Page 1 of 1
Posted: Tue Apr 09, 2019 2:30 pm
I've got a question about normalization...I am making a sample pack which will total about 400 audio samples. I'd like to normalize all the samples, so they peak at -6db, so there is more consistency in volume throughout the sample pack. I use Ableton Live and it seems the default normalization level is 0db and this can't be changed. Is there a programme I could use to normalize these audio samples at my desired level?
Posted: Tue Apr 09, 2019 2:34 pm
ive twisted wave and you can do that, its about 70 bucks
and does batch processing
just checked audacity which is free allows you to dial in the level you want normalising
Posted: Tue Apr 09, 2019 2:43 pm
Have a look at this video. https://youtu.be/q9SQp-NhVjE
So you just need to follow that process, but set the sample maximum level to your preferred normalisation value. Easy.
Posted: Tue Apr 09, 2019 2:55 pm
Bear in mind that normalising the peak levels may not equalise perceived volume as well as you hope. If all the samples have a similar Crest factor it will be okay, but if not you will still have volume variations.
There are batch processing options in some of the R128 loudness normalising and metering programs that do a better job of matching loudness.
Posted: Tue Apr 09, 2019 3:15 pm
I would be VERY wary of normalising. Simply setting a high and a low note to the same peak level will rarely result in some that sounds subjectively natural. If that's not the object of your exercise then this may not be a problem, but if were something like a piano or a flute, for example, it may give some very odd, lumpy results.
Posted: Tue Apr 09, 2019 5:39 pm
The Elf wrote:I would be VERY wary of normalising.
Excellent advice. If I really need to bring the volume up I use something like a compressor or limiter which will give a warmth to the sound as well. I don't think a sample pack needs everything to peak at the same level though.
I'm not an expert and most of the things I do are finished in a studio by an engineer.
Posted: Fri Apr 12, 2019 8:40 am
Worked on a large number of sample packs over the last few years (and bloody boring work much of it can be...!!).
Where the files are purposed for easy assembly by the end user into a functioning mix (a ’toolkit’ if you like), then you should be wary of batch normalisation. Worst case - you may even find that different files of the same instrument end up with different perceptual levels.