Producing a full song/album: Which workflow?

Elektron Gear Digitakt
#1

Hello everyone!

I’ve been thinking a lot about how to improve my workflow with elektron boxes these past few weeks (DT + DN).
The goals were these:

  • make the sounds as professional as possible
  • being able to play these patterns live without too much difference

I tested a lot of different techniques that I summarized here:

[1] Live recording, mono-track, just audio:

  • No overbridge: record audio out from devices, apply some compression and done

con(s):

  • can’t do much once recorded…
  • no multi-tracking = no eq per track
  • no way to blend between patterns without “”“hacks”" (= resampling the whole pattern)

[2] Live record (OB), multi-track, “post” arrangement:

  • 4 daw tracks for DN
  • DT. Either:
    • 8 daw tracks
    • 1 track for all the drums (to make the project less flexible but easier to manage)
    • Mix of both, for example: 1 track for kick, others on the main bus
  • FX:
    • Dry DT/DN FX
    • Using only ableton FX (could be midi controlled)
  • Fix recording mistakes in the arrangement view, add automation on tracks, remove parts, etc…

con(s):

  • rely heavy on automation to fix mistake. ie: use amp automation to fix mute not on time
  • can’t really cut parts because it’s one long recorded file per track
  • no dry DT/DN FX per track which means I have to use daw FX if I change song structure/mute tracks

[3] Pattern-record (OB), multi-track, full arrangement in daw:

  • Record a full loop of each pattern without any live tweak, store thoses recordings as one scene.
  • Do all the arrangement in Ableton: Drag & drops patterns from live to arrangement, as usual.
  • Recording the dry DT/DN fx seems maybe a bit more doable this way as I can have the dry FX track per pattern (it will still be in one track tho)

con(s):

  • I’m slowly being sucked by the software
  • No improvisation at all & no happy accident anymore

[4] Midi record, multi-track, full arrangement in daw:

  • Rather than storing the audio, I use my daw to launch pattern through midi
  • I send all the CC from ableton to simulate mute/filter sweep/…
  • Once all of this is done: just press play+record and wait until the song is finished

con(s):

  • Spending good amount of times mapping all the cc, program changes for each tracks. then making all automations in the arrangements view
  • Well at this point I’m not sure it makes much sense to uses those boxes as my daw is now fully controlling the boxes

Voilà. I’ve been testing them all a little bit, but as you can imagine, none of them really suit me right now (the [2] was maybe the least worst). The “playing live”-style one don’t give me a satisfactory result. I can’t handle all thoses mistakes and I tend to not mute/filter sweep how I want it. On the other hand, programing everything ITB just kill the magic of the hardware and my motivation…

The reason for this post is that I’d really like to know about your own workflow, those of you who have produced EPs/albums with Overbridge (DT, DN or other elektron devices)! Are there any other techniques I forgot? What works best for you and why? Please let me know!

Thanks you in advance! :slight_smile:

4 Likes
#2

Let‘s bump this thread!

I would love to hear from you (Elektronauts) as well how you do it. Same fears and problems and thoughts like you do.

If we boil it down, I‘d say the biggest „problem“ is the send fx of the Digis?

1 Like
#3

You are not alone. I’ve been going in circles, myself. Please excuse the rant:

I finally have a DT track I can sorta do live how I want, but just via the stereo out it’s not quite to the level I expect from myself from say completely tracking in Ableton. I’ve spent days trying most of what you’ve mentioned. At this point, I’m in logic with the DT synced, I’m recording the mono tracks with unchanging sends dry via OB; adding reverb and delay sends after. Kick was recorded dry and looped throughout the session just so it’s there. Have begun individually recording tracks that have panning plocked/randomized via LFO and/or tracks with specific send techniques (like a plocked or lfo effected delay/reverb) via the main out by themselves to retain stereo and at times send effects. Things like control all no longer apply in this case. The performance greatly suffers. The DT compressor gets bypassed because tracks are being recording 1-4 at a time, and that hurts the groove. It is incredibly frustrating. It took me days of just thinking about it to try and decide how to even try it. Nothing had worked great yet. I’ve gotten about 1 section tracked. It’s a little cleaner, but also more static. It’s much easier to mix or edit, but the feel and performance suffered. The fill mode becomes useless.

