On 08/29/2014 08:27 PM, Stian Jørgensrud wrote: > Of course it could, but what do we need lfo controllers for then :) > You are making a sine wave with automation instead of using the lfo > controller!
A cspline interpolated 2-point curve does not create a sine wave... you'd need cosine interpolation for that. Anyway... LFO's and automations work very differently in LMMS... automations "push" their output to the model, whereas controllers are "pulled" to the model by the model. When it comes to sample-exact controls, things get a bit more complex... the sample-exactness in automations is kind of hacky, it simply interpolates between the latest value pushed and the one before, so the resulting curve does not accurately represent the actual automation... I intend to address this later, though - but for that, we need to modify how automation tracks are treated by LMMS - we need to process them per period, not per tick, like other tracks. (Or even better, we could not "process" automation tracks at all, and instead let models just pull data from them...) Again, I'm digressing... automations are a bit more flexible and easy to use than controllers. If you only need the one waveform pattern for the model, then LFO controller works fine, but if you want to variate it, if you want mix in waveforms and longer sweeps... then it becomes really tricky to use LFO for that, and much easier to do with automation. So yeah, I think extending automations would be a good feature. ------------------------------------------------------------------------------ Slashdot TV. Video for Nerds. Stuff that matters. http://tv.slashdot.org/ _______________________________________________ LMMS-devel mailing list LMMS-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/lmms-devel
