just wondering if the OT being slaved makes any difference to a sampling recording, as in the quality of the audio or the resulting algorithm effect for timestretching.
Ableton does not like being slaved to a hardware device when trying to record incoming audio. the resulting warp file is all over the place. but maybe the OT is different.
i remember the OT seemed to have no difficulty playing back and resequencing audio loops when slaved to the midi signal from a machinedrum.
If the clock is stable no worries, if its crappy it might have a lot of jitter which could make an audible difference to your timing and therefore recordings
could be an interesting subject for a series of A/B tests
Iām thinking it could be interesting if exploitedā¦ 
haha potentially ā¦
the thing is, although ableton gets confused by being slaved and recording audio, i donāt think it has to be that way. not being critical of abletonās programming, just saying ā¦ if the actual time of a one bar loop is the same even though some slight variance is present in the midi clock, then the OT could easily analyse and properly use the recording made while slaved.
the reason i ask is that the Machinedrum needs to not be slaved in order to have proper Swing control. anyway shall explore the area of interest once the OTmkII arrives.
imho ā¦ sampling will likely happen in the āsamplesā domain, i.e. in the time domain (i.e. sample for X samples or for Y milliseconds)
in either case it will happen linearly between start and end ā¦ the audio wonāt be compromised if played back without timestretch, this issue would be that a jittery clock will lead to playback consequences for which the algorithm then kicks in
if the recording process was capable of adapting to jittery clock then the pickup machines wouldnāt abort overdubbing when slaved
the reason they abort is that thereās divergence of the varying sequencer-clocked playback algorithm from a fixed(constant/linear) samples-based overdub timeline
1 reason not to slave an OT : pickup machines donāt deal too well with jittery external clock, it doesnāt affect quality, it affects functionality
i would imagine pickup machines are a prime example of when not to use the OT in slaved mode, although i donāt quite see the use for them if not as a live looper, and live looping intrinsically implies the OT is going to be master tempo.
possibly there are cool pickup machines techniques i donāt know about tho
is there any way of telling whether the OT playback is behaving properly if slaved to a Machinedrum and not using pickup machines? From my experience it seemed to work great, but that could have been a combination of imagination and wishful thinking.
Iām not sure what the best test might be to explore its working-as-should envelope
Maybe record an accurately looping (i.e. top/tail zero crossings in right phase) 1 bar loop of a sinewave
That wave would play nicely with time stretch off (may loop okay)
Maybe itāll play okay with time stretch on and internal tempo
what will happen if you turn external tempo on ? will the algorithm show artefacts
then, what will happen if the external tempo is deliberately jittery ? will the artefacts be tolerable
try recordings out too, with both internal and external/jittery tempo, then compare externally in an editor or similar
Iāve never really noticed anything (beyond Pickup machines aborts) either, but I slave the OT the least of all Elektrons (partly pickup m/c driven and partly that it feels like more of a hub)
Using a perfectly cycling sine wave is a good idea. I used a 1 bar 256Hz sine wave at 120 bpm to check overdub with Flex.
Errors, clics, latency might be obvious this way.
Iāve had OT as master since I got it in 2011, but with recent addition of DT Iāve been slaving the OT and have noticed no impact on sampling/sample playback, even with timestretch turned on.
When timestretch is on and sample and sequencer tempo are different you hear the artifacts pretty clearlyā¦ I was expecting to hear this with OT slaved due to clock jitter, but this has not been the case.
For more than a year Iāve been slaving OT to the iPad using the Link to MIDI app. Absolutely no problems sampling into OT.
IĀ“m curiousā¦ getting the Squarp Pyramid soon for slaving the OT to it. Will see how the OT reacs, huhu 
Any experience there so far, somebody?
Why Pyramid as Master?
glad youĀ“re asking.
I donĀ“t really come all along with the xox-style sequencing of the OTā¦ it is so powerful, yeah (all the trigs, etc) - but I really want something being able to cope with quintuplets, polyrythmics, you name itā¦
on the other hand I canĀ“t live without the OT for the moment as a swiss army knife in live performance.
I totally understand if you want to use Pyramid as main sequencer, but I think you can use it clock slaved to OT.
Well, brilliant idea. Got to think about that!..
ā¦all realtime stretch algorithms got their flaws if slavedā¦because there is no true solid master clock out there, as long u donāt invest in an dedicated extra external clocking deviceā¦
and the ot has the least trouble in doing soā¦but even there, some little shake in tempo canāt be avoidedā¦
if u always slave the ot, u canāt tell any differenceā¦the stretch algo of the ot is very musicalā¦especially in direct comparism to abletonā¦
if u want to record the otā¦u might do so completely unsyncedā¦and put it in correct sync later onā¦
So true!
About recording, tempo variations change samples lengths, so if you use short loops it be out of sync / tune.
If the OT is master clock, loops are āperfectā. 
The problem I had when slaving the OT to a jittery click was when I was sampling & playing back at the same time. The slightest jitter can stop the audio (by throwing the playback ahead of the recording).
Could you try syncing just the transport and setting the clock manually? I imagine that would be the best of both worlds (and might work from OT into Live too).
Very important for me ! Thanks, itās also good example if you want to explain why slave can be annoying.