27 Oct
2007
27 Oct
'07
2:58 p.m.
Hi Dragan,> > hammond is crashing just after the gui appears:
Jack period count mismatch: 512 (bristol -count=1024)>
Hm, I am surprised you get much to work here, you have jack configured to have a 512
period buffer size and bristol is going with 1024. Try "startBristol -b3 -count
512", that should get you much better results. I have discussed how to determine this
correctly from the script, and at the moment I parse ~/.jackdrc, the process table and
finally resort to the jack provided default of 1024. I admit that should not cause seg
fault.
The last proposal I was given to resolve this issue was to start jack from the script and
request the buffer sizes in advance (I need these for some internal initialisation
sequences - long story, the MIDI library must be started before the audio is started since
it is a midi command that starts the emulation. To start MIDI I need to initialise code
sections that use the audio parameters such as sample rate and buffer size, to build the
frequency tables amongst other things). Hence the bristol engine needs to know this
apriori. The option to have a program start the jackd to request its audio parameters,
then have it die, then start it again for bristol from the libraries seems daft - I am
happy to take some of this on advice.
> vox: first there is no sound and after i try to
move some sliders, crash:> (moving the first slider from left side labeled
'16')> midi write error, fd 5, size 1> midi write error, fd 5, size 12>
midi write error, fd 5, size 1> return - no data in buffer> socket closed>
request acked: -1> /usr/local/bin/startBristol: line 311: 21925 Speicherzugriffsfehler
> bristol -rate ${RATE} -count ${COUNT} $*> nowhiskey@murija2:~$
This appears
to be the same issue - at least from the diags.
the third thing i found is a small feature, starting
pro10 emulator, the > balance poti is turned hard-left, so i have sound only in the
left > channel. it would be good if per default the balance is centered here.
I
will look at this as it should be centered and it should have been a memory parameter
since different layers may be panned differently.
Regards,
Nick.
_________________________________________________________________
Connect to the next generation of MSN Messenger
http://imagine-msn.com/messenger/launch80/default.aspx?locale=en-us&sou…