What does Waves Vitamin *actually* do? It ain't no saturator...

[Zexila]

Part eq, part multiband compressor, part saturator, part multiband stereo widener.

I to put it on a send, turn down the direct and send anything into it I feel can do with a bit of lift. I’ve also run it into an SPL Vitalizer which can also work quite well.

Part none of the above

It sounds like it uses the same method that the Aphex Aural Exciter(there is also no saturation or transfer curve involved in the Aphex things either) uses but in a multiband arrangement.the way the Aural Exciter works is by synthesizing frequencies above (or below)the harmonics you have present in your sounds using a "heterodyne"

https://en.wikipedia.org/wiki/Heterodyne.

Think of it like frequency shifting,but it's not frequency shifting.i might have my wires crossed though but it is either this or a Hilbert Transform which mixes in phase inverted copies of the same thing at 90 degrees offset.

It doesn't saturate at all though there are no transfer curves or chebyshev polynomials (like in Cobalt Saphira for example)in Vitamin

I personally don't use it because i find the crossovers to be a bit much (they ruin the sound IMO)would be nice if they at least had different options for the crossovers like FIR mode that isn't quite linear,and isn't quite minimal (like the hybrid mode in Izotope's stuff)

As for the dynamics section,i don't think or i am not sure that it's as sophisticated as a compressor. it has more in common with a transient designer in its behaviour,so it is likely just a very simple envelope follower of some sort

In case you was interested the SPL Vitalizer also works by doing something funky with phase shifting using the group delay maybe from Allpass filtering to shift overlapping frequencies into the time domain.it doesn't specify in the literature how it accomplishes that but there are not many ways you can accomplish that anyway so it is likely an Allpass filter or phase rotating circuit(which still would involve Allpass filters).no saturation in that one either.

Great input, cheers.

[imrae]

Shit, why you using waves plugins?

This is a fair question! I don't like this kind of black-box stuff, or DRM, or "noise" buttons.

There was a thread recently asking for an alternative to Doubler. Most of the suggestions were actually chorus effects that can't do the classic static pitch-shift effect. The others were Microshift and CMX, which actually cost more and have less controls.

Generally speaking their old stuff seems to be well-designed with good workflows, clear sound and predictable, well-documented behaviour. Gold is good value for Doubler, RenEQ, C1, Metaflanger, H-delay... MaxxBass/RenBass is another "secret sauce" plugin that people find irreplaceable. I have DC8C which is a *great* flexible compressor, but C1 can do some things it can't. I'm less sold on the modelling stuff but that's a topic for another thread.

Point is: I got it in a bundle, so did lots of other people, I have it, people like it, therefore should probably investigate. If it turns out to be poo, I'll increase my suspicion of modern Waves plugs. If not, huzzah there's another useful tool

[imrae]

TIMT thank you that is fascinating. There was no mention of techniques other than distortion in the other Exciter thread (same one Mathematics linked). The ThrillseekerXTC uses distortion and I have no experience with the Aphex which is often explained as "uses a tube to add harmonics" so it is a breakthrough to recall that there are other options than distortion!

I will look more into how the Aphex exciter works. But in the meantime... wouldn't a "pitch-shift" type exciter also add a noticeable peak or two above the fundamental? Why don't I see that in my spectrum analyser?

[do_androids_dream]

The difference between exciter and saturator was discussed recently:
viewtopic.php?t=383884

No need for a discussion.. there is no difference - just a marketing distinction.

On topic.. off to play with Vitamin and give my take on what it's actually doing. One thing to say about it is that it's a shame it's not linear phase.

[stromkraft]

A quick analysis of a Saturator in Live, which adds fat bands of harmonics, and Waves Aural Exciter, which doesn't do the same at all, renders this statement as useful as "there's no difference between a high-pass and low-pass filter, just marketing distinctions".

A linear phase Vitamin probably would add some hefty latency. Maybe as an option?

[do_androids_dream]

A quick analysis of a Saturator in Live, which adds fat bands of harmonics, and Waves Aural Exciter, which doesn't do the same at all, renders this statement as useful as "there's no difference between a high-pass and low-pass filter, just marketing distinctions".

Most devs use 'exciter' to describe some kind of distortion process which can be described as saturation too. Aural Exciter is an exception. For the most part, it's just a marketing distinction or, worst case, buzzword.

A linear phase Vitamin probably would add some hefty latency. Maybe as an option?

