[quote="greg"]This is a shot in the dark but are there any macro's or GPI loops tied to the timeline or header keyframe?[/quote] Yes, I use boxes that require 7 keyers, so I have a Macro built into a 1 keyframe E-Mem on ME3 recalling an E-Mem on ME 2. (We use tons of multiple video box set-ups.) It just makes it so I dont have to recall the same number in 2 places. (I need the master emem for 3 other things, and I can't seem to get ME3 to control ME2. Its in Suite Prefs, but not functioning.) Nothing is left over from previous effects though. The only 2 levels lit on the master emem are Misc. 2 and 5. (GPI and DDR) I use GPI loopback to run DDR transitions. And, some of my loopbacks trigger macros. So 2 different macros would be running at or near the same time, but not on the same part of the switcher. The problem happens even when no macros are involved. I have a different set-up where I simply cut cameras on ME3, while a timeline is running with no macros or GPI. Same problem. It is mostly as mtiffee described. The keyers on PGM turn on and off. In my set-up, a GPI loopback would trigger trans a keyer on (DDR), then the glitch would turn it off, and then the 'off' trigger would actually leave the keyer on after the timeline ended.

I've seen this before. Are you seeing like a 1 or 2 frame flash where the keyer cuts off/on on PGM/PST? On the NFL Network shows all of the EVS packages had front ends that choma-keyed on. I would trigger the roll of the EVS and at a predetermined time, the EVS front end would cut to the actual package. Over that cut, I would key a transition, so the front end would appear build itself on the screen, then flash off to the package. I would recall an emem on ME-2 that recalled the EVS clean, then copy that to ME-3. (don't ask) The entire front end sequence was keyed over ME-3 and when the final transition was complete ME-3 was revealed with EVS clean. The keying of the front end and the EVS trigger was built into a timeline, but the operations underneath the effect was built into the macro used to run the timeline. Randomly, like described above, the keyers on PGM/PST would cut off then off, maybe a two frames - you would also see the green "on" light cut off. It didn't happen all the time and it seemed to happen at different places in the effect, although it did seem to happen around the time I would recall an EMEM on ME-2 and/or the function to copy me2 to me3. I worked around this by temporarily cutting to EVS on PGM/PST and losing the key before it had a chance to lose itself.

Have you resolved this? I just tried it in the GVG test lab and I could nor reproduce anything like this. I have also done similar operations while TDing NFL, NCAA Hoops, MLB etc. and have not seen this. I think Greg may be on the right track - did you have any macros or GPIs that could have been left over from a previous effect thus interacting with your current effect? Could it have been something as simple as M/E-3 was also delegated on the Master E-mem panel while you were running your PGM/PST timeline?

Way back when, pre-version 6 or 7 there was a bug where if you ran a timeline, say a DVE effect (Chyron fly for example) with a pause in it and you then recalled an E-Mem in a ME the master timeline would rewind. I wonder if your issue is an artifact of that old issue.

This is a shot in the dark but are there any macro's or GPI loops tied to the timeline or header keyframe?