I have tried this experiment. It looks as if the PCs are being sent on the wrong port. The connection for M3-INT-EDS is 131 inside the Midi Manager but kmidimon shows the PCs being sent on port 130. Maybe you can confirm.
BTW apologies - it was Abrolag who mentioned the 'send PC on play new segment' problem, not Michael - my bad. Chuck. On Tue, 2017-07-18 at 21:24 -0400, Ted Felix wrote: > On 07/18/2017 07:32 PM, chuck elliot wrote: > > That has turned up some interesting findings: > > Indeed. > > > under v16.02 selecting [no port] for the relevant M3 bank > > does not send any PCs and nor does switching it back to > > the midi port. I see only ALSA ports being subscribed etc. > > (filtering other midi messages) > > OK. This is to be expected. That feature was added later. > > > This version send PCs whenever a new segment is encountered > > during playback too. > > It is presumably sending PCs on load too so that my M3 is set > > up to start with. > > That behavior was causing problems for some users. But it makes > sense that it was working for you since the channel setup at load wasn't > working with the M3. > > > under v17.12, selecting [no port] for the M3 bank sends out a > > complete set of PCs for the M3 which sets up the M3 nicely! > > such that when it is reconnected to midi it plays all the > > correct instruments (but reconnecting does not send PCs). > > Good to hear. This means we are sending the right stuff at some > point and the M3 is getting it. That reduces the mystery. > > > This version does not send PCs on play it seems. > > It does not. That change was made in r14644. The PCs on play > behavior in 16.02 was causing problems for some users, so it was > removed. What is confusing is that I thought that I had at least > implemented sending PCs for playback from time zero. Apparently I > didn't. So, one must rely on having everything plugged in and ready to > go at load time to ensure a proper setup. That seems a bit cumbersome. > > Sending channel setups (which is what we call this internally) at > playback from time zero would work. We might also add a menu item "Send > Channel Setups" or something similar to let you manually get your synths > in sync with the composition at any time. Just in case power goes out > or something gets unplugged at the wrong time. > > > The one thing that is not clear to me at this point is why > > the 'send PC on load' in this version is succeeding with the > > M1 but not with the M3 ??? > > That's the real mystery. We must be sending something that the M3 > doesn't like in the load case. It wouldn't surprise me at all. The > question is: what? > > I think the test to run would be as follows... First, do the port > switch and connect to kmidimon to see what's going out. Keep a record > of this. I think you can save it from kmidimon maybe. Then modify the > .rg file to connect the M3 to kmidimon's port. Save and reload and it > should show you what it is sending out on load. I can actually do this > test myself since it only requires the .rg file and kmidimon. > > > I note what Michael has said about sending BS/PCs and upsetting > > hardware configurations but it seems pointless to me to have > > instrument settings in a sequencer if you cannot use them to > > configure your whole setup centrally. > > That's the idea. We send them at load anyway, so the assumption is > that we are taking control. By unchecking the checkboxes next to bank > and program, you can configure individual channels to not send BS/PCs at > all. So, if you want to let the instruments have control, you can. > > > For those that don't want or need RG to > > configure their hardware an appropriate send filter ought to > > be able to cater for that no? > > Just uncheck bank and program in the Instrument Parameters box. > > Ted. ------------------------------------------------------------------------------ Check out the vibrant tech community on one of the world's most engaging tech sites, Slashdot.org! http://sdm.link/slashdot _______________________________________________ Rosegarden-user mailing list Rosegarden-user@lists.sourceforge.net - use the link below to unsubscribe https://lists.sourceforge.net/lists/listinfo/rosegarden-user