It would but it would mean it's suitable for mastering then which would be nice. A minimum and linear phase one would be good to have.

[imrae]

If the "wet" side of this thing literally works by screwing with phase (which seems like a possibility at this stage) then I don't imagine that the kind of crossover filtering between bands matters that much. The dry signal doesn't go through any crossovers as far as I can tell.

[simon.a.billington]

Part eq, part multiband compressor, part saturator, part multiband stereo widener.

I to put it on a send, turn down the direct and send anything into it I feel can do with a bit of lift. I’ve also run it into an SPL Vitalizer which can also work quite well.

Part none of the above

It sounds like it uses the same method that the Aphex Aural Exciter(there is also no saturation or transfer curve involved in the Aphex things either) uses but in a multiband arrangement.the way the Aural Exciter works is by synthesizing frequencies above (or below)the harmonics you have present in your sounds using a "heterodyne"

https://en.wikipedia.org/wiki/Heterodyne.

Think of it like frequency shifting,but it's not frequency shifting.i might have my wires crossed though but it is either this or a Hilbert Transform which mixes in phase inverted copies of the same thing at 90 degrees offset.

It doesn't saturate at all though there are no transfer curves or chebyshev polynomials (like in Cobalt Saphira for example)in Vitamin

I personally don't use it because i find the crossovers to be a bit much (they ruin the sound IMO)would be nice if they at least had different options for the crossovers like FIR mode that isn't quite linear,and isn't quite minimal (like the hybrid mode in Izotope's stuff)

As for the dynamics section,i don't think or i am not sure that it's as sophisticated as a compressor. it has more in common with a transient designer in its behaviour,so it is likely just a very simple envelope follower of some sort

In case you was interested the SPL Vitalizer also works by doing something funky with phase shifting using the group delay maybe from Allpass filtering to shift overlapping frequencies into the time domain.it doesn't specify in the literature how it accomplishes that but there are not many ways you can accomplish that anyway so it is likely an Allpass filter or phase rotating circuit(which still would involve Allpass filters).no saturation in that one either.

It’s how Waves themselves define it...

https://youtu.be/VRWWPPnGGOQ

There is always some frequency build up that can do with a bit of notching out with eq. This is usually where I park the crossover of non-linear multiband processors. That way I use the phase shift in my favour. I do the same with LP filters as well.

The original SPL units work with several tubes I believe. And the idea of how it harmonically enhances the frequency is by implementing the Fletcher-Munson curve in its harmonic shaping. At least that’s the goal, what happens in real life with real components is often not as precise as that.

[imrae]

It’s how Waves themselves define it...

It wouldn't be the first time a developer was economical with the truth. Linear transfer function on a static signal and no added harmonics on a static signal (test it if you don't believe me) means that this is NOT the same as parallel compression and/or waveshaping saturation. Some interesting ingredient is missing. It may be an exotic saturation technique, or it might be an ingenious phase/antiphase combination of those things.

[simon.a.billington]

It’s how Waves themselves define it...

It wouldn't be the first time a developer was economical with the truth. Linear transfer function on a static signal and no added harmonics on a static signal (test it if you don't believe me) means that this is NOT the same as parallel compression and/or waveshaping saturation. Some interesting ingredient is missing. It may be an exotic saturation technique, or it might be an ingenious phase/antiphase combination of those things.

All possible.

What’s most important is the bottom line. Does it do the job, does it work as advertised.

I believe it does on a personal level. It sounds to me that it kinda eqs, kinda compresses, kinda saturates and I can get good results with it. I’m sure many others do too.

Additionally, it’s not beyond Waves capability to create new approaches to do the same type of job. Their FIR reverb algorithm for example, or something a bit more unique like the new tech used in Torque and Smack Attack.

https://www.waves.com/sound-synthesis-r ... esynthesis

So maybe they’ve gone outside the box on Vitimin. It’s possible. Would they be disingenuous? Possible, but highly unlikely. They’ve never been dishonest so far, it seems really silly to be dishonest just with Vitamin.

They have too many high profile clients that would just call them out long ago and that would end up being very bad for Waves. It just makes really bad business sense. You don’t make 25 years by making bad business decisions.

It hasn’t happened though. Something worth thinking about.

[stromkraft]

Most devs use 'exciter' to describe some kind of distortion process which can be described as saturation too. Aural Exciter is an exception.

