[gmpi] Re: Req 76,78

  • From: David Viens <davidv@xxxxxxxxxx>
  • To: gmpi@xxxxxxxxxxxxx
  • Date: Tue, 08 Feb 2005 14:08:24 -0500

Hi again

Does anybody have any idea what the ratio is of rewired apps that have
use the IPC hack(s) that Angus mentioned and those that actually split
their DSP/GUI code apart?

Well looking at which rewire slaves supports loading VSTs gives you a head start. (except FL, which uses the other trick, the GUI and DSP, both being in the host space, if i understand correctly )

i share that gut feeling, even for Linux and OSX. i'm actually pretty
impressed by the performance you found - i think linux and osx can
easily replicate that, but also couldn't run many JACK clients (the
obvious equivalent configuration) at those latencies. Maybe 2-4, but
probably no more.

Well there are a coupla catches, we currently use two different IPC messaging tricks (optional), and our most efficient one craps out some drivers (digidesign ASIO) for reasons we have yet to explain :)
- one being perhaps what we're doing might not be considered legal by everyone- But both only works on NT5+ cores _only_, no win9x im afraid (but dont really care)

That said, im not an expert in the field and I mostly experimented for a month trying everything i could find in msdn :) It may well be possible for say a MS low level guru to do a bit better, especially in the stability domain.

however, notice that this actually has little to do with the issue at
hand, because nobody is suggesting that GMPI should use an
out-of-process model. your numbers do suggest, though, that IPC on
windows is certainly fast enough for people who do want to split the
DSP and GUI code; this has to be true, because IPC on windows appears
to be fast enough to even split the DSP code of several rewire clients
across multiple processes. That's a good thing to know.

Yes sorry i know we were running away from the main subject :)

-- David Viens, Plogue Art et Technologie Inc. Montreal. http://www.plogue.com

Generalized Music Plugin Interface (GMPI) public discussion list
Participation in this list is contingent upon your abiding by the
following rules:  Please stay on topic.  You are responsible for your own
words.  Please respect your fellow subscribers.  Please do not
redistribute anyone else's words without their permission.

Archive: //www.freelists.org/archives/gmpi
Email gmpi-request@xxxxxxxxxxxxx w/ subject "unsubscribe" to unsubscribe

Other related posts: