, or even just two people? If you have obviously the dynamics would be very different as you would interact but how else would it differ? Would it take a longer production time? Would it be a completely different set up or simply the same with more mics?

Originally Posted by Sara

You mentioned to think of it as a "one person play" have you ever been involved in an audiobook which had a number of people recording, or even just two people?

I've never done a full cast audiobook where characters are interacting with each other in dialog, like a radio play. Those aren't done as often due to the extra time and expense (gotta pay all those people to narrate).

I have done several books with multiple narrators. Each chapter was written from the perspective of 2-3 different characters of different genders, so I split my time with some female narrators. We each took the chapters that were told from the appropriate gender perspective.

This process has it's own challenges. Since each person is working remotely from their own studio, I (as producer) have to make sure the audio all sounds the same in terms of quality and room tone. This is not easy.

Also, we collaborated via DropBox, which was a huge help. We had a "characters" folder that contained snippets of each character. So if a character appeared in multiple chapters, each narrator would know how that character sounded. It wouldn't work if I read Bob Jones with a NY accent and the other narrated him with a British one.