Hello everyone.
Since Overbridge seems to be taking its sweet time and since i am very new to the Elektron gear and there are experts here with solutions i would never even accidentally come up with i was
wondering what methods there are of exporting the stems of a project.
Currently i have not done it yet but my first step would be to solo each track and record it into Ableton.
Any other more intelligent ways to do this?
You’ve already found the most intelligent way.
If you’re concerned about stem alignment, add an extra (1) one bar (single bar / 16 step) pattern and play it prior to the start of each track stem using a click, clap or pop sound on step 1.
Something short and sharp will work best and shows up as a highly visible blip in the audio wave. If you use the same sound and same duration each time, it should be easy to align the stems visually or by detecting transients.
Actually, if they record it all in one take, wouldn’t it be easier to just align the first transient of the recording to the grid, then everything will already be aligned (as long as the project tempo matches at the time of recording) to the grid and is just a matter of cutting and arranging to separate tracks in live? No need to spend time aligning stuff to each other and then to the grid… no? 
I never trust gate recording to grab the beginning of the first sound.
Not talking about gate recording just simple old hit record in live and hit play on the DT, after recording you set the first beat from the first transient and you’re good to go as long as your live tempo is the same as your dt tempo
It’s not that simple for some people if I’m understanding what you mean. Doing it this way (tracking out by audio) won’t have the luxury of a mechanical start to the recording so alignment is based on visual clues. So, a snare and a kick (just as an example) have dramatically different waveforms, so do most sounds right?
If you have a sound that only hits on the 3 and it doesn’t have a perfect parallel to another waveform on a different stem which hits on the 3, then you’re lining it up by sight or by ear, what I’m talking about is creating a mechanical way of aligning for time and not just for how it looks.
If each track has a perfectly timed (to the project bpm) identical waveform (ie if a pop hits exactly 4 bars before the start of the one), it becomes much easier to align them, then just move them all as 1 group to where you want the downbeat of the 1 to be. If I’m understanding your method, it’s no easier than what I’m suggesting and lacks the safety net for a “simple user”.
Hmmm my method sounds way simpler in my head maybe than it is? I’ve done it plenty of times before, so I’ll run it down perhaps I can make it clearer:
I have a composition in the dt with a few patterns and stuff…
I set the tempo of my dt project the same on live first.
Then I will solo the kick drum (if it starts on the one, if it doesn’t you have to place one there)
Then record a few bars of that and switch to the next sound and so on until all my tracks are recorded
Then I go to my recording in live and set the first transient which is the kick and that starts on the one
Now I cut in the grid at the start of the bar where the kick ended and my next sound plays and bring that to a new track to the very beginning of the project again. Do this again for every sound
It will all be in sync. I recorded several bars of each pattern so if I didn’t switch sounds in time it’s fine cause I got more to choose from, it’s a matter of editing bars
Does that clear it up?
I think I understand but since I’m more of the “compose full songs on hardware” type and don’t really do any arranging in the daw unless I start in the daw, that may be the difference.
I’ll have like 16 patterns that are all different to some degree so for one thing I record a much longer stem than you do. Also, I don’t use ableton so that could be some part of it, but for me what you’re talking about sounds like more work than aligning the same same pop waveform on all tracks and then dragging all tracks to the same point in time.
This may only be preferable if dealing with arrangements and not an intention to arrange further, but I think it would be just as easy with like 4 bars or whatever, it’s just a different way to do it.
Right I’m an “arrange in the daw always no matter where the song started” type, so that’s gonna be different if you want to keep the exact arrangement, yeah. That’s where OB shines (also shines in my use case too of course)
Well, I’m still hanging in there with og digitakt but if I didn’t already have the mpc live 2 I would have probably been compelled to upgrade my DT to DTII a bit sooner. As it stands, I’m just going to wait for prices to come down to the point where it makes sense.
Overbridge is really pretty impressive, that’s why I understand them not wanting to release a buggy version even as public beta. I think it only makes sense that it will come soon though.
Same here, bruvva
That’s how I track stuff out of the OT sometimes. If it’s something small, I just record internally, but if I’m recording longer stems I use this method to only have to align once. But I also do what @shigginpit mentioned above: place a single drum hit on the 1 and play a couple bars before.
So, essentially I record patterns in sequence onto a single stereo track. The first pattern is a mostly silent pattern, but with a hit on the 1.
One way to track out is to make a utility Song Mode song, pointed to a dummy pattern. Use the mutes to play just track 1 as long as typically needed, then silence as desired, then track 2, etc. Then paste patterns you want to track out into the dummy pattern slot, play the song, record, then segment it into 16. That at least lets you grab a cup off coffee in between. Then however is easiest to slice your audio into 16 even slices and move each to its track in your DAW…
Quantized resampling of each track is also another way to go. Then it’s just a matter of dragging those loops off the DT and in to your DAW. I think a lot of people take this approach with OT stems, it’s just as applicable to DT.
