An Engineer's Perspective

Tiffany participates in a few Facebook groups dedicated to audiobook narration.  I monitor them too but don't often post even when there are technical questions posed.  Why?  Call it extreme apprehension as I've seen many civil discussions within comments sections turn ugly as people with opposing viewpoints lash out at each other like cavemen with inarticulate verbal sticks. "Me right!"  *Bonk* "No!  You dum dum!" *Crack*. Luckily, this isn't the case on the audiobook groups but I tread cautiously anyway.  The internet is a weird place... But hey keep reading my blog!!!

Okay, where was I?  Still recovering from that first *Bonk*!!!  Oh yeah... A lot of technical questions get asked on Facebook.  And that's totally understandable as many narrators are serving as their own Engineers/Producers without any prior experience in the field.  Most have a performance background not a recording background and if you watch the videos on ACX, they say find a quiet room, a USB interface, and a mic and go to work.  It's just  that easy!  In theory, that is as basic as it can get.  But while recording you'll have to wear numerous hats simultaneously and that can take away from your main objective which is capturing a strong performance.  You'll have to make sure the levels are okay.  Record too low and once the audio is compressed either by you or ACX and you've invited a world of background noise to the party.  Record too hot and a great take can be ruined by the dreaded digital clipping.  And really, it's more important to focus on the text.  But most people have no other cost effective choice than to do it all by themselves.  Studio time is expensive and if you're starting out, you're working for practically nothing or royalties which is right next to practically nothing. I'm pretty sure they share a duplex with a very thin party wall.

This is not an expert's column.  This is more some lessons learned from our experience and as with each production, we learn new things.  Tiffany and I are lucky in that I'm a trained audio engineer and she's a trained performer.  But we still had a lot to learn when we got started.  For her it was learning to perform in front of a single microphone as opposed to an audience.  Totally different approach.  And for me, it was learning how much editing is enough and how to retain the essence of the performance.  The challenge in mixing music is getting the pieces to blend together.  The challenge in narration is to take a single voice and learn when to accept slight imperfections in order to preserve the performance.  There's no "we'll fix it in the mix!" (Which is a dumb statement anyway but that's another topic!). 

For the longest time I manually cleaned every noise possible which involved scrubbing, zooming in and trimming out any little mouth click I heard.  It was aggravating.  It sounded clean but it wasn't worth the time and effort.  Why would I do that?  Because I wasn't listening to it as a listener would but as an engineer and focused on the micro-details.  I still clean mouth noises but I use Izotopes RX3.  There are many opinions on this but MY experience is that it saves time and if used properly (and there is a learning curve) it works wonders.  I use the DeClicker set to about 4.5 and set the clicks to random.  This has given me the cleanest results without artifacts.  Others use DeCrackler... It's whatever your comfortable using and gives the proper results which is no digital artifacts which distract the listener.  If there is any other noise, I highlight about .1 second of audio and get more aggressive with a 9.0 setting.  In small doses and zeroed in tight on the offending noise, it'll clean it without distorting the audio.  If that doesn't get it, I have to decide if it's truly distracting and manually clean it or rerecord it, or leave it be and move on.

I've seen questions about debreathers.  ACX says get rid off those breaths.  Yes and No... Mostly No!  Breathing is natural. (The longer you're able to breathe, the better the odds are that you're still alive.)  It can become a distraction if the recording sounds like it's being read by someone having an asthma attack.  And some breaths can sound awkward if they aren't taken within the natural flow of the read.  But in those cases I mask them with room tone.  Now a compressed audio file can still make a natural breath sound extremely jarring and inappropriately loud.  So I will use a debreather to attenuate the breath by about -6 to -12 db.  They are still there, still natural but not jarring.  I do not use a debreather to remove anything.   

People want to know about the signal chain and getting consistent levels.  These are two separate issues but my approach makes them a single issue.  First, some narrators are fine with their computers built in pre-amps and that makes me cringe.  I've been in studios, and as a musician myself, I see the signal chain as crucial to a quality sound.  This does not mean that you should break the bank and buy the most expensive mic and preamp.  But for a narrator, the mic and pre-amp you use can bring out the right timbres in your voice and give you the right tone for your performance.  Sure the listener won't care what you used, but when the right combination of performance and sound equipment work, they will hang on your every word.  Shouldn't performance be enough?  Not always.  A great read with a thin sound that takes away the textures of your voice or buries it in tons of background noise negates the read.  Just because it's an audiobook, doesn't mean sound quality should be treated as insignificant.  It's not.  Singers are just as particular about their mic and pre-amps as any other musician is about their choice of instrument... it's an extension of themselves in many ways.

When we started, we used a Shure SM7 into a Focusrite ISA One pre-amp.  Both quality products, but originally bought for a different purpose.  And they were not right for Tiffany's voice.  The pre-amp is especially good.  It has glowing reviews all over the internet from consumer users and industry professionals.  It's a damn good pre-amp.  But it was all wrong.  Why?  Because again, it wasn't a match for Tiffany's voice.  We got work using it and did many books but we weren't happy.  It sounded like Tiffany, but didn't really capture Tiffany.  So I tried a pre-amp from Warm Audio.  I learned recording using analog gear (as in 2 inch tape analog gear) and this was a clone of an old API pre-amp.  We had also switched to a Neumann TLM 103 mic and that combination instantly worked.  There was the warmth in Tiffany's voice (thus making the company name very appropriate).  The Focusrite was clearer and more transparent but no matter what mic we used, it felt a bit distant and clinical.  But with the new pre-amp, we started getting feedback such as "really love your voice" and "very engaging read".  Even our rejections were more polite and personalized... A far more delicate blow to the ego than the cold and impersonal ACX form letter...  From "I'm a loser and nobody likes me" to "They liked me.  They didn't pick me but they liked me."  

We also use an inline compressor.  Singers do this a lot.  It allows them to really let go when the moment calls for it and project or emote without clipping, and used right, it sounds great.  And I applied the same principal to narration.  If set right, it's transparent and maintains consistent levels and allows Tiffany to really get into a scene without me having to adjust pre-amp levels or have her backup, turn her head, whatever...  She still practices these techniques but if a character has to yell... Well she can yell and the compressor keeps it in line before it commits to tape... Err um... Disk!  (Unless you are comfortable with using a compressor, I wouldn't advise this! Because you can't undo the compression once it's recorded.    Even I had to rerecord a chapter because I fiddled with the settings and it was horrible... Lesson learned... Don't mess with the knobs at chapter 39 just because your hands are bored... Um.. Yeah!)

Again, this is not a list of how to's and rules.  This is what we've learned... Or rather I have learned as an engineer and what works for me.  Many of you already have a process and equipment that works for you but for those still searching this is what we've discovered.... At least so far.  I'm always looking for ways to make my life easier without sacrificing sonic quality to Tiffany's performance.