KVR Audio

kunn · Post by **kunn** » Wed Aug 12, 2009 7:22 am

sockofgold wrote:http://www.megaupload.com/?d=GRXL5DE1

Here you go. 24/44.1 glory. There are about as many supersaw samples in here as you could ask for. All middle C, roughly 3/4-second long. Basically, I just moved the mix fader up between each note until it reach the top, and then increased the detune fader a bit and did it again.

Naturally, the very first note has the controls at 0%/0%, and the last note has them at 100%/100%.

Hope these help.

(Oh, and these come from an initialized patch, so no filter/OSC 2/chorus or any other junk.)

Thank you!

Can you upload samples with controls 0%/0% and notes going from C1 to C5. Got to analyse the hi-pass filter.

sockofgold · Post by **sockofgold** » Wed Aug 12, 2009 12:46 pm

kunn wrote:
sockofgold wrote:http://www.megaupload.com/?d=GRXL5DE1

Here you go. 24/44.1 glory. There are about as many supersaw samples in here as you could ask for. All middle C, roughly 3/4-second long. Basically, I just moved the mix fader up between each note until it reach the top, and then increased the detune fader a bit and did it again.

Naturally, the very first note has the controls at 0%/0%, and the last note has them at 100%/100%.

Hope these help.

(Oh, and these come from an initialized patch, so no filter/OSC 2/chorus or any other junk.)
Thank you!

Can you upload samples with controls 0%/0% and notes going from C1 to C5. Got to analyse the hi-pass filter.

Sure, no problem. It will be a quite a few hours again, though--I just got to work (GMT -8 ).

sockofgold · Post by **sockofgold** » Thu Aug 13, 2009 12:40 am

kunn wrote:
sockofgold wrote:http://www.megaupload.com/?d=GRXL5DE1

Here you go. 24/44.1 glory. There are about as many supersaw samples in here as you could ask for. All middle C, roughly 3/4-second long. Basically, I just moved the mix fader up between each note until it reach the top, and then increased the detune fader a bit and did it again.

Naturally, the very first note has the controls at 0%/0%, and the last note has them at 100%/100%.

Hope these help.

(Oh, and these come from an initialized patch, so no filter/OSC 2/chorus or any other junk.)
Thank you!

Can you upload samples with controls 0%/0% and notes going from C1 to C5. Got to analyse the hi-pass filter.

Boom.

http://www.megaupload.com/?d=HS53Q47A

kunn · Post by **kunn** » Thu Aug 13, 2009 8:10 am

Thanks!

The hi-pass filter is scaled to the pitch and seems to have some resonance. There's also a global hi-pass filter with constant cutoff.

valhallasound · Post by **valhallasound** » Thu Aug 13, 2009 8:42 am

kunn wrote:Thanks!

The hi-pass filter is scaled to the pitch and seems to have some resonance. There's also a global hi-pass filter with constant cutoff.

Is the global filter set to a particularly high frequency, or could this be the DC blocking capacitors of the output? Most hardware synths have one or more coupling caps in the signal path, to prevent DC from frying speakers.

Sean Costello

valhallasound · Post by **valhallasound** » Thu Aug 13, 2009 8:46 am

kunn wrote:Seems like the tuning of the detuned saws is not even exponential.

Full detune:

1st pair: f*(1 +/- 0.02)
2nd pair: f*(1 +/- 0.06)
3rd pair: f*(1 +/- 0.11)

I wouldn't be surprised if the super accurate ratios were closer to prime numbers or the like, with more significant digits. The detuning seems designed to have different beat rates between each waveform, for added complexity. For example, if the ratios were:

1st pair: f*(1 +/- 0.02)
2nd pair: f*(1 +/- 0.06)
3rd pair: f*(1 +/- 0.10)

then the beating between the 1st and 2nd pair would happen at the same rate as between the 2nd and 3rd pair.

