MIDI pitchbend input response -- bug?

[Chris Brown-10]

Hi -- 

Don't know if this has already been discussed, but after I upgraded to the 6/02/08 binary I am now getting double responses to a single MIDI pitch bend input -- looks like they are in lo byte hi byte pairs, maybe to take advantage of higher-res controllers, but can't these be combined to return a single value in higher res?  It's breaking old code ... here's the printout of MIDIIn.bend quick start:

[ 0, 4576 ]

[ 0, 20579 ]

[ 0, 3744 ]

[ 0, 20573 ]

[ 0, 2848 ]

[ 0, 20566 ]

[ 0, 1792 ]

[ 0, 20558 ]

[ 0, 896 ]

[ 0, 20551 ]

[ 0, 224 ]

[ 0, 20545 ]

[ 0, 0 ]

[ 0, 20544 ]

......... etc

Chris Brown

Mills College Music Dept.

Center for Contemporary Music (CCM)

http://www.cbmuse.com

[thelych]

hi all,

oops.... made an typo when i made the sysex patch. That should solve the MIDIBend problem (bad values + duplication)....

i'll look further later if it does not fix it (quick patch for now)

sorry for that

tell me if it is ok

best

charles

Le 13 juil. 08 à 08:09, Chris Brown a écrit :

Hi -- 

Don't know if this has already been discussed, but after I upgraded to the 6/02/08 binary I am now getting double responses to a single MIDI pitch bend input -- looks like they are in lo byte hi byte pairs, maybe to take advantage of higher-res controllers, but can't these be combined to return a single value in higher res?  It's breaking old code ... here's the printout of MIDIIn.bend quick start:

[ 0, 4576 ]

[ 0, 20579 ]

[ 0, 3744 ]

[ 0, 20573 ]

[ 0, 2848 ]

[ 0, 20566 ]

[ 0, 1792 ]

[ 0, 20558 ]

[ 0, 896 ]

[ 0, 20551 ]

[ 0, 224 ]

[ 0, 20545 ]

[ 0, 0 ]

[ 0, 20544 ]

......... etc

Chris Brown

Mills College Music Dept.

Center for Contemporary Music (CCM)

http://www.cbmuse.com

[James Harkins-2]

I tried this out and it appears to fix the problem for me. Checked in now as revision 7681.

hjh

On Jul 13, 2008, at 5:05 AM, thelych@... wrote:

hi all,

oops.... made an typo when i made the sysex patch. That should solve the MIDIBend problem (bad values + duplication)....

i'll look further later if it does not fix it (quick patch for now)

: H. James Harkins

: jamshark70@...

: http://www.dewdrop-world.net

.::!:.:.......:.::........:..!.::.::...:..:...:.:.:.:..:

"Come said the Muse,

Sing me a song no poet has yet chanted,

Sing me the universal."  -- Whitman

[thelych]

thx for the commit James :)

I can narrow it down to the patch for sysex (rev 7528). No time to look closely at it now. I don't know why a sysex fix would affect pitch bend...

( just as info to explain )

i looked back at the source code before the sysex fix.

- i was wrong, seems like the bug was already in the stable 3.2 build, even if it does not "pop up" so badly in this version of sc. It misses additional midi messages in the same packet after a pitch bend message. but can lead to unexpected behaviour / corrupted data in some cases.

- after the sysex patch, the "running status" (added too with the patch) catch the invalid data and outputs the following bad values in sc as unwanted-repetitions of strange pitch bend messages.

voila... should be now both fixed

seems ok for me too

is it ok for you Chris after patch ?x

charles

Le 13 juil. 08 à 15:51, James Harkins a écrit :

I tried this out and it appears to fix the problem for me. Checked in now as revision 7681.

hjh

On Jul 13, 2008, at 5:05 AM, thelych@... wrote:

hi all,

oops.... made an typo when i made the sysex patch. That should solve the MIDIBend problem (bad values + duplication)....

i'll look further later if it does not fix it (quick patch for now)

: H. James Harkins

: jamshark70@...

: http://www.dewdrop-world.net

.::!:.:.......:.::........:..!.::.::...:..:...:.:.:.:..:

"Come said the Muse,

Sing me a song no poet has yet chanted,

Sing me the universal."  -- Whitman