u-he

snigelx · Post by **snigelx** » Fri Jan 15, 2010 7:14 pm

Hi Urs (or Anyone who might be able to clarify this),

In the Zebra2 manual it is mentioned that the cutoff parameter value in Zebra's VCF is the midi note number/semitone of the cutoff frequency. You gave an example that if the cutoff knob is set to 81.00 then the cutoff freq will be 440Hz. The lowest value possible is 1 and the highest is 100. When I look at a chart it says that midi note number 81 will give midi note frequency 880Hz. Can you please explain why your conversion is different from this or clarify the entry in the manual? I am trying to relate the cutoff setting number in Z2 to a specific frequency so I would like to get a better understanding of what you mean when you describe this in the manual. Thank you

P.S. I note that ACE's cutoff parameter values have a range of 1-150. That is 12.5 octaves so that means that each number is a semitone. What about Zebra?

xh3rv · Post by **xh3rv** » Fri Jan 15, 2010 9:10 pm

snigelx wrote: P.S. I note that ACE's cutoff parameter values have a range of 1-150. That is 12.5 octaves so that means that each number is a semitone. What about Zebra?

Not a really good answer, but there's a great patch in the factory content 'Tuvatron C' in the category 'Noises' that shows off a self-oscillating filter in Zebra very well, this might be informative. If you coerce the patch into Zebra v2.5 mode, the resonance will be sharper. A spectrograph plug that shows notes (not sure what's out there, Live's internal plug does) would also be useful here.

Basically Zebra's filters are 1.00 in knob movement -> one semitone.

snigelx · Post by **snigelx** » Fri Jan 15, 2010 10:34 pm

Thanks xh3rv I really appreciate your answer! Unfortunately it is not exactly what I am looking for. I know that the Z2 manual states that each value number for the cutoff is equal to 1.00 semitone or midi note #... it also states that 81 is 440hz. I am saying this is not in accordance to most charts I find that chart midi note# and frequency. These charts say midi note# 81 is actually 880Hz. Do you know why this is? Also, if you look in ACE one will notice that the value range for cutoff is 1-150 versus Z2 1-100. Perhaps Urs can shed some light on this at his earliest convenience.

xh3rv · Post by **xh3rv** » Fri Jan 15, 2010 10:53 pm

snigelx wrote: I am saying this is not in accordance to most charts I find that chart midi note# and frequency. These charts say midi note# 81 is actually 880Hz. Do you know why this is?

I have no clue if there's any standard here

Also, if you look in ACE one will notice that the value range for cutoff is 1-150 versus Z2 1-100.

You should have a range from 0 - 150 on Zebra's VCF and XMF as well.

snigelx · Post by **snigelx** » Fri Jan 15, 2010 11:30 pm

xh3rv wrote:
snigelx wrote: You should have a range from 0 - 150 on Zebra's VCF and XMF as well.

oops. yes, i do not know why i said that. they both have ranges of 1-150.

...And after more research on the web i think i see now the the reason for discrepancy but this brings a new question.

the MIDI standard extends midi note number 1 - 127, which corresponds to approx. 8Hz - 12kHz freq range. However, in the ACE manual it states "The cutoff value is measured in semitones from 0.00 to 150.00, a range of 12.5 octaves."

so if 1 midi note number = 1 semitone and 12 semitones = 1 octave. 150/12 is 12.5. OK but what, then, is the 'actual' frequency range of the filters in the two synthesizers? 22Hz-22kHz is about 11 octaves by my calculation

PaulSC · Post by **PaulSC** » Sat Jan 16, 2010 12:01 am

If MIDI note 69* has its standard tuning of 440 Hz, then the usual correspondence between a frequency F and the nearest MIDI note N is:

N = round[12*LOG2(F/440)+69]

And conversely

F = 440 * 2^((N-69)/12)

*Edit: I erroneously had "60" in my original post

snigelx · Post by **snigelx** » Sat Jan 16, 2010 12:10 am

PaulSC wrote:If MIDI note 60 has its standard tuning of 440 Hz, then the usual correspondence between a frequency F and the nearest MIDI note N is:
N = round[12*LOG2(F/440)+69] And conversely F = 440 * 2^((N-69)/12)