I’m not sure what to offer you, but I was about to write this post myself, so I’m very curious what others are doing. My next attempt might be: kick, basic snare, any tracks that are just one synth/bass with no plocked/lfo panning or sends, all individually, plus one stereo mix group for the hats, samples, percussion, etc. - sending just those that benefit from stereo out the mains along with just their sends. Add sends to individual mono tracks in the DAW. Edit where necessary, dub where necessary. At least this way the foundations are mono and solid and independent and can be mixed and edited and have fx added, while some of the feel and space is retained from several stereo tracks and sends in a stereo mix. That could be done mostly as a live take.

One thing I’ve learned is for a 8 minute song, I record about 20-30 minutes of live stuff, especially if I know I screwed up a part, which sucks having to cut it down afterwards, but…I can’t do a full song live and have it be to the standard I expect from ‘studio recordings…’

Honestly I’m not even sure I’m making sense right now. My brain is scrambled trying to figure this out. And that’s just the DT; this song doesn’t even have my DNK…

2 Likes
#4

Because this is always an underlying goal

I record like this

My setup is MNM, DT, and Kaoss Pad.
Recording in this fashion takes a bit of planning about how you are going to structure things and a lot of mixing/making things sound good within the DT & MNM themselves. I find it very important to pay attention to where my sounds sit and give them space to exist in their respective frequencies. Then i record usually the stereo output of the mixer into a multi-track recorder where it is then bounced to the DAW for a little sprinkling of processing. The processing often includes some light eqing, reverb, and compression. But ultimately its really important to have things sounding good before recording, no amount of processing is going to fix something that sounds bad going in. I will record a couple of takes and choose the one i like best as the final product. I don’t master my own stuff but my guy has never complained about mastering tracks of the stereo pair.

To address the cons:
Its ok that I cannot do much once recorded and there is no eq per track because I work extensively to not need those things once i’m ready to record. This is because of the first thing about wanting to have things sounds mostly the same live and recorded. I do eq slightly through my mixer as i’m recording.
I don’t find the need to use an “hacks” to transitions between patterns. I try to keep a songs about 8 patterns so i get about 2 songs a bank. This allows for a good amount of pattern variation especially with conditional trigs. With 8 patterns a song its easy to write transitional patterns. Additionally, i do not really do a whole lot of filter sweeping with the DT, thats what the Kaoss pad is for and i do not really run into issues with value jumps across patterns.

Recording how i plan to play live helps keep me in practice for gigs, and keeps a consistent sound.

3 Likes
#5

I’ve also flip-flopped between almost all the options in the original post :sweat_smile:

To me, the Digitak’s compressor is great, the reverb sounds excellent (especially pre-compressor) and the reason I bought the thing was to get away from painstaking channel mixing in a DAW.

There are certain bits of control I like to hold onto so I record the stereo with printed effects but seperate the kick and subs/lows via OB usually on one channel (so 1x stereo and 1x mono).

I generally make sure the mix is 90% there in the Digitakt by resampling parts through the DT’s compressor solo or my DAW for EQ and creative effects.

Tracks ar pretty much live-jammed in along with seperate stereo recordings for the Minilogue XD and Tascam 414, creative and corrective edits are done in the DAW and I try to finish as quickly as possible!

3 Likes
#6

I was just trying that. I think I may try to see just how close I can get. I need to rethink (and make sure I’m the process I AM actually thinking ahead) and try to put more similar content on the same track, or things that require similar sends, or panning…and also really nail things in the box. And try to level ‘kits’ across different patterns more than I normally would. Perhaps I can get 90% of the way there ‘live’ with a discrete kick track, and perhaps only when I really want to do something fancy for a recording worry about breaking out any parts. It’s another excuse to utilize my DNK, too. Instead of a track being 8 tracks DT, I should try to incorporate more like 4-5 DT and 2 DN, which could provide some extra easy control.

