Export seems to fail on sample dependencies
The entire project is also here, with the samples: https://drive.google.com/drive/folders/14AifqH13y7N_6iInuc-PshmDkhzcXoaS?usp=share_link
I have been struggling with this export the past days. I read an old post where they talked about that the rnbopat needs to be a file, not just a subpatch, and I have been trying with @title, @file and @patchername... There used to be 3 rnbo-subpatches in the top rnbo, one for midi, one for grooves and one for filters, but it seemed like midiin only wanted to work in the top. Eventually I crammed everything in to one singe rnbo.
I made a successful export once, when I had accidentally deleted all the groove-objects.
Putting them back made the export fail again. So now I suspect the sample dependencies is the culprit.
Can anyone make sense of it?
There was a bug a few versions back that caused the C++ export to fail if the resulting rnbomatic_binary.cpp got too large (too many large wav files), my work-around at the time was to use MP3s instead of wavs, I reported it but not sure if it got fixed. Maybe try updating to the latest RNBO or replacing your wav files with MP3s
one could reduce 24bit 48000 to what normal human accepts as good quality.
unless you make this for 1000 $ chinch cable audiophile type listeners ?
that would reduce sample size to almost half.
Few files I listened too have no significant difference through the full length,
a short looped snippet would do just as well.
I was using the latest RNBO, but reducing file size eventually worked. After going down to aroung 100MB I got an error message in the export log about "request too large" or something like that, which was progress since before this there had been nothing in the log pointing to size.
I chipped away and retried a couple of times, and when I got the audio files down to around 33MB the compiling succeded.
Still, Mr. Source, I don't see why you have to be so salty in everything you write? I'm just an old fart dreaming about making my first plugin and trying out ideas. I had no indication that size was going to be an issue, so I made a couple of loops on the theme of "natural noise" and tried to make the looping sound seemless.
One could argue that a normal human also accepts my crude pitch bending...
Nonetheless, thank you all, for 'tis compiled and I am progressing!
Crude pitch bending welcome!
Also, there are actual good reasons to record audio at 24bit 48000Hz or higher (ultrasonic sound downpitching, unmixed material to be mixed with less aliasing/artifacts, preservation...).
recording higher than 48k should not result in anything interesting, except when your microphones actually work above that range. but what surely helps is high upsampling inside effects which are combining delay with pitch-down and pitch up in unforeseen combinations. in this case, only the tapping buffer´s maximal length will be greater, which is hopefully a lot shorter than your longest recording.