On Thursday, April 19, 2012, Tom Breton (Tehom) wrote: > I want to make sure it works right when one changes devices, > especially when one deletes them.
The very first time I started a new session after building the latest changes, track 1 was playing some random program, not the program for the track/instrument I was sitting on. That's the only really obvious and shocking glitch I noticed in this test session. I just had one especially noteworthy finding. I started a new session with my new, otherwise stock autoload that's set to play through my usual MIDI device. I set up distinctive programs on several tracks and noodled around on the keyboard, successfully, including percussion. I want to see what happens when I set up to play drums on something other than channel 10, so I started QSynth and changed the connection to aim at that, instead of my usual hardware. (QSynth changes drum kits in the most standard way, so it's easier for me to use for testing; plus more users use that than any one particular hardware synth.) When I went back through the tracks, QSynth played with the desired programs. That is one very obvious sign that something in Rosegarden has changed. Previously, I would have had to go "jiggle" all the programs to get them set back up for the newly changed device. That's rather nice! NOTES: I did eventually succeed in using QSynth to play percussion on more than one channel, using different drum kits, but there were some glitches. First, after checking [x] Percussion, no drum sounds emerged right away. This almost certainly isn't new, and probably has nothing to do with your work, but now that I'm looking at it, it seems wrong. Checking the percussion box changes the bank to 1:0 in order to access the drum kit programs (changing to bank 1:0 is by far the most common method to set a channel to percussion mode across several different standards) and it changes the combo box to the "Standard 1" drum kit, but neither the bank change nor the program change are actually transmitted upon checking the box. The result is that a piano or whatever continues to sound until you "jiggle" the program combo box. This _could_ be a wrong channel thing, but I think it's much more likely a failure to transmit the bank/program change. Second, after I got that track playing drums, the next couple of tracks I clicked on also played drums. It seems like what probably happened was channel X hadn't been de-allocated, and was in use the next couple of times I changed tracks. Eventually that sorted itself out after I clicked around long enough, but it was momentarily confusing. The experience I had was a little glitchy, but nothing serious enough to concern me. People don't use drums on more than one track all that often anyway, and I did eventually work through the glitches to a successful conclusion. On balance, this latest build is VASTLY improved over the last build I tested. I haven't tried recording, changing documents, and myriad other things. I'll try to really get into some more complicated scenarios later on. In the meantime, everything looks extremely encouraging. Oh. I tried deleting the device I was using for everything. First I created a new completely blank one, then deleted the existing one. Not surprisingly, all tracks stopped making noise. After I hooked the new device to an output and moved several tracks to point to the new device, they all made noise. They played with the programs that were in use by those tracks before I deleted the device out from under them. In this case, there are NO programs in the new device, so what else could it do? So I loaded some programs into the new device and tried everything again. After I did that, I was no longer able to get sound out of anything. It looks like something broke in a pretty permanent way after loading the programs. I saved and reloaded the file. That fixed all the broken gray LEDs that were pointing at a deleted device, but it didn't get the sound issue sorted. After I got that to break, it stayed broken. Here's the resulting file for dissection. I quit and restarted with the same result. Something in the way this file is coded breaks your channel allocator thingie permanently. Even so, this is still a VAST improvement. I'm very pleased with the result of your second stab at all of this. -- D. Michael McIntyre
foo99.rg
Description: audio/rosegarden-composition
------------------------------------------------------------------------------ For Developers, A Lot Can Happen In A Second. Boundary is the first to Know...and Tell You. Monitor Your Applications in Ultra-Fine Resolution. Try it FREE! http://p.sf.net/sfu/Boundary-d2dvs2
_______________________________________________ Rosegarden-devel mailing list [email protected] - use the link below to unsubscribe https://lists.sourceforge.net/lists/listinfo/rosegarden-devel
