From the original SynthMaker forum (2004–2014), with its members’ names and dates.
bootsy ·
Howdy,
I'm currently using the frequency splitter as found in the analog modelling kit here and try to understand the scaling used there, since it's not samplerate independent (corner frequency doubles as frequency doubles). Can someone help here or just elaborat that green Scaling module?
Simplyfied osm attached, check e.g. Buddes analyzer tool too see how curve moves when changing SR.
Cheers mates, bootsy
MichaelBenjamin ·
i think you just need to multiply with 44100/sr to make 0-1 frequencies scaled to sample rate, but correct me if thats wrong.
MegaHurtz ·
The scaling module is a logarithm (of it`s natural exponent i suppose?) Anyway this makes shure your controller has equal steps, so you have more usefull steps to play with. For instance more resolution on lower frequency`s.
You could say the filters in synthmaker always scale up the frequency in relation to the samplerate. This is becouse the feedback they work with gets passed through faster, and so the response gets more accurate. Thus being able to do higher frequencys. The way to make shure this doesnt happen is like MB said.
bootsy ·
I did tried that multiply with 44100/SR already but it did not show up in the analyzer but in your example I saw that you used the sample and hold and that was obviously the reason why the plug-in got no information on the SR change in the analyzer.
So, thanks a lot, everything is fine now!
greets, Herbert
infuzion ·
bootsy wrote: I did tried that multiply with 44100/SR already but it did not show up in the analyzer but in your example I saw that you used the sample and hold and that was obviously the reason why the plug-in got no information on the SR change in the analyzer.
The Sample & Hold should have no effect on the SR rate changes. The S&H is there only to keep the Float Divide before it from being re-calculated every time the cutoff value is changed. If the SR is not sending information with out it, then that is a reportable bug. Likely related to other trigger issues.
oddson ·
Quite apart from any trigger problems on your attempt to scale to the sample rate I have serious doubts that the displayed frequency accurately reflects the true value of the cutoff in the filter itself (even for the 44.1 KHz sample rate).
Since the filter is ASM and I'm illiterate with it I can't say for certain but if the cutoff value the filter code is expecting is the standard normalized linear mapping between zero and nyquist then the frequency being displayed is not at all accurate.
My math skills are not up to an algebraic demonstrastation; this rearranged OSM will show what I mean:
McBarGig ·
Bootsy...Did you get this to work? with all samplerates using the 44100/SR fix?
Wrote something here? To have a post removed or your name hidden, write to [email protected].
SynthMaker