hello, here’re some waveforms I’ve translated; someone works good, others seem to have problems due to the high frequency content (not sure). I read udo suggests to cut frequencies above the ns/4 limit (they wrote 512 hz, but I think they meant 5512,5 considering it’s 44100 khz). However feel free to experiment with them. And let me know if you figure out what’s wrong.
I don’t think it is 44.1kHz, it doesn’t make sense to think of it in those terms, as the frequency of the waveform will change depending upon its pitch. Better to think of it as 4096 points that always play back, irrespective of pitch/sample rate.
Easy to remove any harmonics above 512Hz with a steep LPF, but be careful it doesn’t push the waveform’s level above 0dB, might need to scale it back after filtering.
That’s correct! Feeling a bit dumb, I was thinking in order to not alter the frequency, starting from Ensoniq SQ80 sample , and that’s what I was missing; I was triying to fit them altering the frequency, that’s why they were sounding high pitched. I’ll do few other test and see what happens.
However this makes me think that: if the sample frequency would change depending from the source (this is because I start from an existing sample), that 512hz of cut frequency might change. Also what about the introduction of latency? Audacity does not have FIR filters (I think)
, and that’s what I was missing; I was triying to fit them altering the frequency, that’s why they were sounding high pitched. I’ll do few other test and see what happens.