KVR Audio

aciddose · Post by **aciddose** » Thu Nov 27, 2014 6:46 pm

Lotuzia wrote:For those who are interested in figures just take a look at the middle of this page, and you'll see -litterally- that oscillators are not all similar -by far- in analog synthesizers

So in order to demonstrate that a set of things are not alike, you're showing me a set of things that are exactly alike with small differences in frequency content?

Do you know what a slew limiter is? You can have a slew limiter built from an integrator or lossy integrator or any sort of filter.

Apply a slew limiter with up/down slews at different rates to a pure square and you get the results in the images you've linked. They're all identical.

So the bozos who did this possibly used a crummy approximation for filter coefficients #1 combined with the fact different analogs will have different slopes depending upon temperature, age and pure component variation #2.

Try using a different sample rate
Simply randomly adjust the coefficients

The downward slew here is about half the rate of the xils 3 slew.

The slew here is at least twice the rate, plus the roll-off is much sharper than xils 3. The image shows what appears to be the effect of the audio input's anti-aliasing filter while the natural slew is most likely far higher.

memyselfandus · Post by **memyselfandus** » Thu Nov 27, 2014 6:56 pm

beely wrote:
zerocrossing wrote:I just read an article about how far a Moog Voyager's sawtooth differs from a textbook "ideal" sawtooth. I was wondering if any developers are going through the process of trying to replicate this phenomena. The topic was about how a M-Audio Venom uses sampled waveforms to achieve this.
Oscillators, filters and envelopes are core components of a synthesizer. When a developer models a synth, then of course they analyse very carefully these fundamentals to recreate them as accurately as makes sense.

No analog synth has textbook ideal waveforms - many are not even close. It's part of what gives them their character...

I'm a newb to how this works. What is used to get perfect waveforms? Software?

aciddose · Post by **aciddose** » Thu Nov 27, 2014 6:57 pm

Actually how is the pre-ringing explained in the ramp waveform?

I assumed immediately this was a linear-phase anti-aliasing filter but to be honest this seems extremely unlikely to me. There is absolutely no way that ringing is generated by the analog circuit, it is physically impossible.

I would guess this is due to some sort of weird processing applied to the signal, so this is not to be considered a raw sampling of the output of the analog.

No never mind, it must just be a really bad anti-aliasing filter. They normally have less than a single cycle of ringing while you can see about five cycles in the image.

Lotuzia · Post by **Lotuzia** » Thu Nov 27, 2014 7:45 pm

Mr Bean wrote:
SoundonSound wrote: [...] Unfortunately, there's a significant error in XILS 3's oscillator 1, whose Shape control is linked to the sawtooth wave, not the sine wave. This is a shocking mistake!

Unfortunately, this has been fixed for ages on the Xils 3 V2 and Xils 4

ghettosynth · Post by **ghettosynth** » Thu Nov 27, 2014 7:47 pm

memyselfandus wrote:
beely wrote:
zerocrossing wrote:I just read an article about how far a Moog Voyager's sawtooth differs from a textbook "ideal" sawtooth. I was wondering if any developers are going through the process of trying to replicate this phenomena. The topic was about how a M-Audio Venom uses sampled waveforms to achieve this.
Oscillators, filters and envelopes are core components of a synthesizer. When a developer models a synth, then of course they analyse very carefully these fundamentals to recreate them as accurately as makes sense.

No analog synth has textbook ideal waveforms - many are not even close. It's part of what gives them their character...
I'm a newb to how this works. What is used to get perfect waveforms? Software?

A ruler and a very sharp pencil, very sharp.

Lotuzia · Post by **Lotuzia** » Thu Nov 27, 2014 7:51 pm

Kriminal wrote:
Lotuzia wrote:spam
Shame your company cant spell a simple four letter word like *ramp*

For your information, my company is called Le Lotus Bleu.( like in my sig ) And I dont have anymore free time to represent other companies on forums.( was doing it pro bono )

Try again

hakey · Post by **hakey** » Thu Nov 27, 2014 7:52 pm

Monsieur Spamalot wrote:Unfortunately, this has been fixed for ages

Unfortunately, this hasn't:

SoundonSound wrote:If you're hoping to recreate your favourite VCS3 sounds on XILS 3, forget it. You can't.

AusDisciple · Post by **AusDisciple** » Thu Nov 27, 2014 7:54 pm

ghettosynth wrote:A ruler and a very sharp pencil, very sharp.

lol

Actually, for perfect waveforms, the pencil would need a sub-atomically sharp tip.

