The No.1 Website for Pro Audio
Using compression on audio book project
Old 30th November 2014
  #1
Using compression on audio book project

Hello fellow slutz!

The recording stage of an audio book is now done. No compression was applied during the recording. Dry signal recorded in a sound booth - 24 bit recording.

I am now doing the editing and post on the project. Edit is pretty much done and next up compression etc.

My question is how do you guys do compression o your audio book projects - do you usually compress already on record stage or all in the post? The audio book I am on now is the Quran and the talent is both reciting and reading the translation. It is recorded verse by verse so: recitation - translation - recitation - translation etc. The recitation parts are about 6-9db louder than the english translation.

Using compressor brings out the roomy qualities of the recording up on all the recitation parts which are hard to eq out without affecting the talents natural sound (he sounds great btw.). How do you guys level out vocals in this kind of situations? My goal is to make the levelling sound as natural as possible.

And just as further info, the whole book is about 50 hours of material so we are talking about long recordings. This is just to say that manual editing is out of question :D

Thanks so much in advance for any help! I will do testing with vocal rider and different comps etc. and will let you guys know how they work but would be cool to get any tips and trick to get started!

Old 30th November 2014
  #2
I've done 5 or 6 unabridged Audio Bibles, so this is familiar territory....
I would think the two different levels would problematic for listening in noisy environments - cars, trains, etc. Maybe contain the difference to no more than 5-6 dB.
I do compression normally in post - mild 2:1 to 3:1 ratio, fast attack and release, and set threshold to about 75% of peak. Then normalize back up to -2.
I'm also concerned that compression brings out the room sound. You might be going a little overboard. You might look at a good expander to keep the noise floor in check, or the dialog denoiser from RX4 could help.
Old 30th November 2014
  #3
Deleted 7f9cade
Guest
If the first part of the reading is 6 to 9 db louder than the translation, I would split the two parts onto two separate tracks and process each the way they need to be. That way you don't have one compressor setting for two different ranges of DB. I would bet the EQ would need to be different for each too, As the original reading is probably a more intense a performance than the english translation.

As far as processing, Minimal. HPF at 100. Maybe a slight "air" boost, and just mild 3:1 compression fast attack and fast release only doing 1-4DB Gain reduction.
Old 30th November 2014
  #4
Thanks Bill and Ienjoyaudio.

Yeah I think that I overdid the compression a bit in the first place. I tried to level the recitation and translation to same level which made the recitation sound too compressed. I realized that the translation can be a bit softer and quieter (because it is in real life as well) in the recording as long as it will be comprehensible in noisier environments as well.

Ienjoyaudio: The whole book is recitation-translation-recitation-translation so picking up just the recitation parts on separate track would take ages. This is why i am trying to get a solution that can be applied to the one mono track I am processing.

How do you guys usually bounce audio book stuff. Mp3 / AAC must be the best option? What about kbps? I have been listening to a lot of audible audiobook examples now and noticed that most audiobooks are really bass heavy. Is this to make the material sound bigger on small earbuds (which people most probably use to listen audiobooks anyways)?

Cheers!
Old 30th November 2014
  #5
Deleted 7f9cade
Guest
Quote:
Originally Posted by tv_engineer_lnd View Post
Thanks Bill and Ienjoyaudio.


Ienjoyaudio: The whole book is recitation-translation-recitation-translation so picking up just the recitation parts on separate track would take ages. This is why i am trying to get a solution that can be applied to the one mono track I am processing.

How do you guys usually bounce audio book stuff. Mp3 / AAC must be the best option? What about kbps? I have been listening to a lot of audible audiobook examples now and noticed that most audiobooks are really bass heavy. Is this to make the material sound bigger on small earbuds (which people most probably use to listen audiobooks anyways)?

Cheers!
Ahh. I see. If this is the case, unless you want to spend more time, you wont get the best results. One of the versions will be compromised. Either over compressed verse or undercompressed translation.

Vocal rider may be the way to go, then a compressor set after it. Better to set the comp for the max level. Id rather have under compressed than over compressoed.

Vocal Rider doing fader automation into a compressor might work.