Waves Aural Exciter is modelled on the original exciter, more or less, so I think you're simply over-simplifying this detail.

Saturated and Excited signals don't even sound the same, so you pretending these are the exact same phenomenon is just silly. There may be some overlap obviously in various products, but these are different processes and more importantly these bring different results.

What Vitamin does I don't know, but it's an interesting product that I don't expect to sound like Saphira, that I own, even if some aspects in the processing may be similar. It's all in the implementation.

[do_androids_dream]

Most devs use 'exciter' to describe some kind of distortion process which can be described as saturation too. Aural Exciter is an exception.

Waves Aural Exciter is modelled on the original exciter, more or less, so I think you're simply over-simplifying this detail.

Saturated and Excited signals don't even sound the same, so you pretending these are the exact same phenomenon is just silly. There may be some overlap obviously in various products, but these are different processes and more importantly these bring different results.

What Vitamin does I don't know, but it's an interesting product that I don't expect to sound like Saphira, that I own, even if some aspects in the processing may be similar. It's all in the implementation.

Is there a definitive definition of 'exciter' one wonders? Googling around I've read everything from folks saying "No! An exciter isn't distortion! It's saturation" .... when saturation is just another word for distortion. Others say an exciter 'adds overtones'. How do we get 'overtones'? By distorting/saturating the signal... I don't know who wrote the wikipedia entry on 'exciter' but it's just as much a stab in the dark as any other 'definition' I've read. So, I come back to what I originally said. In the main (not exhaustive) it is *usually* just a marketing distinction of some kind of distortion process.

As regards Vitamin.. well, it certainly isn't doing any form of saturation that I can detect. It seems that it is simply an eq with some very gentle compression introduced and a mid/side balance per band. The punch control is pretty horrid to be honest - usually just mushes up a transient (I'm guessing it's just soft clipping transients a bit with some subtle eq boosting as well.)

[simon.a.billington]

I do know that guitar folks have a definition that describes the differences between distortion, overdrive and fuzz, perhaps this extends to saturation as well.

I would agree, you would saturate a signal by applying some style of distortion as outlined above, it’s perhaps people who get caught up on semantics that have issues. With calling saturation distortion

As far as I’m aware, according to guitarists “distortion” is achieved by clipping components, overdrive is achieved by driving a signal too hard into tube preamps which tend not to clip, fuzz I can’t remember. Oh and there’s also rectifying which inverts the negative phase of an audio signal.

Saturation I believe is from driving a signal too hot to tape, but I’m sure there’s other ways to achieve it, like driving the preamps harder. Given that, I’d go out on a limb that it’s a form of overdrive.

If I were to hazard another guess is that exciting would possibly mean a more efficient but artificial method of achieving saturation. By artificial I mean an effect designed specifically to achieve saturation, not some byproduct you get from driving hot signals into equipment.

But then, that is just an educated guess.

[aciddose]

That distinction is based upon a lack of knowledge of the underlying math.

Every example involves a non-linear equation (a polynomial) with different degrees/coefficients.

"Hard clipping" for example is only an extreme variation of "overdrive", both are pushing the active device into saturation. It's only the manner in which you get there that determines whether the result is "clipping" or "overdrive".

For example:
What is the result of hard clipping from 0 to +V where the signal is centered on 0? That function is called dim(x,y), it's the difference between two values where the difference is positive, otherwise it is zero.

pos(x) = (x + abs(x)) / 2
dim(x,y) = pos(x - y)

Now with that in mind, what do you get when you add 2*dim(x, 0) + x?

2*dim(x, 0) + x = abs(x)

How do you get abs(x)?

abs(x) = sqrt(x^2)

How do we get sqrt(x)?

sqrt(x) = x ^ (1/2)

How do we get pow(x,y)? ( pow(x, y) = x ^ y )

pow(x, y) = exp(log(x) * y)

Round and round and round it goes...

We can also build clip(x, y) from dim(x, y), pos(x) or abs(x), these are all essentially the same function set up a little bit differently.

So while this may seem like a useful distinction to a guitarist, like most things guitarists believe, it is 100% bullshit.

[simon.a.billington]

Yeah so in a scentific manner we could break it down into mathematics. That usually becomes a more susinct way at defining differences.

However, it is very unfriendly for most users. They tend to think in terms of method or how it sounds. We wouldn’t even know the maths for any given plugin to be able to create such delineations anyway.

It really needs to be defined more practically to be more usable by everyone.