From the original SynthMaker forum (2004–2014), with its members’ names and dates.
Tom7777 ·
Well good luck with it, i still would like to see a version made by the devs though and i think it should have just been any input size to begin with rather than 256, hopefully it won't be a massive wait before i can try a non limited version though as i get CM mag.
stefan ·
infuzion wrote: unkargherth wrote: Infuzion: How good is it in terms of aliasing?
Can you show us a splah screen comparing the spectral content of two ( Your wave table and wavedraw one) ?
1) Urrmmm, how do I test for aliasing please?
Play very high notes and see if they sound "clean" or not
The lack of aliasing is the entire purpose of the Wave Table component, otherwise one could just use a ramp oscillator (multiplied with 256 or whatever length the waveform is) and a Float Array to Poly component. Then one can use waveforms of any length, but one will get heavy aliasing.
infuzion ·
stefan wrote: infuzion wrote: 1) Urrmmm, how do I test for aliasing please?
Play very high notes and see if they sound "clean" or not oh yea, loads of Aliasing! Makes me ask:
1) Shouldn't a simple but steep low-pass filter fix that? Which filter would be best?
2) For the original purpose (OSCs for very low notes) does it really matter? Chances are, there will be a low-pass filter towards the end anyhow.
stefan ·
infuzion wrote: stefan wrote: infuzion wrote: 1) Urrmmm, how do I test for aliasing please?
Play very high notes and see if they sound "clean" or not oh yea, loads of Aliasing! Makes me ask:
1) Shouldn't a simple but steep low-pass filter fix that? Which filter would be best?
That's the thing, you need to filter the waveform_before_ you play it back at a too high pitch. That's what the Wave Table component does. It creates a set of 128 "variations" of the original waveform, where each variation has filtered out one harmonic more than the previous one (so the last variation is usually just a sine).
You can't do this filter after you've read from the table, it has to be done prior to the table read.
infuzion wrote: 2) For the original purpose (OSCs for very low notes) does it really matter? Chances are, there will be a low-pass filter towards the end anyhow.
For low notes it does not matter, no. But a "post" low-pass filter really doesn't help to remove aliasing. Or, well, it depends on where, relative to the base frequency, the aliasing is. Imagine playing back the waveform at like... 14kHz, using a sample rate of 44kHz, almost all the aliasing will occur below 14kHz. I say almost, because one harmonic will end up at 18kHz as well. The rest of the harmonics will end up below 14 though
infuzion ·
OK, after hearing some of the low notes of Jobro's Digiwave, I really can see we need higher rez OSCs. [old link]
Wrote something here? To have a post removed or your name hidden, write to [email protected].
SynthMaker