On 29.10.20 22:37, mau...@smoors.de wrote:
That feature reqeust is quite understandable!
Wouldn't be adding noise a task for the software interpreting the MIDI file (Ardour in this case)? In Hydrogen, right now, the humanization does happen in the audio engine by drawing random numbers (each time a different set). Establishing the noise at Song level would be possible right now (as we are only using Gaussian white noise) but would be quite limiting in the future because it would not allow for state-based and correlated noise (which I intend to introduce at some point to achieve a more "human" feeling). On the other hand, those functions deserve to be written in a more modular way and could be reused when exporting MIDI (on other words, the Song would pass through some parts of the audio engine once before being exported).
But is the MIDI format capable of reflecting the onset and pitch variations given in float in the first place?
Best, Phil _______________________________________________ Hydrogen-devel mailing list Hydrogen-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/hydrogen-devel
