From: Tim E. R. <ter...@ro...> - 2011-06-05 19:42:50
|
On June 5, 2011 01:44:00 pm Dave Stikkolorum wrote: > Hi, > > On 05-06-11 18:31, Tim E. Real wrote: > > On June 5, 2011 11:20:57 am Robert Jonsson wrote: > >> Hi Dave, > >> > >> 2011/6/5 Dave Stikkolorum<da...@dd...>: > >>> Hi , > >>> > >>> I was wondering if it would be an idea > >>> to let muse start up some programs. > >>> > >>> I am working with hydrogen en Zynaddsubfx and qjackctrl > >>> > >>> I like to start them from muse. > >>> > >>> especially qjacktrl, is that an idea? > >> > >> Interesting idea, it does pose some problems though, especially with > >> jack since it has to be started before the audio parts go active. > >> > >> I know that at one point we had support for autostarting jack if it > >> wasn't started, due to some problem I can't remember it was > >> deactivated. > > > > LASH, which MusE supports, will try to auto-start Jack. > > > > However I must point out that for some reason this is broken right now. > > When we start MusE with LASH support enabled (no -L switch), > > LASH does in fact start Jack, but MusE is unable to use it and does > > not show any Jack ports. > > But even worse than that, the instance of jackd which LASH starts > > via MusE, hangs around after quitting MusE, requiring a manual kill of > > it. > > > > This really should be fixed, as most people are not going to realize > > that Jack should either be started manually, or else that the -L switch > > should be used, or else realize that this instance of jackd is still > > running. > > > > If Jack is not manually started by the user before starting MusE, > > then I highly recommend always starting MusE without LASH (-L switch) > > until this problem is fixed. > > > > Of course, it really is time that we looked at adding Jack session > > support to prepare for the future. > > > > Tim. > > I tried lash_panel with hydrogen and muse. > Indeed this looks like a nice way of handling things. > I noticed muse (muse 2 beta) says something about jack didn't start, but > he did. > And it won't close by lash_panel. > > So are that the issues at the moment? Yes I think so. I think if you manually start Jack before starting LASH (and MusE etc), then things should go fairly smooth. The last time I checked, doing it that way worked OK. It is only when you don't start Jack manually before hand, that the problems I mentioned start to happen. I will try to check and verify the operation later... > > I also noticed that programs like qsynth and zynaddsubfx don't seem to > do lash. Pity. Not to confuse anyone more, but in the older muse_qt4_evolution branch, a zynaddsubfx native MusE soft synth was included. That would help the situation here in MusE-2, we could automatically start it. I looked at this the other day and it appears it may be fairly easy to import this feature to our code, maybe even bringing in the latest version of zynaddsubfx. Maybe worth a try. But I have heard it said that zynaddsubfx is not a real-time friendly synth, that the coding is not up to real-time standards. So is it questionable to try to do? ... Is this one of the reasons the Yoshimi re-work of zynaddsubfx was done? Tim. > > Dave > > >> I think the easiest (linuxish) way to do this is to create a script > >> that starts all things you need and create an icon for this script (if > >> you like) > >> > >> Example start.sh > >> #!/bin/sh > >> jackd -d alsa ....& > >> sleep 1 > >> zynaddsubfx& > >> sleep 1 > >> muse2 > >> > >> Regards, > >> Robert |