VA oscillator differences
 

I was bored the other day so I thought I'd see what my different pure (unfiltered) VAs saw waves looked like in detail.

I used Logic to capture each "INIT" single saw sound at C2 (about 2 seconds in from note-on*) extracted a single wavecycle sample, normalised each wave and pasted each cycle next to the other.

[* note-on cycle is sometimes unpredictable]

Here they all are next to each other.


I was quite surprised by the differences and maybe this explains why different VAs have their own character (i.e you can quite often identify a Virus sound from, say, a Novation sound). It also maybe explains why the Novation analogue modelling sounds a little "rounder" (more polite) than say the Ion which is quite gritty and biting.

Next, I superimposed a mathematically perfect saw wave in yellow to see how far the modelled oscillator differs from it. I assume the convexity and concavity deviations (from mathematically perfect) are intentional to simulate the behaviour of analogue oscillators ?


It's also quite interesting (to me, anyway 8O ) to see that, apart from the MS2000, the up-cycle is not at the same magnitude as the down-cycle. I wonder whether this is intentional or not ?

Has anyone got a real analogue (preferably vintage Moog or Jupiter - rather than the modern analogues) to see how a pure saw wave cycle compares ? I'd be interested in seeing whether concavity or convexity does occur in the real thing or whether it's a "perfect" saw.

r3dnaX

nice topic, i love to get into the details of the sound! but the pics are not visible m8!

Yup. I can't see them either.

I have an analog saw and pulse oscillator somewhere, but I might have disassembled it. I'll check.

ignore (Juho has kindly sorted this out now)

I replaced the links to the pictures on my web space.

Very interesting graphs! I didn't thought that the saws differ that much. On the analog oscillator I had the saw was also quite distorted on the slope (like on Supernova).

Thanks Juho - you're a star !

Yeah, the comparisons sure surprised me. Obviously they're all going to look fairly saw shaped, but I think the differences are quite telling.

For example, it's interesting the way the Virus saw has that pronounced spike before it turns back up which is completely unlike the Ion's. God knows what's going on with the MS2000 which bows out and in and yet I think the MS2000 does sound very early analogue - so maybe it's deliberate ?

Can you do a pic of your analoge saw ? - I'd love to see how it compares

r3dnaX

Everyone's an expert now!

Edit: When I first saw (ha!) the pictures, I thought it was a discussion of fourier theory, no such luck :(
________
PREGNANT VIDEOS

whats the fourier theory ?

Sums (integral sums if I remember right) of sines and cosines. Every continuous function (wave in this case) can be constructed from sine and cosine components. Named after Jean Baptiste Joseph Fourier, one of the biggest names in mathematics of 18th and 19th centuries.

I also think it's strange that the up cycle doesn't match the down cycle in amplitude. Do you think that this is the type of DC offset that should be corrected by software; or is the general consensus that this, too, is intentional???