seejayjames

I'm making a patch to track beats---kind of a follow-along beat-tapper---and was thinking of using jit.gl.gridshape for the display. Or possibly jit.gl.sketch with points and point_size big. Then they would move (change position attribute) to match the audio/MIDI playback. I also may use a texture on the gridshape to show the pulses, from a jit.matrix that's filled with a matrixctrl.

I plan to have (say) 8 tracks going at once. What I'm wondering is whether it would be better (optimization-wise) to have each track be its own jit.gl.render with one gridshape/sketch, or have one big jit.gl.render and have 8 gridshapes for the tracks? And does having a texture---which I may or may not use---make a difference in this setup?

I'm thinking I'll just move the gridshapes or jit.gl.sketch to match the playback, but I know you can also change the position attribute of jit.gl.render, and if there's only one object in the scene, the result will be the same. Is there a performance trade-off if you move jit.gl.render versus moving the gridshape? And does it matter how many shapes are in the scene when you move jit.gl.render?

It needs to be pretty optimized (say 50 fps) so that the scrolling looks smooth, at 20-30 fps it's workable but looks a little clunky.

Optimization question