AusDisciple · Post by **AusDisciple** » Thu Nov 27, 2014 7:55 pm

hakey wrote:
Lotuzia wrote:Unfortunately, this has been fixed for ages
Unfortunately, this hasn't:

SoundonSound wrote:If you're hoping to recreate your favourite VCS3 sounds on XILS 3, forget it. You can't.

In all fairness, that article concluded with:

I started out by comparing XILS 3 to the VCS3 because the chaps at XILS Lab make it clear that this is what they want potential customers to do. But by the end of the review I realised that it's not a VCS3 soft synth at all... it's much more than that. Sure, it's based on EMS's synths, and the layout and patching mechanism force you to think about sounds in a 'Synthi' sort of way. But XILS 3 is also stable and playable, and the range of facilities and the range of sounds that you can obtain from it go way beyond anything that we thought was possible in 1969. If you approach it as you would a VCS3 you'll merely scratch the surface of what it can do. Dig deeper, and you'll discover a fabulously powerful new synthesizer.

BlackWinny · Post by **BlackWinny** » Thu Nov 27, 2014 7:56 pm

hakey wrote:
Monsieur Spamalot wrote:Unfortunately, this has been fixed for ages
Unfortunately, this hasn't

Unfortunately it has.

hakey wrote:
SoundonSound wrote:If you're hoping to recreate your favourite VCS3 sounds on XILS 3, forget it. You can't.

This article dates of August 2010! Have you followed the evolution of Xils3 since that date?

For your information, today is November 2014.

Sendy · Post by **Sendy** » Fri Nov 28, 2014 12:11 am

A truely "mathematically perfect" square or saw wave would need to have infinite harmonics, and hence would probably tear spacetime apart. So it's just as well nobody's been able to create one so far

While we're on the subject of spelling mistakes on that page, the plural of "matrix" is "matrices". I realize English isn't their first language, but correcting spelling errors helps maintain an air of professionalism.

Tricky-Loops · Post by **Tricky-Loops** » Fri Nov 28, 2014 12:16 am

Sendy wrote:A truely "mathematically perfect" square or saw wave would need to have infinite harmonics, and hence would probably tear spacetime apart. So it's just as well nobody's been able to create one so far

Nobody's TRULY perfect!

aciddose · Post by **aciddose** » Fri Nov 28, 2014 12:19 am

Sendy wrote:... and hence would probably tear spacetime apart.

Well, that, or...

It will create a big bang leading to a whole new universe of sonic possibility in your DAW.

Buy BigBangAudio's all new Infiniharmo™ today!

aciddose · Post by **aciddose** » Fri Nov 28, 2014 12:28 am

BTW, the technical of this is actually that all signals have infinite harmonics and it is no problem at all.

The place where it requires infinite energy is if we wanted to eliminate the latency when moving from one position to another. In order to do this we'd need to expend an infinite amount of energy to at best (to our knowledge) accelerate to the speed of light in an instant.

(Then of course decelerate in an instant once having arrived, using again an infinite amount of energy. This came up when there was discussion about those "bad ripples" which are the difference between digital and analog, clearly.)

Exactly how quanta and relativity all mix together are still a little hazy from what I understand even for advanced physicists. Now, what I understand on the subject is pretty much; forget about it.

We're not absolutely certain what occurs when we apply infinite energy in an instant as we typically aren't playing with that sort of thing in our spare time. It is safe to assume it might make the universe a little bit upset.

mystran · Post by **mystran** » Fri Nov 28, 2014 1:15 am

So after 4 pages, this thread has come to the conclusion(?) that a typical VCO generates near-perfect saw and pulse waves (which could just as well be perfect as far as audio frequencies are concerned, but let's not worry about that) and depending on where we draw the line with "this is oscillator" and "this is after oscillator" there might be a bunch of other circuitry that might or might not result in various linear and/or non-linear distortion to the signal (and yes, "linear distortion" is the correct engineering term here).

What I haven't quite understood so far is how to start modelling the circuitry that might or might not be considered part of the VCO without first generating the near-perfect waveform (which we apparently should not bother generating) that is produced by the part that is part of the typical VCO with high probability (independent of where we draw the line above), nor am I sure how Xils marketing materials are going to help with any of that.

This is getting far too complicated, I think I have to start thinking about an alternative career in quantum physics or something instead.

Analog modeled waveforms. Who's doing it?