On Sat, 11 Dec 2010 21:43:24 +0100 Nick Copeland <nickycopeland(a)hotmail.com> wrote: You're right, I didn't mention system info. I have version 0.60.7 running on Archlinux, standard kernel. Jack is running on laptop built-in audio card with: /usr/bin/jackd -r -t2000 -dalsa -dhw:0 -r44100 -p2048 -n2 -Xseq I found the hanged note stops if I change the midi channel in bristol. Not sure if the note is still there playing when I change back to the original channel, I'll try it next time it hangs. regards, renato

Nick, just updated to 0.60.7 (Gentoo Ebuild), added your suggested cli options to a bristolrc, and all is good so far. Gentoo 64bit, 2.6.33.7-RT. Regards, Alex. -- www.openoctave.org midi-subscribe(a)openoctave.org development-subscribe(a)openoctave.org

The AKS is non functional, it is documented somewhere on the website but I will consider adding a big capital letter disclaimer for the few that are unfinished. If you start it with -libtest you can just see the GUI, I use that during development. The Rhodes is the DX algorithm with a preset. Will have a look at it here. Bristol is not going to be the best Rhodes you will find on open source, It's not even the best Rhodes emulator in bristol! some soundfonts and most samples will give you better results. Regards, nick

Nick Copeland a écrit : Hi Nick, OOPS ! I tell Renato to stay in tune, and run 0.50.1... Will try to compile 0.60 to follow your advice, and let you know ASAP. Thanks for replying, Fred

In release 0.40 bristol would not configure Jack MIDI interface when you used the -jack option. I had not tested this interface that much so when you requested -jack then bristol would still default to ALSA MIDI. Along the way as more testing was done I took the obvious choice to default audio and MIDI to Jack when you give the -jack flag. That means bristol will now appear in the Jack MIDI tab of qjackctl, not in the ALSA tab. Here is how you work around that: startBristol -jack -midi alsa -mini That will give you Jack audio and ALSA MIDI connections. If you want this to be the default behaviour then you can edit your bristolrc and put in one line: vi $HOME/.bristol/bristolrc -jack -midi alsa In my opinion, if you are using bristol 0.60.7 with jack drivers you should also use the following options: -jackstats -count <period size> -rate <samplerate> What this means it that you tell bristol, a priori, what buffer sizes and sample rates it should use to be consistent with Jack. If you leave out these options then bristol will attempt to find the correct rates using a poll to the daemon. It works 99.999% of the time, but when it doesn't then things get a bit ugly: -jack -jackstats -count 256 -rate 48000 -jack -midi alsa This will add these options every time you start bristol. Somebody recently commented that the bristol command line options are a real inspiration killer. No shit! Either bury these options once in the bristolrc or try using monoBristol to launch the app, it will hide a lot of this stuff. Let me know how it goes with the sticky notes and clicks/pops. In 0.60 there is one known case where they still happen: monophonics without note precedence (-nnp) however you have to ask for that to be configured, it is not a default. The next release is intended to have a fix for this last issue. As I said before, I can't rule out you still having the issue, there may be other cases I have not considered but either send reports on this list or send me an email and I will look into them. In one of your other email you asked about Jack Session Management: Bristol has an interface to the Jack Session Manager (JSM) so that you can save application states and recover them in their entirety. I only include the drivers in bristol if I find the correct Jack header files. The Jack dev team will have to comment on which version that would be however last time I checked it was not yet in general release. I would say that the header files are not in the jack version you have installed but it shouldn't give you any problems (unless you were going to use JSM of course). Kind regards, nick. "we have to make sure the old choice [Windows] doesn't disappear”. Jim Wong, president of IT products, Acer Date: Sat, 11 Dec 2010 22:59:49 +0000 From: f.rech(a)yahoo.fr Subject: RE: [LAU] bristol synths hang To: rennabh(a)gmail.com; nickycopeland(a)hotmail.com CC: linux-audio-user(a)lists.linuxaudio.org --- En date de : Sam 11.12.10, Nick Copeland &lt;nickycopeland(a)hotmail.com&gt; a écrit : De: Nick Copeland &lt;nickycopeland(a)hotmail.com&gt; Objet: RE: [LAU] bristol synths hang À: f.rech(a)yahoo.fr, rennabh(a)gmail.com Cc: linux-audio-user(a)lists.linuxaudio.org Date: Samedi 11 décembre 2010, 21h43 Bristol has had these issues although they were supposed to have been iron out in the later releases. The cause of the clicks was largely due to voice reassignment with mono synths fixed by implementing a different note assignment method (note preference with droning) and the sticky notes were due to scheduling issues between threads fixed with a jack ringbuffer message passing method. Now a lot of the distributions are still packaging bristol-0.40 which is a couple of years out of date. If you have newer versions then I can work on some debuging with you, if you don't have 0.60 then it would make sense to try that first. I cannot rule out other issues but lets start with the versions you have and take it from there. Kind regards, nick. Have compile 0.60 with no problem, and no clicks anymore :) But now Bristol lives -alone- in Qjacks'MIDI tab, and the rest of MIDI stuff are in ALSA tab... What did I missed ?