Aaron Laws
2017-06-27 02:21:20 UTC
How do release samples work?
For instance, say I have a .5 second attack and a 3 second loop and .5
second release. If I play the note for four seconds, I will be .5 seconds
into the second iteration of the loop. When I release the key, does the
audio driver skip ahead to the release sound immediately? What kind of
effort is made to avoid the blip that might be caused by a mis-alignment of
phase?
I can think of two possible strategies off the top of my head:
1.) Quickly fade out the loop and fade in the release, and
2.) Wait until the loop signal becomes compatible with the beginning of the
release signal. For instance, if the release signal begins like a sine
wave: ascending from 0, the software would wait for the loop signal to
begin to cross 0 ascending.
I assume different software handles things differently, so I'm interested
in hearing how this is accomplished, and whether this is specified in SFZ,
or if that is up to the particular SFZ implementation.
In Christ,
Aaron Laws
For instance, say I have a .5 second attack and a 3 second loop and .5
second release. If I play the note for four seconds, I will be .5 seconds
into the second iteration of the loop. When I release the key, does the
audio driver skip ahead to the release sound immediately? What kind of
effort is made to avoid the blip that might be caused by a mis-alignment of
phase?
I can think of two possible strategies off the top of my head:
1.) Quickly fade out the loop and fade in the release, and
2.) Wait until the loop signal becomes compatible with the beginning of the
release signal. For instance, if the release signal begins like a sine
wave: ascending from 0, the software would wait for the loop signal to
begin to cross 0 ascending.
I assume different software handles things differently, so I'm interested
in hearing how this is accomplished, and whether this is specified in SFZ,
or if that is up to the particular SFZ implementation.
In Christ,
Aaron Laws