Isolated saturation doesn't sound good

[sambaji]

Today I tested some saturation plugins and I was surprised by how bad most sounded on my mix when they were isolated. In case there was phase shifting due to the plugin, I put the plugin to be tested on two tracks, with one of the tracks phased reversed, turned up the saturation on one of them, and adjusted the gain to isolate the saturation as much as possible. In most cases, it added not just fuzzy noise but crackling and other non-musical noise. (This is in contrast to testing with a pure sine wave, where the resulting isolated harmonics are more pleasant, coherent). Why would we want to add such artifacts liberally to our pure audio? Sure it may add some excitement with the added harmonics, but what about the unwanted artifacts? My answer would be that the unwanted artifacts are largely masked by the full audio.

While I can see using saturation on particular instruments in specific cases (e.g., bass guitar), I am a little hesitant to add saturation to every stage of the mix that some people advocate--console channel saturation on every track/group bus, tape saturation on the mix bus, etc? With blanket saturation on the mix bus, there is often something I like and don't like about it--e.g., loss of clarity, unwanted changes in the tone of cymbals, etc. If I am to add saturation, I favor Krazog's subtle True Iron, and sometimes Voxengo's free saturation EQ, Overtone, where I can control specific frequencies.

Are we a little zealous with adding saturation to our mixes?

[bmanic]

I don't think you quite understand the differential signal you get when you phase reverse something like a saturation plugin. Ever tried doing the same to a compressor?

That "crackling" is of course not representative of what's actually going on. It is the differential signal, the stuff that's being processed, probably mainly transients/peaks. As they are super short it'll show itself as crackling. That doesn't mean you get actual crackling in the signal itself.

Not sure why you are expecting pure sine waves for distortion. That's not how it works.

Even the most "gentle" of saturation will manifest itself as "crackling" if you perfectly phase reverse the rest of the signal.


As for tonality changes on cymbals. If they get "nasty" and you feel like your audio is getting a lot of extra harsh high-mid energy, what you are hearing is most likely aliasing. It can be especially troublesome on cymbals and hihats (and vocal esses).

A cool little trick you can do to minimize the bad artifacts, even if your saturation plugin doesn't support oversampling, is to first pre-EQ the signal, before the distortion/saturation, then post-EQ it in the exact opposite direction an equal amount. For instance using a high-shelf EQ to remove 12dB from the highs.. then put it back after wards in equal amounts. This can help change the amount and pattern of aliasing (though very much depending on the saturation shape itself of course) and thus minimize harshness.

As for "details", in my opinion it's the exact opposite. Good saturation actually helps with details as the signal gets stronger representation and separation of the fundamental frequencies. Thus saturation is a great way to get things to sit in a mix and fill it up. It also helps tremendously on instruments that you want to heavily EQ/filter and position back in the mix. By saturating the sources you can be more heavy handed with the EQ and still keep them perceivable within the mix.

[sambaji]

I don't think you quite understand the differential signal you get when you phase reverse something like a saturation plugin. Ever tried doing the same to a compressor?

That "crackling" is of course not representative of what's actually going on. It is the differential signal, the stuff that's being processed, probably mainly transients/peaks. As they are super short it'll show itself as crackling. That doesn't mean you get actual crackling in the signal itself.

Not sure why you are expecting pure sine waves for distortion. That's not how it works.

Even the most "gentle" of saturation will manifest itself as "crackling" if you perfectly phase reverse the rest of the signal.


As for tonality changes on cymbals. If they get "nasty" and you feel like your audio is getting a lot of extra harsh high-mid energy, what you are hearing is most likely aliasing. It can be especially troublesome on cymbals and hihats (and vocal esses).

A cool little trick you can do to minimize the bad artifacts, even if your saturation plugin doesn't support oversampling, is to first pre-EQ the signal, before the distortion/saturation, then post-EQ it in the exact opposite direction an equal amount. For instance using a high-shelf EQ to remove 12dB from the highs.. then put it back after wards in equal amounts. This can help change the amount and pattern of aliasing (though very much depending on the saturation shape itself of course) and thus minimize harshness.

