Take the 2-minute tour ×
Sound Design Stack Exchange is a question and answer site for sound engineers, producers, editors, and enthusiasts. It's 100% free, no registration required.

Hey All,

I'm interested to hear everyone's thoughts on normalising sound libraries (or 'normalizing' as Firefox seems to want me to call it!)

How do you go about it? -0.1/-0.3/-1/-3dBFS? Not at all? Play it by Ear? I've seen quite a lot of divided opinion elsewhere, intrigued to see what people here do.

Many thanks!

share|improve this question

3 Answers 3

Surely the idea of "normalising sound libraries" is dependent on content? I surely would not want ambiences normalised to the same level as a gun shot...

share|improve this answer
    
@tim attenuating a commercial library ambience to -45dB so it plays right = not fun –  Stavrosound Oct 3 '12 at 8:06

I agree with @tim that normalizing depends on content.

I've written up what I call a "style sheet" for our team. It's guideline of how certain effects should be normalized. It's not rigid, purely a starting point, but is something like this:

  • Specifics: -10 to -3 dBFS (door knob moves on the low side, gunshots on the high end)
  • Atmospheres: -12 to -8 dBFS (birdsong on the lower end, traffic higher)
  • Wind: -15 to -10 dBFS (higher end for the gusts or piercing whistles)
  • Room tones: -15 dBFS, sometimes lower

We usually discard any sound below -15 or -20 dBFS as it is too quiet or hissy to be useful for our projects.

Of course, it all depends on the content, and how you like to use it. Each effect's level is customized.

Whichever your choice, avoid normalizing everything to the same level. This is murder to audition: wind will be the same level as a metal crash or jet pass.

Paul

share|improve this answer
    
interesting. I've found that with 24 bit recording i can easily add 25 db or more of gain to a sound with no degradation if needed. I tend to cut soft :) –  Rene Oct 3 '12 at 15:01
    
@Rene 25 db untainted increase sounds great! Cutting soft sounds like a good idea. Unfortunately, I think my microphone has a poor S/N ratio. It's amazingly lush for prominent material. Great soundstage. However, haven't had much success getting good results from quiet specifics. Matrix setting plays a part as well, I believe. –  Paul Virostek Oct 3 '12 at 23:05

Just peak normalizing does not really matter. Normalizing by RMS level could get closer. But ideally the files should be balanced by ear, taken that it's done with care so that you aren't gain riding or equalizing the files excessively to achieve that, but just balancing them mildly without causing some irreversible changes to the sounds that may not be fit for someone's intended use. Or if you apply considerable amounts of processing, then you should supply the end-user with both processed and unprocessed versions, so they can make the pick of which ones to use.

share|improve this answer

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.