If you don’t care that all sampling / resampling is normalized by DT on the way in, that might be the easiest approach.
Here’s my approach with Syntakt and Live (I don’t use Overbridge so this applies equally to the DT2 or in fact any Elektron device). This may be more detail than the OP needs but I figure other people may be facing the same issue so there’s no harm in a bit of a walkthrough.
Create tracks on Syntakt as normal.
Syntakt is connected to my computer via USB. In Live, create a MIDI track and add an External Instrument from the Instrument selection. In the External Instrument, select Syntakt as the MIDI output device and select the MIDI channel I’ve set up as the Auto MIDI channel in the Syntakt’s settings. Finally, in the External Instrument audio drop down menu select the audio inputs in my audio interface that are connected to the Syntakt’s main outs.
In the ST MIDI clock settings, set MIDI clock to Receive and make sure MIDI+USB is turned on in the MIDI port settings. The ST will now receive MIDI clock and transport from Live.
Create an audio track in Live and select the ST MIDI channel as the audio input. This audio track will print whatever the ST is playing as an audio clip.
This next bit is optional but is worth doing as part of the initial setup. Record 16 bars of a hi hat or snare (anything with a v short attack will do) playing 8th or 16th notes on the ST. Arm the audio track in Live that’s being fed from the ST MIDI track then click on record in a clip and let it run to 16 bars. Stop recording, zoom into the waveform in the clip and work out how much latency there is by looking at the size of the gap between the grid marker and the start of the transient waveform. Live measures this in samples when you zoom right in. FWIW the ST has around 3ms of latency on my system but this may vary for other people. In the External Instrument, dial the latency adjustment into the drop down box. The ST will now be synced to Live (although not perfectly down to sample level… that’s fixed later on).
On the ST, use global or pattern mutes to mute all tracks apart from the one I want to capture. I tend to use Push for this next bit as Push has the option to set a Fixed Length for clip recording that IIRC you can’t do on Live in software. I set Fixed Length to whatever is the length of the track I’m about to record on the ST, usually it’s 4 bars but could be a lot longer with conditional trigs. The key thing is it speeds up the process a lot if the audio clip recorded in Live is exactly the same length as the track being recorded from the ST… there’s no need to trim and crop after the audio recording ends. It works without a Push but just needs a bit more focus after clip recording stops.
Make sure the tempo in Live is the same as the tempo in the ST. I then go through each track on the ST in turn, unmuting it (and muting all other tracks) and capturing the output as an audio clip in Live. Record one clip, mute/unmute ST, select the next slot down in the same audio track in Session view in Live and repeat 11 more times until I’ve captured all 12 tracks from the ST (assuming I’m using all 12).
Final step is to correct any timing lag in each clip. Two ways to do this… either zoom into the audio clip on the computer and pull the whole clip backwards or forwards by a couple of samples until the start of the first waveform aligns exactly with the grid in Live, or (faster and simpler) just hit quantize (cmd+U) and let Live lock the transients to the grid automatically. Change the Warp setting if needed (the default Beats setting usually works well or Tones or Complex/Pro for melodic stuff). Warping doesn’t really alter the sound that much IME because the adjustments to the clip overall are pretty tiny.
Switch to Arrangement view in Live. There is now one track in Live for each ST track, and just drag and drop clips from Session view into Arrangement view to get a full multitrack layout.
This may seem laborious but it isn’t once you get into the flow. It took way longer to type this than it takes me to capture all 12 tracks from ST into Live. Doing this with a Push is particularly good because of the Fixed Length thing plus quantizing audio is one click of a button and each clip is then bang on the grid.
I am not concerned about alignment tbh. I find this easy. I was just hoping to avoid the time consuming way.
But it is what it is for now i guess. Or i would have to wait for Overbridge. Dunno
If your samples are dual mono (both channels identical) and not true stereo, I think you can hard pan two tracks and record one as left audio and one as right audio, then split them and convert them back to dual mono in the daw.
That would cut the recording time in half.
Not sure if that’s an exotic elektron solution, but it’s kind of a typical “tracking out a drum machine” solution.
Thats about the way i do it too.
Was just hoping someone has an exotic elektron solution like „look deep in the buffer zone where the project is stored for 4 seconds. Then using X extraction tool you can access the stems before they self desctruct „ 
But no joy yet
Your solution is also valid.
For me, if the Tempo in Live is the Same as in the DT and you drag over the stems or one stem with a sequence of all tracks , you will only have to find one clear point-one drum hit that has a clear transient and if you set this well to the grid then everything else will be perfectly aligned also.
Its not that hard. However its much easier as long as you dont use complex polymetrics or things like this.
I would love to but ofthen i find i can finish arrangements and make much finer transitions in Ableton.
So i will build 80% of a track on the DT and then finish it (if it is something i want to finish in the first place) in Live. At least that was my plan when i bought the DT2. My brain is simply not the type that can survive and live solely in the DT hardwear. I sometimes really miss what i can do in Live