https://www.kvraudio.com/product/kaze-by-xoxos
i made this fast, i posted it fast. when i put the demo together last night i discovered the noise level detection algorithm was 44100 centric, i've just thrown in a scalar for that but i would still anticipate better results with 44100 since only so much testing occurred.. (i did do a quick brush of my old wav loading system, it might load a few more 8/16/24 bit files with longer headers now....)
it was a rush to make, this, so fast. wow so simple. but i think i was stoned enough to make a funny in the documented code. thank god for weed.
thanks to julius smith for the basis. "hope it works well for you". just from yesterday evening, ..getting a flat pitch with all harmonics does sound like just vocoding everything with a saw.. so eg. analysis of percussion, stuff with inharmonic partials.. isn't that awesome. but wow on the horn for me. i found maybe up to 4 octaves playable range "believable". haven't tried it with strings, or other "harmonic series" instruments yet but a bit of fiddly programming with some decent samples (iirc sydney uni acoustic studies have several quality one shots of instruments across dynamic range), might make some nice patches.
of course no one is reading this but note, the wav file loading thing does take a good while, about a second for every second... why i don't know. also note that once a file is loaded, there will be no sound if the "read" position is static in a silent frame... ..you just get a "read" position param.. you gotsta move it yoself.. quickest ways to achieve constant speed playback are linear lfo or envelope..
once the files are loaded... aaaand analysed.. cpu use is nice, i had 6 running 25-30% of single core (with filters off.. no filters in audio demo..)
kaze
-
- Banned
- Topic Starter
- 12368 posts since 30 Apr, 2002 from i might peeramid
you come and go, you come and go. amitabha neither a follower nor a leader be tagore "where roads are made i lose my way" where there is certainty, consideration is absent.
-
- KVRAF
- 3477 posts since 27 Dec, 2002 from North East England
I completely missed the last few and I'm delighted to see new synths from you. Wild ideas as always. Thank you for sharing these.
-
- Banned
- Topic Starter
- 12368 posts since 30 Apr, 2002 from i might peeramid
ty, discussion here.. big name synths with multiple iterations.
had no idea/intention of doing audio, but south node, mercury retrograde, there it is. i'll have a quick go at using cepstra for reverb removal effect, and then i expect i'm absent again.
any developers, i think this technique would make a fastantic customisable voice synth product (i am not interested in commercial development myself).
have i missed any freeware releases in the last three years that do anything one couldn't already do..
had no idea/intention of doing audio, but south node, mercury retrograde, there it is. i'll have a quick go at using cepstra for reverb removal effect, and then i expect i'm absent again.
any developers, i think this technique would make a fastantic customisable voice synth product (i am not interested in commercial development myself).
have i missed any freeware releases in the last three years that do anything one couldn't already do..
you come and go, you come and go. amitabha neither a follower nor a leader be tagore "where roads are made i lose my way" where there is certainty, consideration is absent.
-
- Banned
- Topic Starter
- 12368 posts since 30 Apr, 2002 from i might peeramid
made my idea for reverb removing, uses 30% of single core for mono, 60% for stereo. "seems to do something to remove resonances" but too bloody lazy to make a proper "sound with reverb" file, and the rooms here have remarkably damped acoustics..
..the test files i've auditioned it with are too complex, "it might do something to cancel reverb" in files with "just voice and reverb" but totally can't tell due to extreme disinterest. more gonig on in the file.. "i could hear footsteps on leaves crisply" a bit like an expander. god knows if i'll ever use it.
but i can't post it, file is 1073kb, and too untested to make official thingy.
but if anyone is curious, i of course just took the cepstral magnitudes and slapped them back on the original phases, with parameter for setting gain of liftered bins. it does recreate the original signal and waveforms look like they are exactly losing resonances.. like the source waveform is the smoothed version.
..the test files i've auditioned it with are too complex, "it might do something to cancel reverb" in files with "just voice and reverb" but totally can't tell due to extreme disinterest. more gonig on in the file.. "i could hear footsteps on leaves crisply" a bit like an expander. god knows if i'll ever use it.
but i can't post it, file is 1073kb, and too untested to make official thingy.
but if anyone is curious, i of course just took the cepstral magnitudes and slapped them back on the original phases, with parameter for setting gain of liftered bins. it does recreate the original signal and waveforms look like they are exactly losing resonances.. like the source waveform is the smoothed version.
you come and go, you come and go. amitabha neither a follower nor a leader be tagore "where roads are made i lose my way" where there is certainty, consideration is absent.
