Does It Make Sense to Oversample a Hard Clipper?

[crickey13]

As far as I understand, hard clippers are used predominantly to trim drums on a buss or a finished mix to gain a few dBs, producing short bursts of distortion that are largely imperceptible to the human brain because of their very momentary duration.

But if hard clippers are used specifically to clip short drum transients, meaning that they work on inharmonic signals for the most part, does it really matter whether the clipper is oversampled and does not alias? They work on inharmonic signal anyways and besides, such short bursts of distortion aren't such a big deal because we don't really hear that and clipping drums can actually improve clarity.

Does a non-oversampled hard clipper produce a brighter sound than an oversampled one, is there any perceptible difference between the two?

[DCrown]

Well, like always use your ears to judge, oversampling is not always the better way

[chk071]

I could be wrong, but, I doubt that you'll come by a hard clipper which isn't oversampled these days?

[plexuss]

The issue is more to do with the density of the clips - the number of clips in a period of time. This is because, as you say, one clip spits out distortion but its so fast its imperceptible. but if there are a lot of bursts in a period of time the distortion becomes perceptible. I hear it as a kind of broad band "fluff" that fills in the "space" between the short term dynamics sucking up dynamic range or "contrast". I don't find it pleasing and work to avoid it as much as possible. So, I generally oversample clipping because I prefer the sound. YMMV do what sounds best to you.

[crickey13]

Well, like always use your ears to judge, oversampling is not always the better way

I know, I just wonder whether there is any technical reason or rationale for oversampling hard clipping. I guess it's more of a theoretical question.

Feeding a hard clipper a burst of noise should result in a burst of noise.

Also, Fabien from TDR has stated that hard clippers are largely resistant to aliasing, I saw it in a GS thread.

I could be wrong, but, I doubt that you'll come by a hard clipper which isn't oversampled these days?

That's because they usually offer soft clipping as well and oversampling soft clipping is an absolute must.

[Frankie.T]

If there is a process who need oversampling, is certainly clipping.
It's the process who make more distortion than everything else, i would never use it without OS

[ampetrosillo]

Aliasing makes little to no difference, in my opinion, on enharmonic signals such as drums. Any distortion is going to produce lots of harmonic and enharmonic content anyway, at most you're gonna get some more distortion up in the treble region where our ears are not as sensitive anyway (but it can still appear as harsh, which is why oversampling does little harm if present). That is even more true of hard clipping that basically produces content not quite up to infinity but in a very wide band, some oversampling might reduce the distortion by a few decimal points but it's unlikely to make much difference.

[vanerio]

In my experience, it removes aliasing if you check it in a spectral analyzer, but the difference is not really audible.
I think it could make sense in the master, but it is not a must.
In my opinion hard clipping just a few decibels, does not really need oversampling.

[Ploki]

Non oversampled clipper has a harsher sound because it also produces all the mirrored frequencies. Depending on how hard you drive it, they might mirror well into the lower mids.
Especially because long bass wavelengths produce a long high amplitude clip - but there’s also other things usually present in the same time window:

I prefer adding aliasing with lofi resampler/bitcrush plugins when i want it. On drums it’s often pleasant but a lot of clippers are meant for mastering duties as well, where you don’t want aliasing.

On single drum hits i just use logics bitcrusher

[Tj Shredder]

I doubt its relevant, even if there would be a difference in sound. If you could have a setting where you have the choice, try a null test and listen to the difference…
On the other hand, drums do have tonal aspects, but most likely in the low end…