Re: [Lmuse-developer] Midi sync problems

[<son...@ii...>]

On Wed Feb 11 11:02 , "son...@ii..."  sent:

>
>
>On Wed Feb 11 10:55 , Tim  sent:
>
>>On February 7, 2009 09:06:06 am allan k wrote:
>>> HI
>>>
>>> Thanks for this version.
>>>
>>> Unfortunately it seems that MIDI sync is broken both slave and master
>>> modes....
>>>
>>> Is there any way I can help debug this?
>>>
>>OK, I tested midi sync in and out with my 1986 Ensoniq ESQ1 keyboard.
>>When the keyboard is master and muse is slave, things seem to be
>> working OK, except for a few issues with positioning the muse cursor etc.
>>
>>When muse is master and the keyboard is slave, things aren't so good:
>>If I hit play in muse, the keyboard runs through its song much too fast.
>>By the time muse reaches bar 2, for example, the keyboard is at bar 12,
>> and I hear the keyboard 'race' through its song way too fast.
>>And if I manually position the muse cursor past bar 7, the keyboard
>> responds with an error and resets itself.
>>
>>So I'm looking at the muse sync coding... Complicated stuff!
>>I have to admit I don't know a lot about midi syncing.
>>I did not change anything related to midi sync in all my fixes.
>>
>>What is your experience with the sync issue?
>>Thanks. Tim.
>>)
>
>I had some much more serious issues than you. 
>
>MusE would start in slave mode but the bar and beat position seemed all wrong. My
>external sequencer only works at 96ppq I had changed the resolution down to that
>in MusE but it didn't seem to have an effect.
> 
>When pressing stop, I would lose the cursor altogether and I couldn't restart
>again without restarting MusE altogether. 
>
>In Master Mode, the external sequencer would get the start and stop messages, but
>again the timing between the 2 systems didn't seem to correllate. 
>
>A couple of things about my system. It is 64 bit and uses GCC 4.3. I am building
>another one in 32 bit (Gentoo) over the next couple of days.. again this is with
>GCC 4.3 for core2 optimisations (which seemed to make a big difference on 64
>bit).. However if this is going to cause compatibility issues let me know and
>I'll rollback the system to gcc 4.1 and generic i686 optimisation. 
>
>I also conducted these tests using the emagic AMT-8 and Unitor 8 racks I have. I
>am going to try another re-test with the MIDI ports on the Delta 1010 (unload
>snd-usb-audio and patch a cord from the front bay on one of the emagics to the
>Delta) this is just to see if latency over USB is causing further issues
>considering your results were a bit better than mine. 
>
>Are there any debug settings (at compile time or otherwise) that I can enable to
>help debug this issue??
>
>cheers
>
>Allan  
>
Postponing the 32 bit install till the weekend and will most likely be Arch linux
(I donwloaded the iso last night Geoff!!) . However I did unplug the usb cable
from the emagic units and try patching a cable to the input of the delta 1010...
No good same results..