Hi!
A short question:
I think I should be able to use sample chains like chopped breaks using the MIDI CC for the sample start parameter.
Any unexpected glitches with this approach? For instance, if I have the sample start CC and the note on event on the same step in my external sequencer, will the Rytm still trigger the right slice?
Thanks!
Anyone?
I figured this would be a hot topic 
I guess not!
how long is a piece of string ?
it maybe depends on your external sequencer i’d suggest
.
i’ve not tried, so can’t help, but i’d be massively surprised if the arrangement didn’t work if the cc was sent ahead of the note on and i’d be less surprised if it worked if the cc message followed - i doubt this is subject to the issues that plague the OT implementation
.
if we were discussing filter freq - then i think it’d play ball both ways - but with the AR in particular, there’s a few params set-in-stone at ‘launch’ and unlike many on the A4 there’s no changing until a retrig - i imagine that samp start is potentially one of those, but as it can be readily modulated anyway, you may be in luck whether the cc is pre or post - easy to test
how long is a piece of string ?
it maybe depends on your external sequencer i’d suggest
.
i’ve not tried, so can’t help, but i’d be massively surprised if the arrangement didn’t work if the cc was sent ahead of the note on and i’d be less surprised if it worked if the cc message followed - i doubt this is subject to the issues that plague the OT implementation
.
if we were discussing filter freq - then i think it’d play ball both ways - but with the AR in particular, there’s a few params set-in-stone at ‘launch’ and unlike many on the A4 there’s no changing until a retrig - i imagine that samp start is potentially one of those, but as it can be readily modulated anyway, you may be in luck whether the cc is pre or post - easy to test[/quote]
I don’t currently have a Rytm, so I thought I’d ask owners.
I’d be sequencing it with a Cirklon.
The reason I ask is that there are some things with the MIDI implementation of the Elektrons that you’d expect to work one way, but they actually don’t. The only fear I have with triggering slices with the Rytm is that I could imagine the Rytm needing the sample start parameter a few milliseconds before the NOTE ON message. The (most straightforward) way the Cirklon works is that the sample start CC would be sent to the Rytm at the same moment as the NOTE ON message.
If memory serves, you can’t modulate sample start after the note on. (I used to own one)
okay - i got a bit into this - big mistake !
.
it strikes me that there is undeniably an issue here
.
to explain it fully is beyond my patience and this is in part because it’s got a bit of shifting sands about it
.
but i’ve sequenced 4 sequential identical and seperate note ons and note offs and i’ve placed a cc in front of the 1st and the 3rd to change the start param
.
depending on the gap between the cc and the associated note you will get differing results
.
it has to be before to work
it has to be circa 100ms before to produce sorta identical pairs
.
if you have a shorter gap between the cc and note - say 70ms, the first note will sound different to second
.
now - it doesn’t seem to be quite that simple - it seems to relate a bit to the sample length and the selection of start positions … yeah, wtf
.
now i’ve tried to do this quite methodically having been thrown by the weird results along the way - but it’s very very odd when you vary the start param of one pair and it changes the other one (all other things the same) !
.
things settle down when the cc is sent way ahead, but even then i can hear audio issues between first and subsequent, so i find it quite hard to pin down exactly what the bug or issue is
,
… and besides i’m frankly quite pissed that many of the bug reports i’ve sent recently have been blanked and i’m increasingly disinclined to waste my energy trying to sort things out - i’d add that the samples i was testing against were perhaps not ideal, they were on the long side, but it actually helped to hear differences at the start and end of the note down, something that would perhaps be lost with a sample chain - it feels like the bogus shift in playback position may be finer than the step size
.
either way, afaict it’s not good news, it’s certainly not consistent and it’s likely buggy too - all of which may be more or less sample dependent to make things trickier - i’ll run it past someone else or set up a more rigorous test sample
Avantronica, thanks. You’ve been a huge help 
this info is plenty to me.
Damn, I wish they were more sequencable from external sequencers.
like a moth to a flame, here i am … !
.
right
for posterity, as it stands w 1.30b
.
If the cc you send is being adjusted a little from current value, you need a smaller amount of time to set up the change - yes, you read that right !
.
so if you want to shift from 0 to 1 you need less time than if you want to shift from 0 to 119
.
i reckon the ballpark figures are 110ms and 190ms
.
i’ve checked this through methodically and i’m in no doubt about the behaviour, the values are merely indicative - it’s strange indeed, but there it is - it doesn’t seem to be file size related - it’s just the start point differential which affects things
.
so with a cc message 200ms ahead of time everything will play predictably nicely
if you use a 100ms interval the 0 to 1 transition may sound okay but the 0 to 119 will not
.
i did this with a 120 wav chain spoken count and the error will be a complete unit, so if you expected the number 0 to play (having been sat at 119) it will play 1 if the cc ms gap is too low, it will play nicely if the gap is right - it is definitely time-gap dependent, 200ms works
.
the other observation i made holds, it will not play a precise slice of the sample at a slice point, it will commence playback in-between slices, so you may hear the second part of a word on the first occasion the note is trigged after the cc, it will play perfectly subsequently
.
this is intriguing as it’s counterintuitive that it should take differing times to get to and from different parts of the sample in a digital context (it’s almost like there was a little guy in there having to run further when the differential is wider) - bizarre !
.
anyway, to answer the question, you need to send the cc about 200ms before the note, that covers all bases, anything under 100ms will definitely be flaky
.
just a bit odd, though not buggy if you accept the need to send well in advance !
What a champ, thanks Avantronica!
And yes… That is VERY odd! Like a playhead scrolling through the sample… Or better yet, that little guy Maybe 1.31 will give him better running shoes.
I’m going to wager that this will not be a priority to fix for them, since CC change of slices doesn’t even work well on the Octatrack
i’d wager it’ll never even be news in gothenburg - it’s a strange quirk, interesting perhaps because of the anecdotal discussion above - but my head has turned to thinking it’s a thing related to percentages or fractions/rounding and possibly sloppy code - i think when i report that the smaller jumps sound right at a lower cc time in advance it’s likely just that the majority of that jump is mostly straddled and sounds okay, whereas for a longer jump, the offset error becomes obvious - maybe this is actually indicative of buggy code, especially as it only afflicts the first hit after the cc is sent - somehow after this is set and played once it plays at the correct slice point
.
either way - there are far greater demonstrable bugs being ignored, so i agree it isn’t going to be a priority issue