For exporting just voice, Ive done 128k mp3 and it is just fine but check your clients delivery specs.
Old 3rd December 2014
  #6
Lives for gear
 
mikevarela's Avatar
 

Quote:
Originally Posted by Ienjoyaudio View Post
For exporting just voice, Ive done 128k mp3 and it is just fine but check your clients delivery specs.
No,
this is not the case.

audiobook processing is governed mostly by Audibles standards. They are strict on what they except.

Each chapter, prologue, epilogue, preface, introduction and credits must be separate files delivered in MP3 format at 192kbps mono.

They must also have .5 sec of room tone at head of each file and 3.5 sec of room tone at tail.

They must also abide by an average RMS between -18 and -23 dbfs and have no peaks over -3dbfs.

Your compression settings vary on material but overall, for you to hit the settings above, you might need to do some harder limiting.
Old 3rd December 2014
  #7
Gear Maniac
Audible Processing is, in this order (and cut/paste from their site):

1. EQ. Remove low (80hz and lower) and high (16kHz and higher) frequencies by using a high-pass and low-pass filter, respectively. Set the high-pass filter to remove sounds below 80hz, and set the low-pass filter to remove sounds above 16kHz. If available, set the Q to the highest-possible setting for both filters. Usually, that setting is 24dB or 48dB per octave.

2. Normalization. Typically, you should normalize your peaks to -6dB.

3. Comp & Limiting. To properly utilize limiting on your files, start by setting your limiter’s maximum output to -3dB. Then, turn up the gain on your limiter until you have achieved a loud, clear, and consistent sound. Don’t boost the level too high.

Also, you can save tons of time by using something like "fre;ac" (windows) to queue up and encode all the files at once.

Audible has a script that processes the files after your submission, the script fails under certain conditions, like processing joint stereo files or anything that was bounced outside of 44.1. To be honest, it's one of the few, and I say this loosely, "universal standards" I've seen. People all over the globe are hopping the same circles.
Old 4th December 2014
  #8
Deleted 7f9cade
Guest
Quote:
Originally Posted by mikevarela View Post
No,
this is not the case.

audiobook processing is governed mostly by Audibles standards. They are strict on what they except.

Each chapter, prologue, epilogue, preface, introduction and credits must be separate files delivered in MP3 format at 192kbps mono.

They must also have .5 sec of room tone at head of each file and 3.5 sec of room tone at tail.

They must also abide by an average RMS between -18 and -23 dbfs and have no peaks over -3dbfs.

Your compression settings vary on material but overall, for you to hit the settings above, you might need to do some harder limiting.
Thats interesting. I only have two audio books to my name so Im not well versed. I never received any detailed delivery requirements like that, however they did specifically ask for 128 kbps mp3. That was all.

Are those universal specs or are those specific to a certain distributor?
Old 4th December 2014
  #9
Gear Maniac
Quote:
Originally Posted by Deleted 7f9cade View Post
Are those universal specs or are those specific to a certain distributor?
There's no universal specification for audiobooks, just lots of different workflows. I have seen:
  1. Narrator records, processes and uploads the files to ACX.
  2. Narrator records, sends wav files to editor, editor finishes and sends back, narrator uploads the files to ACX.
  3. Narrator records, sends wav files to studio, moves on to next project
  4. Go to studio, eye the donuts, drink all their tea, record. Ask for lunch (pickups) or dinner when leaving.
Audible adds DRM to the file then places them in the catalog for sale, publishers selling audiobooks for download through Amazon have to conform to their standard. It's pretty straight forward.
Old 4th December 2014
  #10
Lives for gear
 
mikevarela's Avatar
 

Audible is the undisputed king of audiobook production and distribution. It's sister site ACX is a free and open marketplace designed to allow a wider body of work to get completed by a community of self taught/directed narrators and editors.

In both cases the requirements are the same, as all files for both sites end up on Audible's store.

Even Random House distributed through Audible.

https://www.acx.com/help/acx-audio-s...ents/201456300
Old 4th December 2014
  #11
Gear Maniac
 

I would think something like waverider could help wading through the huge amounts of material that needs handling.
Old 8th July 2019
  #12
Here for the gear
 

If you do compression in stages, it will even things out more without sounding over compressed.

For something like this I would use something like the Waves MV2 with the low level set to 4 and the high level set to 2. This will bring up your quieter moments.

Then through something with a soft knee, like the Waves RennAx plugin, input at -6 and output at -3.