I briefly tried to setup a copycat delay; reverb, and compressor in the DAW to translate tracks over to discrete without things changing much. Didn’t spend a lot of time on it but it wasn’t as easy as it should be. The DT verb and compressor really ARE good. They just fuckin’ work!

2 Likes
#7

I’ve been working with exactly this problem too. My view is that the best compromise option might be for Elektron to release the Digitakt and Digitone FX as plugins we can use in the DAW and/or somehow part of Overbridge (ie maybe tied to Overbridge for copy protection perhaps).

Because the problem is, as we know, you take away the FX - which is what you need for the true flexibility of muliitrack recording and post-processing - and you’re left with something much less special than the Digitakt/Tone mix output with FX.

So what if we could virtually exactly re-produce those FX in the DAW with Elektron plugins? We’d obviously miss out on the spot P-locked FX but we could keep those in the FX track and those might still be useful.

I really think Elektron should consider this so we can truly fully integrate these boxes into a full-bore DAW production workflow.

1 Like
#8

Well that sure would be something! Also, while we’re dreaming - when recording T1-T8 via OB, since they are mono files, there surely could be a way for each track’s panning (including plocks and LFO pans) to become DAW automation? Or even an update to allow 4 stereo pairs to be configured if 8 is too many to stream? Or trade the 2 inputs for an additional stereo out, perhaps with delay to one, reverb to another? I’m not positive it’s built that way, but certainly not the same size ask as reverb and delay per track; we know that’s not possible without a processor refresh.

There are a lot of potential ways to make things work better. As of now, overbridge destroys the sound of all my sessions. If I start building patterns with OB in mind, or use it as a ‘tape deck’ while building patterns separate from the live use of the device, perhaps it’ll be better. I have no doubt that if utilized in specific constraints, it affords incredible control not available otherwise…but generally, OB is rarely even remotely ‘backwards compatible’ with my sessions.

And for the record I’m honestly not complaining! I truly do appreciate the added features and flexibility, and I understand it’s been a long time coming, and that there may be no ideal way to implement anything much better with the current device offerings…so I’m just brain storming here. But it sure would be a shame if OB ended up rarely used after all that time and development because it worked against the strengths of the machines as a whole; and I do believe that the digitakt offers an experience greater than simply the sum of its parts when played stand alone.

What other whacky ideas are folks coming up with?! :crazy_face:

#9

They’re send effects, so you can always set up your audio channels to only record the FX on the main outs channel. In fact, it’s more flexible, because you can raise or lower the volume of the FX track in a DAW to taste. When the FX are recorded into the tracks themselves, you’re kind of stuck with them.

#10

I think you might have missed the point - the fact that they’re send FX is exactly the problem. You record the tracks + FX to the DAW and, if you want the sound with the FX, you have no flexibility with regard to mixing and mutes/arrangement, let alone track post-processing. Try this and your FX track soon becomes useless. So the point of my plugin suggestion is a less than ideal way of getting around this - ie of replicating the Elektron send FX (and actually the insert/master compressor) in a DAW.

1 Like
#11

This is my basic setup, most times. I has Digitakt only I go stereo outs into Zoom H6. I take these recordings to DAW to tweak, (sometimes trim out a piece I have fluff, sometimes I record a little low and must turn up a bit), but mostly just trim the silences from tops and tails.
I turn into the wind and tack with the strengths of the takt!
I perhaps misunderstand, but I don’t know why you say no way to blends from pattern to pattern? I use many patterns with no problems.
I don’t use compression every track, only when I feel to need it. I set levels by ear as I goes.

#12

Very interesting thread, I´m just on this things now.

