Closed derselbst closed 1 week ago
1427 is not related to the filter changes
I don't think I have claimed that. However, #1427 seems to be caused by the filter's behavior, or at least the filter is involved here. That's why I currently believe it makes sense to keep it in the iir_filter
subfolder.
This adds a custom target to the build system
check_manual
that invokes rendering several test-MIDI files. The test files, along with potentially pre-rendered demonstration audio clips are checked into git lfs. The audio files rendered during build-time are stored inbuild/test/manual/
and can be used as reference, whenever a part of fluidsynth's logic is about to change. Esp. useful when changing legacy code, as in e.g. #1345.For the beginning, test files for the IIR filter have been added, taken from issues #1415, #1417, #1424, and #1427.
Additionally, an interactive Bode-Plot for the IIR filter has been added written in Matlab.