Hi
The last day's I'm working on a little tool which allow me to generate a
SondFont 2 to be used with fluidsynth from a single audio file.
It allow now to clip out a part of a file to be used as OneShoot
instrument sample, select a part to be used as Looped Instrument. It
detect the dominant Frequency and set the RootKey and PitchCorrection to
be used in the generated SoundFont. Additional it allow to set the
values for Chorus and Reverb.
This allow to generate quickly a well tuned SoundFont from any sample
you've laying around, or grep out a sample from a audio file to use it
as Soundfont.
Note, it will use only the first channel of a audio file, regardless how
many channels it have.
Supported audio file format is what ever libsndfile supports.
Testers been welcome.
https://github.com/brummer10/sf2generate
regards
hermann
Unfortunately mhWaveEdit is abandoned, but IMHO is still very valid for
quick audio editing of single files, trimming, fade-in / out, etc. and
as discussed in a recent thread I haven't found alternatives on Linux.
It also looks like it won't build any more 'out of the box' from the
source package or AUR packages (I'm on manjaro)
The 'latest' source package gets a bunch of errors / warnings
https://bpa.st/4TLQ
This was by doing a:
./configure --prefix=/usr --without-oss --without-portaudio
--without-oss --with-double-samples --without-check-casts
and then just
make
I was eventually able to successfully compile and run by brute force
suppressing two of the most relevant warnings:
make CFLAGS="-Wno-incompatible-pointer-types -Wno-deprecated-declarations"
Ideas / thoughts?
Lorenzo
