On 04/02/2014 03:22 AM, midi-pascal wrote: > > On 69-12-31 07:00 PM, wrote: >> On 04/01/2014 09:00 PM, Stian Jørgensrud wrote: >> I like the first idea. The second really only apply for Sf2 drums :-) >> Could be implemented if it is simple, but I can understand that it >> might overload LMMS too. >> >> For SF2 it might actually be more difficult, since SF2 is a MIDI-based >> plugin that only gives LMMS a single audiostream, we can't divide it in >> any way. So it would not be simple at all, it would require some heavy >> architectural changes. >> >> For native instruments, dividing notes to different fx channels based on >> pitch might actually be not that hard (technically - UI considerations >> aside), but then, would there be much of a benefit in doing so? I can't >> see any of the native instruments really benefiting from it. Like you >> said, it's only really useful for SF2 drumkits (and maybe Zyn drumkits >> too, but Zyn is also MIDI-based). > Before beginning a deep diving in the code, there is something I don't > quite understand: > > How lmms manages to display the events (notes and all) in a track, > allow editing with a piano roll with all its capabilities to make > changes on the fly from "a single audiostream" from sf2, as says dii? > > Where does this magic come from? =-O > If we can display and edit the notes of a track , I cannot believe we > cannot clone a track and set a limit on the notes by filtering the events. > Perhaps I am completely wrong in my understanding about the inner > working of lmms but this sounds a bit strange to me. > > I'm really interested to understand this phenomenon. ;-)
Well there might be a misunderstanding here. If you just want to divide the notes themselves to different tracks, that's of course possible. If you want to divide the *output* of those notes, ie. the actual sound they make, that's impossible. The note events themselves are separate, and we can do whatever with them. But once the note events get converted to MIDI and sent to the SF2 plugin, then the SF2 plugin only gives back a single audiostream where all the notes are mixed in. At that point, we can no longer separate the notes. So if you only want to statically divide the note events themselves, and not dynamically divide the audio output of the notes, then yes that's possible. You could even write a plugin for LMMS that does this for you (split notes from one track to several). I'm not 100% certain but I think that should be possible within the current plugin framework. Of course, then you'd have to have different instrument plugins (or copies of the same plugin) for each track. Because a single SF2 plugin can only give a single audiostream.
------------------------------------------------------------------------------
_______________________________________________ LMMS-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/lmms-devel
