Slice chains and normalizing audio

franq · 2023-03-24T22:47:09.369Z

Hi guys,

I’m preparing some sample chains and I was wondering about normalizing every sample of the chain, is this a good practice? or could this broke the dynamics if for example want to launch a sequence using different samples of the chain in the same track and every of them are normalized?

I hate when I’m switching samples over the chain and I found the level droping from one sample to the other so that is why I’m thinking on reworking my chains but I found this question about dynamics so I’m not sure how to proceed

yeah I could use a compressor but…

how do you work your chains?

Microtribe · 2023-03-24T22:49:11.955Z

Theres no right or wrong answer. Do it the way you want to do it.

Your music is yours, no one elses.

warpigs330 · 2023-03-24T22:53:01.140Z

To expand on the answer above, if you are trying to get varying dynamics of drums or something like that, normalizing might be counterproductive, whereas if you have lots of let’s say hi hat samples at various volumes, but when you play them you want them to be consistent, then normalizing would make sense. It just depends on what effect you want.

For your specific situation it might be worth adjusting the volumes of each of the samples manually to get them consistent, but still with the desired dynamics between them.

franq · 2023-03-24T22:59:52.315Z

I see…

to be more specific I want to make chains grouping drum hits so one chain would be for Kicks, other for snares and so on so I can switch over them, the thing is I tend to use more than just one sound on the track so if I normalize all the hits at the same level I’m worried about losing dynamics or sounding “over compressed”

Microtribe · 2023-03-24T23:03:17.040Z

So there you go, you’ve answered your own question. Dont over think it, or over complicate it.

Dynamics are cool. Over compression is not. Thats my opnion like, but its true

franq · 2023-03-24T23:12:07.404Z

thanks for the feedback to be honest I wasn’t very sure about doing this, you saved me lot of time reworking my chains

warpigs330 · 2023-03-25T14:32:47.476Z

I think it just depends on the samples, Like if one kick drum sample is aggressive tonally but quiet, and another is mellow, but loud, that doesn’t make for great dynamics really. My answer would be that instead of taking a one size fits all approach of normalizing or not, I would adjust each sample volume until all their dynamics mesh well with each other, with the tonally soft sounds actually being lower in volume than the more aggressive ones. Like if your favorite big bass drum sound is just quieter than the rest, the dynamics are working against you.

In a recording of a drum break the dynamics all make sense relative to each other because they all come from the same recording, and if the drummer hit a drum harder it would be both louder and more aggressive sounding. But as soon as you take things from two different recordings the dynamics no longer match. One recording might be twice the volume of another, so a ghost note on one would be louder than the main note on the other. If you just take those two samples and don’t adjust the volume, you don’t really have intentional “dynamics” between the different samples, they are just different volumes.

But again, your ear is the key, if it sounds good to you then it is good. But when you say “I hate when I’m switching samples over the chain and I found the level droping from one sample to the other” that sounds to me like just one or two samples need manual volume adjusting rather than a one size fits all approach.

franq · 2023-03-25T23:52:28.918Z

that’s the point, and that is what I’m finally doing, instead of normalizing all the samples from the chain, I’m picking the ones I found “weaker” and pushing them a bit, I think this will be the best solution, thanks a lot for your feedback!

SKull · 2023-03-26T00:17:24.791Z

Normalization doesn’t change dynamics.
It just adds |(0 - the loudest point)| equally to every sample*. The dynamic stays the same.

*Sample as in sample rate, not as in sound sample.

sezare56 · 2023-03-26T02:12:55.610Z

Auto normalize is unacceptable.
(Cf Digitakt)

sezare56 · 2023-03-26T02:19:39.293Z

I agree with @Microtribe
Just make music.
Depending on audio source, even normalized, level isn’t appropriate with others so…
Wtf…