Dr_wav only loads the wav file from the disk to an array. It does not play the wavs.
I use Rack sample rate converter only to resample from the wav sample rate to Rack sample rate. I had thoughts about what Vortico says. Use the resampler to change the pitch but I did not code anything. I just do what every other open source sampler does. iterate the array fast or slow.
Like Squinky. Labs says simple linear interpolation improves the sound quality a lot over no interpolation. Other methods like BSpiline only give you more s/n ratio.
I also looked at my code and the interpolation is broken. It allways uses no interpolation. I will update the module as soon as possible.
Yeah, thatâs the one. Sounds great. Or maybe it was a newer one thatâs in the spirit of that? The one that lets you switch between two waveforms that are in sync?
good catch! I found long ago that I make mistakes like this all the time. I once made a plugin that played every group of 4 samples backwards! Since then I try to write unit tests to keep myself honest. Most of my VCOs have unit tests for alias, like these ones: https://github.com/squinkylabs/SquinkyVCV/blob/master/test/testVCOAlias.cpp
I remembered seeing a video on how the Functional VCO did have a little bit of aliasing⌠when I test it, it only aliases at very high frequencies, outside of the musically useful range, so they seem to be really good, and Vult oscillators canât even reach those frequencies so I guess it belongs in that category as well. Itâs better than the Fundamental VCO, which also does alias at high frequencies, but is probably still good enough for most common use cases.
Saws aliases a bit more and Chebbyshev is limited by itâs uppermost fundamental, which folds right back if it crosses nyquist. But theyâre probably still good.
I only started recognizing aliasing for what it is recently, so I probably should collect more listening experience before judging oscillators⌠but with most samplers itâs really harsh.
Both of the manuals for these are quite explicit about that. Saw attempts to accurately emulate exactly how the JP-8000 aliased, but has two options to reduce it dramatically. And, again, the Cheby manual tells you right up front that the top harmonics can fold over, but if you keep the fundamental below 2khz it will not alias at all.
My only point is that these VCOs are carefully designed to trade off CPU usage for sound quality, and most importantly are very up-front about exactly what they do.
Yes, for sure there are some good dirty sounds out there. I never could stand the sound of the DX-7, but I know some people love it. I can assure you that at the time is was not marketed as a âgritty aggressiveâ synth. It was sold more more on digital purity and sounding exactly like ârealâ instruments.
I can assure you that at the time is was not marketed as a âgritty aggressiveâ synth.
Thatâs for sure. But the DX7 certainly has some aliasing, especially when compared to some of itâs successors. But I tend to agree about the DX7, not a huge fan, even though it does have a time and place in certain genres that could likely never be completely replaced.
Itâs better than the Fundamental VCO, which also does alias at high frequencies, but is probably still good enough for most common use cases.