As for "details", in my opinion it's the exact opposite. Good saturation actually helps with details as the signal gets stronger representation and separation of the fundamental frequencies. Thus saturation is a great way to get things to sit in a mix and fill it up. It also helps tremendously on instruments that you want to heavily EQ/filter and position back in the mix. By saturating the sources you can be more heavy handed with the EQ and still keep them perceivable within the mix.

Thanks for the helpful info! I wasn't expecting the tested plugin to produce pure waves as distortion from my mix. I comparing the lack of artifacts from pure sine wave distortions versus a complex signal. But as you mentioned, the "artifacts" I am hearing such as crackling are just quick transients peaking through. How does saturation create stronger representation and separation of the fundamental frequencies by adding additional non-fundamental frequencies?

[Ploki]

its hard to phase-flip non-linear transfer functions - most saturators are wave-shapers, they affect dynamics - so it affects waveform and flipping phase will also show all those non-linearities, which sound artefacty and crackly and weird.

Harmonics are part of every sound (except pure sine waves), so playing with the harmonic content changes timbre. By accentuating upper harmonics you effectively make the sound "brighter" and "more present" so it creates a stronger representation of it in the mix.

A cool little trick you can do to minimize the bad artifacts, even if your saturation plugin doesn't support oversampling, is to first pre-EQ the signal, before the distortion/saturation, then post-EQ it in the exact opposite direction an equal amount.

aah bias for saturation. Why that isn't standard on saturators is beyond me, so much more unnecessary work.

[Tj Shredder]

A cool little trick you can do to minimize the bad artifacts, even if your saturation plugin doesn't support oversampling, is to first pre-EQ the signal, before the distortion/saturation, then post-EQ it in the exact opposite direction an equal amount. For instance using a high-shelf EQ to remove 12dB from the highs.. then put it back after wards in equal amounts. This can help change the amount and pattern of aliasing (though very much depending on the saturation shape itself of course) and thus minimize harshness.

That does not work, as aliasing would still occur and you would boost the aliasing as well. It does help if you don’t boost the highs after the distortion though, as the distortion will add highs on its own. The principle of the Aural Exciter...
What would help as well is to split the sound into two bands and only distort the low band...

[Ploki]

That does not work, as aliasing would still occur and you would boost the aliasing as well. It does help if you don’t boost the highs after the distortion though, as the distortion will add highs on its own. The principle of the Aural Exciter...

will it does work because the high frequencies will be pushed into the saturator LESS so they will have LESS harmonics and LESS foldback due to it.

[TIMT]

A cool little trick you can do to minimize the bad artifacts, even if your saturation plugin doesn't support oversampling, is to first pre-EQ the signal, before the distortion/saturation, then post-EQ it in the exact opposite direction an equal amount. For instance using a high-shelf EQ to remove 12dB from the highs.. then put it back after wards in equal amounts. This can help change the amount and pattern of aliasing (though very much depending on the saturation shape itself of course) and thus minimize harshness.

That does not work, as aliasing would still occur and you would boost the aliasing as well. It does help if you don’t boost the highs after the distortion though, as the distortion will add highs on its own. The principle of the Aural Exciter...
What would help as well is to split the sound into two bands and only distort the low band...