Hi Paul! yes, i saw that formula but midi note 60 is 261.6 Hz according to midi note chart. also, i am wondering now about this "12.5" octave frequency range in Zebra2 and ACE if value 1-150 is a 12.5 octave range, what then is the frequency range being used? (??Hz to ??kHz)

PaulSC · Post by **PaulSC** » Sat Jan 16, 2010 12:12 am

Aargh, i meant MIDI note 69, the orchestral "tuning A". Sorry!

groundhum · Post by **groundhum** » Sat Jan 16, 2010 12:43 am

cf. http://www.kvraudio.com/forum/viewtopic ... highlight=

PaulSC · Post by **PaulSC** » Sat Jan 16, 2010 12:47 am

As to your question about the actual frequencies involved, I don't know; it would be interesting to hear from Urs or from someone else who knows these instruments at that level of detail.

Meanwhile, the Empirical Method...

When resonance in ACE is high enough to produce stable self-oscillation, the filter tuning seems just like the oscillator tuning, with note# 69 = 440 Hz (if you set the master transpose to zero). So we might safely calculate as follows:
- The bottom of the range is 440 * 2 ^ ((1 - 69)/12) = 0.866195722 Hz
- And the top of the range is 440 * 2 ^ ((150 - 69)/12) = 4735.928611 Hz

It surprises me a bit to see the top end of this range so low (less than 5 KHz), because I think about a fully-open LPF and expect it to pass higher frequencies. But maybe it shouldn't -- the filter isn't infinitely steep, so it slopes off gradually above the cutoff frequency.

Please don't take this as a definitive answer! [Edit: Especially since I missed a decimal place when calculating both figures in Excel; multiply by 10 for answers that are basically correct, and see Urs' comments below for the real deal...]

PaulSC · Post by **PaulSC** » Sat Jan 16, 2010 12:47 am

Thanks for the link to the other thread, groundhum!

Urs · Post by **Urs** » Sat Jan 16, 2010 12:48 am

I don't have the figures in my head, but all filters open as far as they can at 150. The last octave is basically either stretched or truncated depending on the actual algorithm. Some algorithms get instable if the -3dB cutoff point gets higher than 15kHz, some other algorithms happily go up to 20kHz. Don't ask me which is which. It's been a long course of test and try to max out the values as good as possible.

The XMF in 2.5 mode goes up to a bit above 20kHz for the -3dB Cutoff point.

Urs

Urs · Post by **Urs** » Sat Jan 16, 2010 12:56 am

Paul,

You've missed something in there... it's off by factor 10.

The frequency at 150 would be 47,35 kHz. That's why they're transposed down one octave so that 81.00 corresponds to 440Hz. This makes 150 a frequency of a bit more than 23000 Hz.

Urs

PaulSC · Post by **PaulSC** » Sat Jan 16, 2010 1:04 am

Ah, that seems much better

I must have had a typo in my Excel spreadhseet the first time around; this time I'm getting the correct figures.

snigelx · Post by **snigelx** » Sat Jan 16, 2010 1:31 am

Big Thanks to all!

@Paul - thanks for that descriptive insight and calculation. now i think i can better guess the freq in both synths based on all of you guys input.

@groundhum (i hope you are isolated

) BTW,

NICE chart. thank you for taking the time to do the extended calculations above the standard! SAVING THIS

@URS,... so THAT is what you meant by "...1 octave below..." in the Zebra2 manual. You transposed the midi note number scheme or semitone value by an octave from the _actual_ frequency scale for 'technical' reasons. I'll just say that makes sense based on what you said about instabilities in certain algorithms. I know little to nothing about the deep mechanics of DSP. AAMOF, I did not know all that about the instability factor. Interesting part about the 2.5 XMF 3dB filters. It makes me want to dive into it even more to see what can be tricked out of this thing at those extremes.

Thanks again, all of you.

u-he

KVR Audio

Cutoff Frequency in VCF's of Zebra2