first i'll describe what it is, then i'll give my speculation ..
this is a technique i used to use in Live 4 alot; its a way of getting new loop varations on the fly. example:
- take a 4 bar percussion loop in session; say a fairly complex rimshot pattern with individual hits and rolls. loops switch on.
- play the loop over a basic 4x4 kick/snare pattern
- select the loop braces and CTRL-downarrown to narrow it down to a small value - like a quarter or half note over a small section of rimshot pattern.
- now move the loop braces back and forth (left/right arrow) and make it larger /smaller (CTRL - up/downarrow) as you play your scene, to make rythmic patterns over your base loop
- when you find somehting you like, hit the record button on top of the transport to record this to the arrange grid. if you have set up an arrange loop starting a 1.1.1, ending at 5.1.1, and use the Punchout switch, this will automatically stop your new pattern as a 4 bar length recorded into arrange
- now flip over to Arrange. what you will see is a collection of individual clips, that should exactly represent what you did in session. now if you select all the clips and CTRL-J, you have a brand new loop custom made, played in realtime over your track.
all of the moves that you made should be precisely recorded. ie. when you play it back, it should sound exactly the way you played it in session. consolidating cements it that way
rather .. this is how it worked in 4 and previous versions. but (suprise?) not in 5
the new clip sections are not accurately recorded. some are, some arent. its like that arrange view is not tracking your changes properly and setting the start markers correctly. so that when you play it back .. its alternately a mess.
technique doesnt work anymore
all of this starmarker / arrange-session relationship / timing problems all started with the release of 5 - including this new problem (feature?) in session view where the clips are not firing back at 1.1.1 on transport restart
what changed in 5? a lot of things .. but the most fundamental from what i can see is related to the orange dot / global quant / new scrubbing features.
in other words - it seems like in order to enable the new DJ features, changes were made to the transport / start marker / timing algorithms (sorry im not being so precise but i hope im getting the point acrost)
these changes added some complexity, and a lot of things are happening that were not anticipated. EXAMPLES - losing the instant scrub behavior of the start markers; clips not firing correctly in session; loops not splitting predictably in arrange;
and now this 'correction' of the session grid that is supposedly a feature that slipped thru the cracks, and theyre now getting around to correct, that for all intents and purposes is digging a deeper hole for most people.
are we seeing a pattern here? all of these bugs/odd behavior are related to the same basic function(s) in the program. speculation: in order to enable some new features, changes were made that are breaking the application. this new error being just one more example and i know not the last
also: i think maybe this 'correction' in 5.03 to session seems to me might be another way of trying to adjust and fix some basic problem without scrapping the whole paradigm. because, its way too obviously a radical change that seems like it was "slipped in" without announcement. a little .. desperate?
if so - this is very bad. am i totally off base here, or is there someone that has better insight into how these functions are related that can either back up what i think or disprove it?
.