by Jerm » Thu Dec 18, 2003 1:01 am
The only reason to do this, like I said before, is if your soft-synth supports more than 1 MIDI channel. In this case you only need one instance of the soft-synth and it provides up to 16 channels of MIDI which you most likely would want on separate MIDI tracks.
For example take the DLS-MusicDevice. This soft-synth takes a Soundfont such as a general MIDI bank. Then with 16 channels setup on it you can assign bass as MIDI channel 1 to track 1 and guitar on MIDI channel 2 to track, etc, etc. This way you will definitely get a big processor savings especially for synths that have effects that can be shared.
In the above example all 16 MIDI channels share the same reverb. If you do it the other way where each track gets a separate instance of the DLS then you are creating 16 reverbs which will definitely start to bog down most processors. There is also CPU overhead in the instances themselves. And last but not least, you may run into problems, especially if you have a setup like the following:
Track 1 MIDI Data -- the DLS Music Device plug is on this track
Track 2 MIDI data -- to DLS Music-Device (on track 1) Channel - 2
In this case since the plug-in is on track 1, when track 1 gets muted, so does track 2 because the plug-in is getting muted. If the plug-in is placed on the aux buss this problem is avoided.