Turning a DAW track into a Digitakt live performance (Tutorial / How to)
I shall here describe a workflow that turns a track produced in a DAW (e.g. Ableton) into a Digitakt (DT) live performance. There are two main challenges that have to be solved: the ‘vertical challenge’ and the ‘horizontal challenge’. The vertical challenge is to reduce a larger number of tracks to the eight tracks of the DT. Not only will most DAW-produced tracks will have more percussive elements, there will usually also be textural elements and multiple higher frequency elements (e.g. shakers, rides multiple OHH, CHH etc), which create a richer sonic spectrum. My solution to this challenge is the recording of high frequency elements on top of low frequency tracks and then separately treat those with the DT. The horizontal challenge is the live performance of a DAW arrangement over time. My solution to this challenge involves mutes, FILL and 1ST conditionals. Producing tracks in a DAW there is another horizontal challenge, which is however no problem for Elektron machines: producing variations in time is easy with Elektron machines, thanks to the sequencer with parameter locks and conditionals.
As an example, I shall consider here a DAW track with following individual tracks within the DAW:
1. Sidechain trigger
2. Kick / bassdrum (Kick)
3. Bassline (BL)
4. Noise FX / textual bed
5. Clap
6. Percussion (Perc)
7. Closed Hi Hats (CHH)
8. Rides
9. Shaker
10. Open Hi Hat (OHH)
11. Plucky theme
12. Lead
13. Violin I
14. Violin II
15. Rev Crash
16. Crash
17. Noise Riser
18. Impact
19. Vocal FX
The vertical challenge is to bring those 19 tracks on the DT with 8 audio tracks. The bassline, plucky theme and lead are obvious candidates for tracks that one could play on an external synth, using the DT’s MIDI channels. For the example here, however, I shall consider the DT standalone.
Looking at the frequency spectra of the tracks, one can often find that they are reasonably well separated. In my example, I record the kick, the CHH and Shaker together into one track. I use on the kick track a lowpass filter and on the CHH and Shaker track a highpass filter to better separate the two parts (kick vs CHH+Shaker. In my case I recorded a 1/4 bar and named the file Kick-CHH-Shaker 1/4 bar.wav, which will have a file size of 69 kB. On the DT one can then image automating the higher end of the LP filter using the LFO, to name just one idea for subsequent processing.
In Ableton, I record the DT clips by soloing the tracks of interest and resampling the master track, which then includes effects and mastering in the recording. I then isolate a 1/4 bar (“consolidate” that piece) and export the clip with 48 kHz and mono conversion turned on.
Because the kick has elements in the higher frequencies, the separation from CHH and Shaker is not perfect. Most DAW produced tracks will however have textural, noisy or atmospheric elements that add an important spectral dimension but whose frequencies are quite high up. In my example, I recorded the Bassline together with the Noise FX track. Using light EQing, it is then possible to completely isolate the two elements with the FLTR page of the DT. Through this ‘trick’ have already reduced the required number of audio tracks by two.
Carefully observing the MIDI pattern in my DAW, I noticed that elements in the Perc and Clap tracks do not overlap. I could thus programme these on one DT track. I recorded two percussive hits ‘Perc 1’ and ‘Perc 2’ (each 17 kB) from the percussion track in the DAW and a single clap. (Note that I also record these through the mastering chain in the DAW). On the DT, I then programme this double track as follows: Perc 1 on 3,8,12,15 and Perc 2 on 6,10. Clap with 3:3 condition on 5 and 13, and on page 3 on 5,8,13. Rather than occasionally adding claps to the percussion track, one could also use the FILL condition. More on how I use the FILL condition comes below.
Looking at the arrangement of the track in the DAW, I noticed that in some parts the Ride was combined with the OHH, in other parts it was just the ride. Again, we can realise this with just one DT track. I recorded the ride into a file ‘Ride 1/4 bar’ and place it on trigs 3,7,11,15 using the FILL conditional. Another recording combines the Ride with the OHH (Ride-OHH 174 bar.wav) on trigs 1,9, 5,13 with the NO-FIll conditional.
For the two atmospheric violin elements, the Violin II DAW track occurs in a break. The idea is then again to have a one bar recording of the Violin I track on trig 1 of all four DT pages, using the NO-FILL conditional and a four bar Violin II recording (1,1, MB) with the FILL conditional on trig 2 and using microtiming to move the sample start all the way to the left, on top of trig 1.
The allocation of tracks on the DT looks finally like this:
1. Kick+CHH+Shaker 1/4 bar on trig 1,5 ,9, 13. with NO-FILL. On steps 2,6,10,14 have CHH-Shaker 1/4 bar with microtiming to the left and cond FILL. One page only.
2. BL+Noise 1 bar. trig 1. Four pages.
3. Ride 1/4 bar on 3,7,11,15 (FIll cond). Ride+OHH 1/4 bar auf 1,9, 5,13 with NO-Fill cond.
4. Perc 1 on 3,8,12,15 and Perc 2 on 6,10. Clap with 3:3 condition on 5 and 13, and on page 3 on 5,8,13.
5. Vocal FX 4 bar on trig 1, NOT-1st cond, four pages. Impact 4 bar with 1ST cond on T5, trig 2 shifted to left with microtiming.
