Page 2 of 2

Re: MyCo's Oversampling Toolkit

Posted: Wed Apr 20, 2022 1:10 pm
by Tepeix
For the particular square form, here's a way to simulate oversampling.
(when the polarity change, it calculate where between two sample it must occur and blend accordingly
with the last sample)

But whatever i try, changing the sampling rate or using a simple prim sin.
I could never have note at 10000 hertz. The volume drop or we could only hear aliasing when it occurs.
Is it the limitation of computer ? Or maybe my particular setup ?

Re: MyCo's Oversampling Toolkit

Posted: Wed Apr 20, 2022 2:50 pm
by tulamide
Spogg wrote:
tulamide wrote:... I double, quadruple, or whatever the rate increase is, each sample...


Once I’d kinda worked out what was going on I started to wonder if you could generate 4 non-identical values per sample period. But then my brain hung up.
:lol:

I'm aware, that this is totally offtopic, but technically you can, if you don't care for the type of value (wether NaN or NiL). Every number is just a set of bits. How they are interpreted makes them a number. You can split the 32 bits of a sample point into 4 groups of 8 bits and fill up each one with 24 zeros. However, as mentioned, they lose their quality towards the original number, and when doing calculations and later put the 32 bits back together, the new number will be drastically different from the original. I wouldn't know any use case (apart from heavy distortion near noise).

Re: MyCo's Oversampling Toolkit

Posted: Wed Apr 20, 2022 8:41 pm
by martinvicanek
martinvicanek wrote:You can cascade the 2x oversampling units to achieve 4x, 8x, etc.
On second thought, it does not seem to be as straight-forward as I thought. In the meantime, I updated the tutorial a little.

Re: MyCo's Oversampling Toolkit

Posted: Thu Apr 21, 2022 1:06 pm
by Tepeix
8-) Was believing i get something good whit my square oversample simulator..
But i make a false comparison with your square osc Martin, confounding the naive and bandlimited out ;)
Now i could say that in this case your Polybleep method is better and faster !)

Re: MyCo's and MV's Oversampling Toolkit

Posted: Sun Apr 24, 2022 10:06 am
by martinvicanek
Hi folks,

I have tweaked the oversampling filters a little to make them more efficient and more accurate for the higher orders. Also included 4x oversampling. Added to the first post.

Also updated the demo & tutorial (third post).

Re: MyCo's and MV's Oversampling Toolkit

Posted: Tue Apr 26, 2022 3:23 pm
by Tepeix
Thanks very cool !

I tested it a little but for now i could not make a better overdrive with it.
(not surprise because my overdrive prototype doesn't sound greater at higher sample rate..)

Re: MyCo's and MV's Oversampling Toolkit

Posted: Wed Apr 27, 2022 11:42 pm
by StereoSpace
Hi guys!
I found an excellent compressor circuit on the forum, started experimenting with oversampling modules, but I don’t understand at all how to calculate the delay automatically in order to pass the oversampling delay value to the delay compensation module. The Bertom - EQ Curve Analyzer shows phase shifts and if you set the values to 3 samples, then everything is aligned. And sorry for the mistakes in the text, I use google translator))