On 07/11/12 15:27, Kassen wrote: > On Wed, Nov 07, 2012 at 03:00:39PM +0000, David Griffiths wrote: >> Hi all, >> >> I've pushed some more changes to fluxa, work on the temporal recursion, >> but also a big fix for the internals of the sound engine. > > Awesome! > >> Anyway, it's now possible to have much more complex synth graphs playing >> faster music! > > Would that mean I can stop searching the node-cycling stuff I have > been at for 2 days now? I had some cases where fairly fast but not > overly complex music would consistently drop some notes (say a highhat > every 2 bars) in ways that did not seem affected by max-synths and did > not seem to stem from proper node-cycling.
This is quite possible under some circumstances... if you want to check - get the latest, and change the value COMMAND_DATA_SIZE back in CommandRingbuffer.h to: static const unsigned int COMMAND_DATA_SIZE = 4096; It will now print out a message when the ringbuffer is exhausted. > It sounds like what you did now might explain what I saw; I went up > and down the graph and cycling and after I figured out how it works I > had to consider that it was either fine or there was some very obscure > bug. My plan was to go into debug-printing mode tonight. > > If so I'm going on to "unversalise" my scales stuff and push that in, > then pitch it to you as "proper behaviour". Sounds great. cheers, dave