How does this makes sense to you....ofcourse attenuating the amount of high end that gets fed into a non-linear processor pre that non-linear processor would reduce the amount of foldback at Nyquist with a couple of exceptions e.g anything that folds the transfer curve to produce wrapping of the phase, abruptly asymmetric transfer curves

Splitting presents it's own caveats for instance like the disconnect you can get in the signal depending on the order and topology of the crossover, and the amount of and type of processing you do in each band. Fair enough, it reduces intermodulation, but distorting/saturating things without the presence of any intermodulation from lower frequencies can get nasty and present a mixing issue fast. Well, at least its usually the case ITB.... Sounds incredibly unnatural to me. Digital IM can be unpleasant though in most scenarios possibly due to a lack of recursion(overly simplistic result) in the thing producing the non-linearity e.g basic transfer functions/waveshapers and that

[ramseysounds]

Its what it sounds like in the mix that matters, not when it's isolated.

[zerocrossing]

It’s funny, when I was first learning the ropes of audio engineering, there was sort of a push to make sure everything was recorded as cleanly as possible with as much isolation as possible. Any saturation was specifically added as an effect. Pushing things into distortion was frowned upon. Now we love it, but I agree that in some cases it’s used to excess. It can sound great on a bass drum, but too harsh on a drum bus, especially if there’s a lot of cymbals or even a bright snare. As usual, use your ears.

[jancivil]

"surprised by how bad most sounded on my mix when they were isolated"
happens all the time, because it's mixing
CF: "Its what it sounds like in the mix that matters, not when it's isolated."

[ramseysounds]

What does CF stand for?

[jancivil]

confer

[sambaji]

Harmonics are part of every sound (except pure sine waves), so playing with the harmonic content changes timbre. By accentuating upper harmonics you effectively make the sound "brighter" and "more present"

A great free plugin to explore how even and odd harmonics and different limiting shapes affect the timbre is Melda's MSaturator, which has anti-aliasing. It has different limiting shapes and an "analog" slider that adjusts the ratio of even harmonics from these shapes, a harmonics panel where 2,3,4,5 order harmonics can be individually adjusted, and a compression panel that adjust compression on the saturation itself. Considering how flexible MSaturator is and that it is free, I am curious why it's not more promoted in the online community (e.g., for example, "The Best Free FX Plugins 2021" lists).

In my current mix, I am experimenting with routing all the instruments besides the cymbals to one group bus to subtly saturate using MSaturator (single or double in series at very low settings) first before it reaches the mix bus. Let's see if I can find some degree and type of saturation that sounds good to my ears with a largely acoustic mix.

[Ploki]

Harmonics are part of every sound (except pure sine waves), so playing with the harmonic content changes timbre. By accentuating upper harmonics you effectively make the sound "brighter" and "more present"

A great free plugin to explore how even and odd harmonics and different limiting shapes affect the timbre is Melda's MSaturator, which has anti-aliasing. It has different limiting shapes and an "analog" slider that adjusts the ratio of even harmonics from these shapes, a harmonics panel where 2,3,4,5 order harmonics can be individually adjusted, and a compression panel that adjust compression on the saturation itself. Considering how flexible MSaturator is and that it is free, I am curious why it's not more promoted in the online community (e.g., for example, "The Best Free Mixing Plugins 2021" lists).

In my current mix, I am experimenting with routing all the instruments besides the cymbals to one group bus to subtly saturate using MSaturator (single or double in series at very low settings) first before the mix bus. Let's see if I can find some degree and type of saturation that sounds good to my ears with a largely acoustic mix.

you either love or hate melda.
I hated it, now i love it.
I bought the free bundle (among others) and i have to admit i never even tried the MSaturator

[plexuss]

Saturation is distortion. I get that analogue gear imparts distortions and those that work ITB want those analogue sounds. But just adding distortion with no strategy or specific goal is just looking for a crappy sounding mix.

I've tried many of the distortion tools out there and yes they do affect the sound and generally in a crappy way. Its distortion!

Now I no longer call these tools and techniques "saturation" - I call them what they are: distortion. this helps keep me focused from doing damage to my audio with these tools.

Distortion appears to be psychoacoustically similar to loudness in that the as you add it is appears to sound good but on more objective listens it actually sounds bad - loss of dynamic range, clarity, transients and over-all bloated weight to the whole thing.