Hi Yoshimi-developers,
so it seems the "padthread" effort could be ready for integration.
All intended features are implemented, we did some smoke tests and
could not provoke any further crashes, and the systematic tests
did not yet show anything of real concern. We see slight sonic
changes in the testsuite, but below -70dB, so these might be
qualified as rounding errors. The overall run times have
been improved by a very tiny amount (which is due to the
reworked memory management). However, when using the
new background build on PADSynth, we notice a small
overhead, and cross-fade or retriggering obviously
also cost some additional performance, when used.
Any comments, concerns, remarks etc. would be welcome now.
Moreover, I'll be happy to explain more technical details
or the reasoning of some of my design decisions, in case
someone is interested in those deeply technical aspects.
On my part, I have still to investigate two minor problems,
not sure if these are bugs or irrelevant. And I'll also write
some technical notes for the developer documentation.
The "padthread" branch is now rebased on top of current master
(with the license clean-up). Incidentally, this branch also
contains the commits with Will's initial work for this feature.
-- Hermann
_______________________________________________
Yoshimi-devel mailing list
Yoshimi-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/yoshimi-devel