First of all, I don´t have experience with this but recently I found me doing your third technique:

It was for me a huge eye opener, and now I´m precisely trying to solve the cons you pointed:

It´s ok for me to use the DAW (Ableton in my case) for arranging and mixing. I like to create a pattern, or several, in the machines, I feel more motivated, creative, etc. you know. But, FOR NOW, I think a good way is record the patterns in Live and continue from there, this way I can use unison in the Digitone, for example (but no CTRL ALL sadly). I don´t use Overbridge ´cause my computer doesn´t like it.

About the second point: I made a track to practice, most of the sound are barely twaked presets (my first one, lacking a lot of things in terms of arrangement, mixing, sound quality, etc.) and what I did is record an “instrument” or sound, this time an arpeggio, and tweak it live. This way it can be the reference for the rest of the sounds. I accomodated the other sounds to what happened with the arpeggio (it was a bit too long), and in this case some sounds are the pattern as it is, others played with the keyboard, etc.

A thing that worked for me was to record the same sound (loop, pattern, notes…) one take after another, in the same clip, let´s say 2 minutes. Then select the bar, 2 bars or whatever I liked and create a loop in the clip. Then duplicate the clip and move the loop to the next take. This way I can have several versions of the same pattern. Ah, I recorded each sound with it´s effect.

The next thing I want to do is the same but, once I have an arrangement, re-record some sounds again, this time not the patterns (loops) but the complete tracks, or at least different takes for different moments, live tweaking them.

I think this way I could find a balance between recorded and live stuff. Only used a bit of delay in Ableton in a specific moment, and reverb in another moment, almost nothing of automation and only edited and reordered the kick and snare. Almost all panning was recorded like that, I didn´t want to do a lot in Ableton really. Then mix and master (the most complicated things for me).

So:

1.- Machines (jam, sound design, idea, draft)
2.- Ableton (record, create different clips and scenes, arrangement)
3.- Machines (re-record any sound tweaked live if necessary)
4- Ableton (mix and master)

This is the track, I think it needs to be shorter, better mixed and compressed, more volume…

3 Likes
#13

I’m currently in the process of tightening up some tracks for recording but I’ve started to notice that some kits are louder than others. I’ve just been boosting and compressing things trying to get them a bit more similar but no I wonder if it wouldn’t be better to fix those types of things at the mixing stage. I want them to be even in the box (in this case hardware) so I can trust my levels won’t be too erratic when I play live.

Any thoughts on this? I worry that some of the tracks may change too much in sound while doing this. Working with the MD/MnM/OT so the compressors are a bit finicky.

Aside from this I’ve gotten a pretty good workflow going, MD stereo feed, using all MnM outs in stereo pairs and a few mono, some of which are routed through the OT. Also have midi from both silver boxes being recorded along with the audio, in case there are some basses or kicks I might want to double up at the mixing stage. I feel lucky to have this much I/O capability, the newer machines would drive me wild! Though maybe I wouldn’t have to think about it so much.

#14

Really nice track!

Coming from the thread „Losing the will“ this really makes me want to try this out and eventually make a finished track – although the workflow seems obvious I somehow never actually tried do it like this.

#15

Thanks!

I don’t know if it is the best way to work, or even the same I’m going to use in the future, but for now it helped me.

I think it’s key to make tracks, even if they are bad, it forces you to solve problems and do the whole process.

#16

This method seems to work for me at the moment…
Just keep recording what I’ve created into ableton basically use all the methods…
Record the mains if want the effects, if I don’t need them record individual tracks.
arrange add more from a machines if needed etc…
Thinking about it too much seems to stop me getting anything done

1 Like
#17

I would normally multitrack with overbridge with edits in post-production, but after hearing Ess’ 0-Day Sysex albums, which are great, I’m definitely gonna try more often recording everything from the analog outputs in one single wav. It has it’s disadvantages of course, but if done right it’s the best way to capture all the work you put in making those patterns in the machine, while also avoid all the meticulous/neverending editing in post.

The good thing is that multitracking and recording a single wav from analog outputs can be done at the same time, haha just in case :stuck_out_tongue:

2 Likes