On Mar 6, 2011, at 12:33, David Kastrup <d...@gnu.org> wrote: > James Lowe <james.l...@datacore.com> writes: > >> The google tracker issue does work in as much as the main point of this is >> to split the Lilypond score into 'lines' and each line is exported to an >> EPS file - this bit is done by Lilypond (IIR) you can then manually put >> each line into a video using most standard video editing software, this >> bit is done by the 'user'. > > That would be more like flashing line by line rather than a continuous > horizontal scroll, right? > >> Then (again using the video software) you use the timing marks created >> manual, the music itself (that can be generated separately as midi >> from LilyPond or be a real recording) and the snippets of music to >> create a 'video'. It works but it is very clunky, takea a fair bit of >> work. If you are comfortable with your video editing software, and >> have a good recording of you music then fine, but I am not sure it is >> quite the same as a 'follow along' video. > > I'd likely be able to write (in a nontrivial amount of time) the > required code for matching Midi time to sound file time. That's > reasonably manageable signal processing. > > Correlating grobs to miditime would be another task. And then, of > course, generating the equivalent of one long strip, and making that run > through, likely in a nice smoothed motion with some cursor following the > notes more accurately. > > -- > David Kastrup > >
Just to avoid potential wheel reinventing, make sure to check out the conference proceedings of the 2010 Journées d'Informatique Musicale. I remember someone presenting a time-point based score follower there. ~Mike _______________________________________________ lilypond-devel mailing list lilypond-devel@gnu.org http://lists.gnu.org/mailman/listinfo/lilypond-devel