Then follow it with a brick wall limiter with its output at -3. Adjust input for the overall level that’s needed.

Each stage should be doing about -1 to -3db of compression, but by doing it in stages with different compressors with different settings, you can get about -8dB of compression without it sounding bad.

Also, the FabFilter Pro-C2 in “vocal” mode is very transparent at high compression settings.
Old 10th July 2019
  #13
Lives for gear
 
Mundox's Avatar
Seems like a spam bot brought this old thread back.
But anyway, this is a perfect candidate for WaveRider Tg which can do offline processing now.
Old 6th September 2020
  #14
Here for the gear
 

This thread seemed apt for my needs. I'm about to record my first audio book.
I have set up my equipment after following the requirements & watching several videos from ACX.

I have also used some of the instructions & advice from this thread (thank you!).

Here're my settings (afterwards I will point out my issue):

Recording peaks at -12dB. I am recording about 12 inches from my AT 20-20 condenser mic which is shielded with a wind cover. I have many years experience as a singer & so have quite natural mic control.

At the mastering stage I use Cakewalk:

EQ - Bass HPF - 80hz & Highs LPF - 16,000hz (Both with 0 gain & bandwidth (oct) 2.0)

Compression - Threshold -6.0 / Wet output -3.0 (limit output on) Ratio 2:1
Attack & release maximum (I suspect this ought be changed)

Then I normalise by -6dB

Compression - Threshold -6.0 / Wet output -3.0 (limit output on) Ratio 3:1
Attack & release maximum (I suspect this ought be changed)

Than I normalise again -8dB

(Noise floor after mastering is about -74, this seems to good).

The issue I have is that the range of the recording is supposed to be between -23dB & -18dB, but mine occasionally reaches -27dB & -15dB. When I play back the (mono) recording after export (through VLC), the volume seems low, even though I have my laptop & VLC volumes at 100%. On re-importing the exported track, it is visibly lower in volume than it's original counterpart, so, somewhere during export it is losing volume. The only setting I have altered in the export menu is to choose mono. The recording & export are both at 44.1. Export at 32 bit.

To be honest, I'm not experienced enough to understand to what degree all these parameters effect each other. I am aware that the best way is to nail the recording in the first place, but I thought I had done so. (I am using a Zoom R8, 8 track digital recorder & have had experience recording songs with vocals, guitars, etc), so I'm not a complete newbie.

I feel as though I'm close to getting it right, but there's perhaps something I need to tweak.

Last edited by justinpbrown; 6th September 2020 at 05:36 PM.. Reason: Extra Information Added
Old 7th September 2020
  #15
Lives for gear
 

The ACX specs call for an RMS value of between -23 and -18. That's an average RMS reading over the course of a chapter. The maximum allowable peak is -3TP. So your levels can vary as much as you like, as long as your RMS average falls between -23 and -18, and you don't exceed -3TP anywhere.

Your dynamics processing seems unnecessarily complex, but if you like the sound, then whatever gets you there.
Old 7th September 2020
  #16
Here for the gear
 

Quote:
Originally Posted by seanmccoy View Post
The ACX specs call for an RMS value of between -23 and -18. That's an average RMS reading over the course of a chapter. The maximum allowable peak is -3TP. So your levels can vary as much as you like, as long as your RMS average falls between -23 and -18, and you don't exceed -3TP anywhere.

Your dynamics processing seems unnecessarily complex, but if you like the sound, then whatever gets you there.
Thanks for the clarification.

My processing in this way was an attempt to shorten my RMS value parameters at playback. Obviously, now I am the wiser the process can be simplified.

Thanks for the feedback seanmccoy.
📝 Reply

Similar Threads

Thread / Thread Starter Replies / Views Last Post
replies: 264 views: 93178
Avatar for MattFoRill
MattFoRill 8th March 2013
replies: 62 views: 45986
Avatar for marcus
marcus 10th July 2012
replies: 9967 views: 477959
Avatar for Mr Funk
Mr Funk 2 hours ago
replies: 0 views: 1477
Avatar for Guitarking
Guitarking 4th November 2018
Topic:
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
🖨️ Show Printable Version
✉️ Email this Page
🔍 Search thread
🎙️ View mentioned gear
Forum Jump
Forum Jump