View Issue Details

Jump to NotesJump to History
IDProjectCategoryView StatusDate SubmittedLast Update
0003356ardourbugspublic2010-07-23 13:152020-04-19 20:14
Reporterdanboid Assigned Tocth103  
PrioritynormalSeveritymajorReproducibilityrandom
Status closedResolutionfixed 
Target Version3.0-beta1 
Summary0003356: MIDI automation is temperamental
DescriptionSorry for continually bringing this problem back up but IMO it is the big, cool new feature in not only A3 but the world of Linux audio so we need to get this right. I won't be so hasty to say MIDI automation is fixed this time round so I'll do some thorough testing before we close this one.

MIDI automation is temperamental, I can't really describe the bug any better than that as sometimes it will be working before you save it then when you reload that session it is often only partially working (or not at all) but I've also had it where I've drawn some automation and then it doesn't fully work UNTIL I've saved it and loaded it.

I've attached an example test session where you should see an automation curve consisting of several peaks but only the first peak actually plays back. I hope this provides a good clue as to whats going on.

The example was created under and this description applies to 7476
TagsNo tags attached.

Activities

2010-07-23 13:15

 

hhh.tar.bz2 (4,380 bytes)

cth103

2010-07-23 15:47

administrator   ~0008607

Can you attach a screenshot of that MIDI automation track as you see it? I'm only seeing one peak, and you mention several.

2010-07-23 16:21

 

hhh - Ardour_001.png (81,638 bytes)   
hhh - Ardour_001.png (81,638 bytes)   

2010-07-24 02:45

 

3356.patch (573 bytes)    
diff --git a/libs/evoral/src/ControlList.cpp b/libs/evoral/src/ControlList.cpp
index bfe4c95..96f8dca 100644
--- a/libs/evoral/src/ControlList.cpp
+++ b/libs/evoral/src/ControlList.cpp
@@ -1040,7 +1040,7 @@ ControlList::rt_safe_earliest_event_linear_unlocked (double start, double end, d
 			return false;
 
 		/* Step is after first */
-		if (range.first == _events.begin() || (*range.first)->when == start) {
+		if (range.first == _events.begin() || (*range.first)->when <= start) {
 			first = *range.first;
 			next = *(++range.first);
 			++_search_cache.range.first;
3356.patch (573 bytes)   

cth103

2010-07-24 02:46

administrator   ~0008615

I think the attached patch will fix this, but it needs some review.

danboid

2010-07-25 12:01

reporter   ~0008634

Last edited: 2010-07-25 12:02

 7489 under Deb squeeze/testing amd64 w/o this patch played hhh as under 7476 ie only the first peak got played but 7489 w/ this patch on the same OS played the automation back fine so I would say you prob commit the patch but lets not close this issue until I've had time to experiment with the automation more under the latest builds?

Thanks Carl!

cth103

2010-07-25 23:20

administrator   ~0008641

Patch applied to SVN. I'll mark this resolved; please re-open if you see similar problems. Thanks!

system

2020-04-19 20:14

developer   ~0022176

Issue has been closed automatically, by Trigger Close Plugin.
Feel free to re-open with additional information if you think the issue is not resolved.

Issue History

Date Modified Username Field Change
2010-07-23 13:15 danboid New Issue
2010-07-23 13:15 danboid File Added: hhh.tar.bz2
2010-07-23 14:12 cth103 cost => 0.00
2010-07-23 14:12 cth103 Target Version => 3.0-beta1
2010-07-23 15:47 cth103 Note Added: 0008607
2010-07-23 15:47 cth103 Status new => feedback
2010-07-23 16:21 danboid File Added: hhh - Ardour_001.png
2010-07-24 02:45 cth103 File Added: 3356.patch
2010-07-24 02:46 cth103 Note Added: 0008615
2010-07-25 12:01 danboid Note Added: 0008634
2010-07-25 12:02 danboid Note Edited: 0008634
2010-07-25 23:20 cth103 Note Added: 0008641
2010-07-25 23:20 cth103 Status feedback => resolved
2010-07-25 23:20 cth103 Resolution open => fixed
2010-07-25 23:20 cth103 Assigned To => cth103
2020-04-19 20:14 system Note Added: 0022176
2020-04-19 20:14 system Status resolved => closed