• We're currently having issues with our e-mail system. Anything requiring e-mail validation (2FA, forgotten passwords, etc.) requires to be changed manually at the moment. Please reach out via the Contact Us form if you require any assistance.

How Audio Works/Avoiding Clipping

mobius017

Aspiring ∞ Creator
Apr 8, 2018
1,993
Clipping has been a bane of my existence for as long as I've been trying to produce. It's annoying as all heck trying to make a song, having it sounding good in the DAW, and then being unable to render it because it warns you that it's too loud in spots. As I understand, it's related to whether or not a WAV file (I guess?) will be able to support the volume level that you're putting out, or maybe the amount of audio data?

It seems like it should be simple to avoid having the volume level go above a certain amount, but I've noticed that the effect seems to be cumulative. For instance, in my current project, I have a crash sound right at the start of the song, and it seems to be the cause of a clipping warning maybe 15-20 seconds later, when all of its playing should be over with (i.e., there shouldn't be reverb from it at that point or anything). I say it seems like the cause because if I mute it, no clipping warning occurs.

When the crash is soloed, it fills the master bus's volume meter and drains away within a few seconds. But it seems like when the other instruments are there, it's like it must not drain away as quickly, if that makes sense, since the other instruments are filling the master bus also.

If I think of the master bus as sort of a bucket being filled with water, it makes sense. But I don't see why that should be the case when we're talking about audio input from separate sources that either is or isn't there. Why should the presence of other instruments cause the crash's audio data to drain away from the master bus more slowly?

Just curious if someone who knows more about how audio processing works might be able to explain this.
 
  • Like
Reactions: WyndReed

inactive

Passionate Fan
Jun 27, 2019
179
How long is the crash tail? If it's 100% gone by the 15 second mark then something else might be at play.

The master bus is a literal sum of all your tracks, and if they were all playing exactly the same thing at exactly the same time, then the result would be very loud. It's just basic addition. However, in the real world you have different pitches, timbres, and timings, which makes the summing much more unpredictable. (One reason why EQing in solo is often a bad idea.)

Regardless, when I read your statement about the solo crash "filling" the master bus meter, my immediate thought is that it's too loud in the mix.

If you just want to render successfully without changing anything, you can always render to a floating point format. These formats allow for really loud spikes, which can then be fixed in mastering. But honestly, you shouldn't be relying upon that, and it's kind of a terrible suggestion. But it would get your track rendered!
 
  • Like
Reactions: mobius017

mobius017

Aspiring ∞ Creator
Apr 8, 2018
1,993
Thank you for the explanation! The crash is 100% gone after about 6 seconds. The first instance where clipping was occurring was around 32 seconds.

I've got the mix working at the moment. What I ended up doing was basically taking the volume of all the instruments down by maybe 3-4 dB, leaving only the vocalists as they were. Surprisingly, at least to me, I couldn't really tell the volume difference. I figure that's either because 1) the difference was relatively small, or 2) maybe Studio One is doing something automatically to sort of balance them out/deal with how loud they were. Or maybe 3) they're filling up the available audio "bandwidth"/"space" without clipping, and when that happens they all seem a little less "powerful." When all the instruments got quieter (including the crash), the crash surprisingly sounded louder to me, which to me suggests option 2 or 3....

By default, instruments in Studio One are set at like 0 dB, and while I've been happy just to leave them there, I'm thinking that they probably should usually be set lower than that. When they're all at 0, there isn't much headroom to expand into. And it seems like you can apparently set them all lower and have the overall experience not really change.

As I was changing the volumes, new instances of clipping emerged, which sort of also suggested options 2/3, as well as the uncertainty you mentioned. So I just gradually took the instruments down by a dB at a time until the clipping stopped.

As a side effect of this work, I found a new merit in the Mixtool plugin, with which you can adjust the volume of your track by +/- some number of dB. I would've thought that would be useless unless you needed to go above the track's volume slider, but it's also very useful because it's relative. So when I adjust the volume of the instrument as a whole up/down (i.e., with the track's volume slider, or iZotope's Visual Mixer), whatever the Mixtool was doing can be left alone because it's working relative to the instrument's basic volume. This is much better than specifying volume changes to the instrument slider itself via automation, because once you've set that to an explicit number of dB at various points, those won't change unless you go back into the automations and change them (I assume.). So part of my work was also to change volume automations I had made on the tracks themselves into automations made on new Mixtool plugins that I added to the tracks.
 
Last edited:

inactive

Passionate Fan
Jun 27, 2019
179
Never be afraid to lower the volume of your tracks. Most modern DAWs and plugins work in 64-bit floating point, which essentially gives you 52 bits of resolution for every 6 db of signal. This means you can go very quiet without losing information.
 
  • Like
Reactions: mobius017

mobius017

Aspiring ∞ Creator
Apr 8, 2018
1,993
Thanks again for the feedback! Studio One's default volume is 0db. I'm not even sure if you can do it, but I'm wondering if it wouldn't be a smart idea to change that default to maybe -10db or something. I can always turn up my headphones while I'm working, and I think the difference can be made up during mastering. That would give me increased headroom by default so I don't have to worry about clipping issues as often. I don't know what the rationale was behind the default 0db choice, but I can't really think of a downside to doing this. Can anyone else?

Edit: Oof. If anyone knows differently, please correct me, but it looks like Studio One doesn't have an option to set the default volume for tracks (i.e., make a setting somewhere so that when you create a new track, it automatically is set to -10db). Bummer. :ring_ani_lili: No matter, you can control volume after the fact with VCA, MixTool, individual faders, etc. Just would've been nice to set it and forget it, at least for the most part.
 
Last edited:

Users Who Are Viewing This Thread (Users: 0, Guests: 1)