VO-BB - 20 YEARS OLD!

cyclometh · Posted: Sun Nov 18, 2012 6:37 pm Post subject: Limiter vs RMS Normalization

Hey folks- I posted this to an Audiobook group on Facebook so you may have seen it there also, but I think there's more technically-minded folks around here, so I'd like to get some insight on this.

I have a question about mastering for places like Audible. Their guidelines call for using RMS normalization. My DAW (Cubase) doesn't have that function, having only peak normalization.

After doing a bunch of research into it, I determined that I could produce the appropriate results by using a combination of a limiter and compressor (or what Cubase calls a "Maximizer"). At least, it sounds fine to my ears and my RMS meter shows an average of around -19dB, which is what they call for.

I'm curious what others who work on the editing/mastering side have to say about this, and what techniques you use to meet the various standards set by audiobook production entities like Audible when you don't have the exact feature set their documentation expects.
_________________
Corey "Vox Man" Snow
http://voxman.net

Scott Pollak · Posted: Sun Nov 18, 2012 6:59 pm Post subject:

Interesting, Corey. I'm running into the same issues here using AA 5.5. There is no RMS function per se', but in the edit mode I can click "Effects -> Match Volume" and then get to an RMS function THAT way. Kinda awkward, but it works. Now I know NUTHIN' about Cubase, but I'm betting if you do a bit on Googling you'll find a roundabout way to do it.
_________________
Scott R. Pollak
Clients include Pandora, NPR Atlanta, Wells Fargo, Cisco, Humana, Publix, UPS, AT&T, HP, Xerox and more.

www.voicebyscott.com

Frank F · Posted: Sun Nov 18, 2012 7:50 pm Post subject:

You are on the right track or thinking process. I will sometimes use a limiter - NO COMPRESSOR (or any other processing for that matter) while recording for audiobooks; then use the "normalize function AFTER editing (and decreasing stray peaks volume to match the rest of the file).

This takes time, but when done properly the sound is what they are paying you for.

Frank F
_________________
Be thankful for the bad things in life. They opened your eyes to the good things you weren't paying attention to before. email: thevoice@usa.com

Philip Banks · Posted: Mon Nov 19, 2012 2:23 am Post subject:

Bish · Posted: Mon Nov 19, 2012 9:25 am Post subject:

Here's my process:

1) Record everything as aiff, raw with no processing. Try to get as good a level as possible. Shooting for peaks around -6dB and absolutely no higher than -3dB. Clipping is not an option.

2) Visually scan the file and manually tame any excessive or unusual peaks. (just a couple of dB off the top). I never apply any noise treatment. My environment is kind, and if I'm diligent with the s/n ration, there's no need.

3) Normalize to 100% and then add mild compression (usually Apple AU Dynamics compressor "Light" setting)... barely noticeable audibly. ("Sweetening" has been left out as beyond the scope of the original question... the jar of fairy dust would be uncorked around this time if needed.)

4) Normalize to -6db. Save file as "XXX Master - prepped". Convert to mp3.

As far as Audible and their "RMS normalization" is concerned, don't make the mistake of thinking that they actually know what they're talking about. In my experience (message exchanges and telephone conversations with ACX) they are not in touch with the tools used by their "producers", have no idea about what constitutes good audio, and they mangle what they do received beyond all recognition. In my opinion, they're reading verbatim out of a playbook that was written back in the days of tape. The simple fact that their published documentation regarding levels and compression is almost impenetrable is testament to that. Nearly any one of us here could write "an idiot's guide to recording and prepping an audiobook file" that makes more sense than their stuff. It would probably take five minutes.

Check any files that Audible has published recently (don't just go by the audio samples on the book's page) and they're all over the place with regards to level. However. the majority seem fall into one broad range, which my quick and simple method seems to satisfy.

Ooops... I slipped into rant mode Smile

_________________
Bish a.k.a. Bish
Smoke me a kipper... I'll be back for breakfast.
I will not feed the trolls... I will not feed the trolls... I will not feed the trolls... I will not feed the trolls.

Jason Huggins · Posted: Mon Nov 19, 2012 10:37 am Post subject:

You are totally right! The first book I did with ACX, I asked them to check my audio sample to make sure it was right. I sent them a file that I process the normal way (similar to Bish, but normalized to -3db and a tiny compression from the Pre with limiting in post) and sent it to them. They said, "If your final productions is of the same quality and standards as this sample, it will definitely pass our final QC."

Jeffrey Kafer · Posted: Mon Nov 19, 2012 10:42 am Post subject:

Philip Banks · Posted: Mon Nov 19, 2012 10:59 am Post subject:

Seriously, I have no idea what "normalise" means in this context.

Frank F · Posted: Mon Nov 19, 2012 11:58 am Post subject:

Hey Phillip,

I didn't say we were paid well. Just paid - sort of.

FF
_________________
Be thankful for the bad things in life. They opened your eyes to the good things you weren't paying attention to before. email: thevoice@usa.com

Philip Banks · Posted: Mon Nov 19, 2012 12:00 pm Post subject:

I knew what you meant. We're paid to do the job properly for the agreed fee.

cyclometh · Posted: Mon Nov 19, 2012 12:41 pm Post subject:

Jeffrey Kafer · Posted: Mon Nov 19, 2012 4:12 pm Post subject:

If you look at the more extended Audible spec, it says to use RMS normalization and then in parentheses, it says something like "should be peaking between -3 and -4 db." So when I normalize, I peak it at -3.5.
_________________
Jeff
http://JeffreyKafer.com
Voice-overload Web comic: http://voice-overload.com

georgethetech · Posted: Mon Nov 19, 2012 4:40 pm Post subject:

I have Charles from ACX/Audible on a Webinar with me Dec 11th for Voiceover Xtra.
1 hour will be covering editing techniques, incl. punch n roll, 2nd hour will be QnA with Charles and I.
Perfect time to question him and hopefully get some clear answers.
However, after speaking with him last week, they know what they like when they hear it, but don't seem clear how to best achieve it.
I need to do some mastering, send it to someone there in charge, get their approval, then teach others the right way to do it. Not the awful Sound Forge instructions they provide.
_________________
If it sounds good, it is good.
George Whittam
GeorgeThe.Tech
424-226-8528
VOBS.TV Co-host
TheProAudioSuite.com Co-host
TriBooth.com Co-founder

cyclometh · Posted: Mon Nov 19, 2012 5:23 pm Post subject:

George, let me know if I can assist in any way- I'm very interested in coming up with some good techniques for acheiving the right sound that Audible wants to hear, which could be applied to more than one DAW or workflow.
_________________
Corey "Vox Man" Snow
http://voxman.net

Jeffrey Kafer · Posted: Mon Nov 19, 2012 9:06 pm Post subject:

If they get "what they want" will they still run it through their own audio meat grinder anyway or will they leave it as is?
_________________
Jeff
http://JeffreyKafer.com
Voice-overload Web comic: http://voice-overload.com