6. Violin 1 1 bar on step 1 of each page (NO-FILL condition) and Violin 2 with FILL cond on step 2 of page 1 and moved left with microtiming.
7. Plucky theme
8. Lead
For the Kick track on the DT, I have used two recordings - the kick together with the CHH and Shaker and a recording of just the CHH and Shaker. One could use jus the one recording of Kick+CHH+Shaker and use a low pass filter on the DT during breaks, but I found that using the FILL condition to create short breaks works quite well. The idea is then to place a second sample next to the other and use microtiming to move them on top of each other.
A drop is often combined with some impact or crash element and I added this to the vocal track 5 on the DT using the 1ST conditional. Exiting a break to the main loop of my song, I want to have an impact and this is done by either changing pattern, or by re-loading the current pattern. The impact will then only be heard once, following by the vocal elements that are otherwise on that track.
Regarding the allocation of tracks on the DT, I placed the most important elements, kick and lead, on the outside (track 1 and 8) and have the low frequency bassline next to the kick, then percussive elements and melodic and synths elements near the lead on track 8.
This brings us to the horizontal challenge. A ‘typical’ DAW track for electronic music in the genres of house, tech house and techno, has one core or main loop, which would be all DT tracks playing at the same time, the most rich part of my song. In addition to this core drum loop, there is an intro and outro, which is easily realised by muting tracks on the DT. Looking at many Ableton templates, I noticed that there are typically two types of breaks: ‘percussive breaks’, with melodica elements muted and ‘melodic breaks’ with kick and rhythmic or percussive elements muted. Percussive breaks are used before the core drum loop, sweeping frequencies of the high frequency elements to create a rising tension that leads to the drop. The melodic breaks follow the core loop, with tracks 1,3, and 4 muted on the DT. For the percussive break, I mute track 8, or track 7 and 8. Muting tracks is thus the simplest solution to the horizontal challenge with what I have described here, the entire track is realised in single DT pattern.
My allocation of samples above, allows however for another solution to the horizontal problem. Using the FILL condition, I create a break by just pressing the PAGE button. The kick track is not muted but plays the CHH+Shaker recording. the BL+NoiseFX runs throughout the entire track anyway. Also, rather than muting track 3, using the FILL condition only the Ride is played not the RIDE+OHH. Furthermore, on track 6, the Violin theme is changed with a more dramatic violin stab sample playing during the break.
In total I used the following sample recording to recreate the DAW track with the Digitakt:
1. Kick-CHH-Shaker 1/4 bar, 69 kB
2. CHH-Shaker 1/4 bar, 69 kB
3. Bl-Noise 4 bar, 1,1 MB
4. Ride 1/4 bar, 69 kB
5. Ride-OHH 1/4 bar, 69 kB
6. Perc 1, 17 kB
7. Perc 2, 17 kB
8. Clap 69 kB
9. Vocal FX 4 bar, 1,1 MB
10. Impact 4 bar, 1,1 MB
11. Violin 1 1 bar. 274 kB
12. Violin 2 4 bar, 1,1 MB
13. Pluck 4 bar, 1,1 MB
14. Lead 4 bar, 1,1 MB
For the plucky theme and lead I was lazy. It would be possible to record here only single shots and programme these tracks then individually on the DT. This is therefore not a good example on how to the use the DT but for the purpose of my exercise, this was ok. (In actual fact, these elements could be fed into the DT externally from synths. The same is true for the bassline. Using a 4 bar recording isn’t the best use of the DT and its fantastic sequencer. The goal was however to prove whether it is possible to recreate a DAW-produced track, in principle.
One can now see that it would be quite possible to realise an entire set of tracks from your DAW on the DT and perform them live. The bassline and lead tracks on the DT are ideal candidates to justify GAS related purchases of additional hardware 
Playing pad sounds with a real synths is also possible without loosing focus in the live performance. In my case, I use the Waldorf Blofeld and have ordered the new Roland SE2 to complement the DT. Once Overbridge is there, one could of course also complement the DT with soft synths playing in the DAW but for me the interest in the DT came from the idea to perform my tracks, produced with a DAW, live.
In my example, it was possible to almost completely reproduce the track from Ableton in the DT. I have yet to think of ways to have the riser and crash included as well. One option would of course be to use more pattern for this purpose but the goal was here to keep it as compact as possible.
In summary, the key ideas presented here are solutions to the vertical and horizontal challenge. The former was addressed by recording combined tracks with low and high frequency elements and then use the FILL conditional to play alternative elements on one track, or play a kick with some OHH and shaker, or a bassline with some noise together and play with the elements using the DT filter. The horizontal challenge was solved by concentrating on reproducing the core drum loop first and then realising that there are two types of breaks - a melodic break that follows the core drum loop (all DT tracks playing) with kick and prominent percussive elements muted and a percussive break in which not melodic elements are played and the filter is used to create tension on the high frequency elements so as to lead to the main, core drum loop for the drop.
Sharing is caring and I therefore hope this is useful for some of my fellow Elektronauts
Best
Olaf