Sean Costello

kunn · Post by **kunn** » Thu Aug 13, 2009 12:52 pm

valhallasound wrote:
kunn wrote:Seems like the tuning of the detuned saws is not even exponential.

Full detune:

1st pair: f*(1 +/- 0.02)
2nd pair: f*(1 +/- 0.06)
3rd pair: f*(1 +/- 0.11)
I wouldn't be surprised if the super accurate ratios were closer to prime numbers or the like, with more significant digits. The detuning seems designed to have different beat rates between each waveform, for added complexity. For example, if the ratios were:

1st pair: f*(1 +/- 0.02)
2nd pair: f*(1 +/- 0.06)
3rd pair: f*(1 +/- 0.10)

then the beating between the 1st and 2nd pair would happen at the same rate as between the 2nd and 3rd pair.

Sean Costello

From FFT I got

1st pair: f*(1 +/- 0.019)
2nd pair: f*(1 +/- 0.062)
3rd pair: f*(1 +/- 0.108)

and rounded them a little.

fliffo · Post by **fliffo** » Fri Aug 14, 2009 5:03 am

kunn wrote: From FFT I got

1st pair: f*(1 +/- 0.019)
2nd pair: f*(1 +/- 0.062)
3rd pair: f*(1 +/- 0.108)

and rounded them a little.

@ what amount of detuning does correspond this values? Full?

kunn · Post by **kunn** » Mon Aug 17, 2009 9:55 am

fliffo wrote: @ what amount of detuning does correspond this values? Full?

full detune.

fliffo · Post by **fliffo** » Mon Aug 17, 2009 2:34 pm

ok, thanks.

SoulState · Post by **SoulState** » Fri Sep 25, 2009 4:39 pm

And what about Virus Ti Hypersaw?
It's using 9 detuned saws.
With which detune values?
Similar 3 pair's to Jp's? and one extra pair(what value?) Or different?

michu · Post by **michu** » Sat Sep 26, 2009 10:48 am

valhallasound wrote:
kunn wrote:Seems like the tuning of the detuned saws is not even exponential.

Full detune:

1st pair: f*(1 +/- 0.02)
2nd pair: f*(1 +/- 0.06)
3rd pair: f*(1 +/- 0.11)
I wouldn't be surprised if the super accurate ratios were closer to prime numbers or the like, with more significant digits. The detuning seems designed to have different beat rates between each waveform, for added complexity. For example, if the ratios were:

1st pair: f*(1 +/- 0.02)
2nd pair: f*(1 +/- 0.06)
3rd pair: f*(1 +/- 0.10)

then the beating between the 1st and 2nd pair would happen at the same rate as between the 2nd and 3rd pair.

Sean Costello

my understanding of this is totally rusty, but with naive saws in fixed point math phase offset would be a simple add/sub per additional saw and for detune you'd need to increment it once per cycle? the detune amount would be increment amount...

Substance 8080 · Post by **Substance 8080** » Fri Jan 08, 2010 8:42 pm

Dont know if anyone's still watching this thread but i'll ask anyway. If the full detune values are,

+/- 0.019
+/- 0.062
+/- 0.108

What would the detune values be at no detune, full mix?

valhallasound · Post by **valhallasound** » Sun Apr 17, 2011 5:13 am

Some recent research in this area:

http://www.ghostfact.com/jp-8000-supersaw/

Swiss Frank · Post by **Swiss Frank** » Fri Sep 02, 2011 11:51 am

A question about the killer article at: http://www.ghostfact.com/jp-8000-supersaw/

"This makes sense when you think about it - efficiency was king, and it's hard to beat the efficiency of two extra adds per sample - things are a lot easier when you embrace aliasing rather than reject it, huh?"

I don't get it-two extra adds to do what? Or do you mean the 7 sawtooths really are just 7 naive, non-BWL sawtooths (which I guess may only take a few additions each)?! Thats the secret?! And if so why only for non-BWL sawtooths?

Roland Supersaw - any idea how the original was done?