I was thinking about clippers… Those often come with oversampling options. I think I really tried all available. To my ears, clipping only works on a bus or even master, because then the added distortion will be somehow masked in the mash of full band sounds… So also only if you mix fullband-width-alike, not classic or so
But when it comes to the oversampling options, in the end I always tend to use no oversampling. Why it is even provided, I don’t get this: Clipping is clipping, cutting of the waveforms in a very rigit way. Why should oversampling help here? Will the curve then smoothed out over some samples? Still it would result in drastical distortion on the transients. It will alias as hell, no matter of oversampling… Seems to be a voodoo topic to me.
And then do you ever use soft clipping, which simply is “fading in” the clipping? I only seem to like Renoise’s own soft clipper, sometimes it actually improves the sound, but I cannot tell under which conditions this is (Also its annoying the soft clipper is not available as device simply). What was interesting maybe, would be a softclipper connected to an env follower, or lets say a waveshaper. Maybe this is what a FET compressor actually does…
There are a lot of tutorials (also by quite skilled people) in which they use clippers everywhere esp. on the drums, the single drums, then the bus, then the master… I don’t like that harsh, edgy sound, would only fit to hard genres, IMHO. For example, in Baphy’s videos, the clipping ruins the dynamics, more specifically the sound of the transients. Maybe you need very good, fast responding monitors to have a clue what coloration the transients have. On the other hand, clipping was a thing in a lot of Amiga tunes, and it sounded great.
What are your finding with clippers? Do you use it regularly? On what kind of sounds, or only selective for specific genres? Do you use oversampling? I hope we can start some discussion about the topic here!
I think the oversampling has a very legitimate reason if you’re running at 44.1 or 48 khz. The reason is, that the clipper is a waveshaper, it will reshape the sound so that it will generate harmonics from all affected frequencies, also harmonics of high frequencies, which are then above the nyquist frequency (half sample rate, i.e. above 22.05/24 khz). This means that they will be reflected/mirrored down into the range sub nyquist, and as such possibly into the audible domain. Long story short term: non oversampled waveshapers can subtly mess up your highs, and when you use them as a noticeable distortion effect, even lower frequencies. Oversampling shifts up (doubles at 2x oversampling, doubles again at 4x, …) the reflection frequency, so that (hopefully) the reflected frequencies are no longer contained in the (filtered, then downsampled) resulting signal. Especially distortion effects can greatly benefit from oversampling, leading to more transparent, clear and defined high frequency action…!
Hard clippers are imho a good tool to tame transients, but I usually rather like to use dynamics processors (fast compressor), so I can set a threshold for the peaks of a rhythm track, and then dial in with the ratio how much it should be squashed. Using hard clippers for mastering you have to either work very accurately, or it will grind your bass kaput.
Other people use them for mastering, and then soft clippers, which are like a “shape” distortion. They will not just chop off the peaks, but have a smooth distortion curve. With such clippers you get a slight warming of sound, even distortions in the bass, and the cutting off of peaks will still work but be a bit more soft. It is a means to really crank up and make warm a full mix to -0db for (dance) music that is like…no dynamics, just bassy full throttle all the time with maximum loudness.
I also enjoy softer mastering without the aggressive clippers, so there can be more subtle dynamics and dynamic variations, and I give away the maximum loudness crap for it gladly, because I believe dynamic music is more fun to listen to. Maybe on master a soft clipper, but only subtle.
Here are a few tid bits from the Fabfilter help file to Pro-L limiter.
Oversampling
The limiting algorithm often needs to make very quick changes to the audio in order to remove peaks while preserving transparency and apparent volume. These sudden changes can introduce aliasing, which causes distortion and generally reduces the quality of the audio signal. Oversampling is a way to reduce that aliasing by running the internal limiting process at a higher sample rate that is a multiple of the host’s sample rate.
“When do I need to turn on oversampling?”
You need it more when the limiting process operates faster (using short lookahead times), and when limiting more heavily, both leading to a higher level of aliasing. Aliasing has the effect of adding spurious non-musical frequencies to the audible signal which degrades the audio quality. In addition, aliasing causes higher inter-sample peaks and these can cause distortion later on, for example during D/A conversion or conversion to MP3. There are only two small drawbacks to oversampling: it increases CPU usage, and it can introduce a very slight pre-ring due to the phase-linear filtering that is needed. Generally this effect is so small that it’s inaudible, but it’s good to be aware of this and not blindly assume that oversampling is always better.
My opinion is I can’t hear the difference, I just turn it on if I have the CPU for it
In my case it’s a combination of what all of you said. Yes, I generally use a clipper in every song in the master track for some slightly soft clipping to “get the song louder” without getting a harsher sound, and I fully agree with you that clippers only work on busses or master. Mostly I don’t activate oversampling while clipping because like @Garf I usually don’t hear any significant difference. But I’m thinking about using it by default, especially when I see interesting videos like the one @toimp has posted. It shows what @OopsIFly has already explained quite well, so even if you can’t hear a difference there might be one. If your CPU can handle it, why not?! And I prefer using a compressor instead of a clipper for taming transients or peaks in general within single tracks, too. Therefore I like to use the Renoise compressor, which can also be used as a limiter if necessary.
