Thank you so much for this - very helpful and hard to figure out!!
But, man - the fact that it is this convoluted to just record perfect loops from external gear into the 404, is crazy and reminds me of why it has ended up being relegated to just an end-of-chain fx box in my setup mostly 
Tasks that should be seemingly so simple are often buried in that Roland-y haze of difficult UI
That’s because perfect loops are a prison and you’re better off without them 
it is convoluted, but at least it is possible.
still think an extra COUNT IN option = MIDI Start + bpm etc would be most straightforward to achieve this in regular sampling mode.
what are imperfect loops? freedom from the grid, drifting into purgatory.
if i was making stuff standalone in the 404 i’d agree with you, it can really create unique rhythms and lovely movement in the sound to let things run free.
but when your wardens are elektron sequencers then you need the stability.
First off, I’m being deliberately cheeky and apologize for that. I don’t know your use case and this could be crucial to you for all sorts of legit reasons. But also, turns out I want to talk about this so I’m going to lean in using you as a spring board! Forgive me. This isn’t directed at you, per se. Just me rambling.
I’d say stability is also a prison. Or maybe an illusion? And it’s certainly possible to use an OT without it (I’m less familiar with DT or M:S).
This is a super good question, and I don’t know that I have an answer that everyone will agree on, but I’d start by saying a “perfect loop” is “a sound file that is the exact number of samples as the time you want it to play.”
And an “imperfect loop” is anything that’s not that. 
So if I want a 48k loop to play for 4 bars at 120BPM, to be a “perfect loop” it must be exactly 384,000 samples long.
Samples per second = 48,000
Samples per minute = 48,000 × 60 = 2,880,000 S/m
Beats per minute = 120
Samples per beat = 2,880,000 ÷ 120 = 24,000 S/q
Samples per bar = 24,000 × 4 = 96,000 S/b
4 bars = 96,000 × 4 = 384,000
If I start looping something and let it run free, if it’s shorter or longer than exactly 384,000 samples it will eventually go “out of time” with my clock/click track/backing/whatever. This leads to the drifting purgatory of different rhythms playing off of each other that I agree can be quite lovely.
But there’s nothing that says loops have to be left to run free. If you want a loop to play for 4 bars, rather than letting it run free just trigger it at the beginning of every 4 bars. If the loop is a few thousand samples long or short, you’ll never know. If the loop is so long or short that it gets off the beat before you get to the end of your four bars, slice it into four and trigger them at the beginning of each bar. If it’s so, so long or short that it gets off the beat within a bar, split that in 4 and trigger it each beat.
So just different techniques, right? But my point, more over, is that the 404 makes this sort of “trigger and slice” beat matching silly easy. And it makes “perfect loops” style beat matching crazy hard. So maybe just do the easy thing?
Maybe you’d prefer not to and you don’t mind the extra work? That’s cool, but “trigger and slicing” loops is pretty forgiving over reasonable variances in length and BPM. “Perfect looping” needs to be incredibly precise by its nature. So it’s just a more fragile solution — less flexible. And less flexibility in music making usually bites me in the ass. That, and the more flexible a thing, the more happy accidents I find in it.
love this kind of over think.
and agree with everything you’ve said.
i purposefully import my breaks into the machinedrum at different speeds, if they were all locked in at “162bpm” for example. things would get stale quickly, allowing samples to collide when you start repitching creates far more interesting interplaying grooves.
i find the trigger every 4th, or just trigger the rough loop at the beginning of the bar ends you in the same static place as the perfect loop. as you say “you’ll never know.”
but case in point; with the mk2 when i first tried to record loops (which i assumed i was entering the correct bpm for, but as the start/end time wasn’t precise they were imperfect to their bpm) and free trigger them (and other loops) with bpm sync enabled to the bank bpm outside of pattern mode, i assumed everything would be fine, but the drift was inescapable. interesting and cool. but not how i imagined/wanted that to be working.
with this latest approach that is now possible, to have loads of handmade in box loops that you know (beyond the timing of your pad press) are going to sync up and not drift. and even then you could use imperfect loops alongside for lovely organic drift alongside.
extra work maybe, but compared to cranking up ableton and plugging in USB cables it’s an easier and more flexible life . and tobe honest i haven’t gone to town on this yet, it may end up being that midi jitter and other micro variations still cause issues for this approach.
interested to hear if it works out for others.
Very cool. I was planning on taking the SP with me on holiday. Do you know whether I can trigger the recording process via an iPhone or laptop via the USB-C?
Maybe an S-1 would be a good travel companion with the SP. Load up some drums on the SP, a synth and field recordings.
Looking at the S-1, does anybody know whether I can record directly from the S-1 USB C output?
DAMN!!
Just tried this out and it is amazing !!
Feels like a new machine with this workflow.
If only one could start the resample with midi transport to capture things which isn’t starting on the 1.
Shaq and Kobe together. I go audio out to instrument in at the bottom of SP. You can also send midi either way. USBC audio/midi easily to Mac, IPad or iPhone but I don’t think to other external gear because it doesn’t have power.
If someone else gets it to work, please enlighten me. Although the way it works now is fine.
I’m in the habit of fripping almost every evening ~ a peaceful way to interact with the gear~ cheers 
Still no looper function?
different approach for accurate loops is send clock from the mk2 to an external device, and use the blank pad method to resample when pattern is played with count in off.
might be tighter!
though it’s funny to me using different clock sources and hearing subtle changes depending on the direction of travel.
Ooh good find! It seems to still be listening to ext sync after you exit that screen, try it and let me know what you see
ETA: I get this little lock sign on my screen after choosing ext sync … no clue how to turn it off yet that’s just live mode 
yeah, thanks, saw it, not sure if it doesn’t sync the bank before you open shift+11 or i just never noticed. christ, this machine.
Yeah, I think that screen is just telling you that it’s set to EXT SYNC … I remember having trouble getting ext clock working too but now that I’m playing with it I don’t think that doing it (shift + pad 11) is actually doing anything, just a information message.
i just turned off and on.
it seems that EXT bank sync is OFF from a fresh load. it only becomes enabled after shift + 11. then is permanently on.
so you can trigger sampling in bank mode with audio in auto trigger this way, harnessing ext bpm/end snap.
Sounds awesome but what’s “fripping?”
ok, this doesn’t work if your BPMs aren’t whole numbers.
Roland rounds them up.
which is a shame. why with two decimal places of bpm precision would they not translate accurately from external clock to the recorded sample?
ah thank you 
I was referring to “frippertronics”
Robert Fripp/Brian Eno technique of ambient looping with tapes, basically a short, degrading looper … I’m using Ko-Do-Ma
Ah, cool
