Jump to content
Sign in to follow this  
The Magic Autist

Elektron Octatrack MKII

Recommended Posts

 

 

Fuck it, just a straight FM/PM machine. Let it use incoming audio as well as delay lines as modulators in the algorithms. Make it cute and tiny like the digitakt.

 

Physical modelling in hardware really has been dead in the water for a long time, I would love to see something like this.  Just a single polyphonic part with some kind of modular (or at least semi-modular), with audio input as an operator (so that you could choose where to inject external audio in the algorithm you built, rather than being limited to a fixed point in it).  maybe a couple velocity sensitive pads on it or something.

 

Basically something that was to Tassman what the OT is to Live (i.e. not actually the same at all but similar enough to make superficial comparisons).

 

Yeah this is a cool idea actually. I had an idea for a tracker that the Octatrack ended up being really similar to, minus a few things, one of which was this idea of being able to use arbitrary signals as modulators. Maybe I will revisit this. 

 

 

 

I actually had meant to say something about a physical modeling engine with audio OUTS that could be inserted at arbitrary points in your patch being potentially amazing, in that you could tap audio at some point in the patch, run it through whatever external effects you wanted, and feed it back in at another point.  I'm sure there would be potential for a lot of issues with latency and phase, but the potential for sound design seems so massive that it would more than compensate for that.

 

If Axoloti had more i/o you could probably patch up something like this (and I guess even as is you could if you limited yourself to a mono output and use the other output channel as your send), but that's hardly the same as this fantasy Elektron box we're talking about.

 

 

EDIT: honestly, just a branch of the OT operating system that replaced the machines, effects and sampling engine with some high level physical modelling building blocks like membranes and tubes and things like that, so say every track could have an excitation source at the point in the OT where the machines go now, would already be incredible.  But it's not like Elektron would ever have any incentive to fork the OS on an existing piece of hardware in house, or open the source up or anything.

 

But just imagine.

Edited by RSP

Share this post


Link to post
Share on other sites

Get an Analog Four and control Mutable Instruments Elements imo

Share this post


Link to post
Share on other sites

Yeah, digital delay lie module with the sort of feedback RSP describes has so much opportunity for gathering grit and warmth in the context of a modular system. That's probably the way to go here. The A4 would probably add some of that turn-on-a-dime Elektron magic.

 

I'm wondering if you can't already do this to some degree with the OT just by sending a thru machine with one of the delay effects on it into the cue outs, and running that through external processing and back into the OT ins. You'd have a minimum latency just by virtue of going A/D -> OT engine -> D/A and back through, but the reciprocal of that delay might be a high enough pitch that it'd give you a decent range. I'm guessing it's less than 10ms/100Hz which would make it usable at least for bass.

Share this post


Link to post
Share on other sites

After I finish moving I was already planning to spend a lot of time experimenting with the cue ouputs on the OT being fed through the Axoloti and back into the inputs, and this is just giving me more ideas. 

 

But yeah, thinking about it now I bet some basic Karplus-strong stuff could be done with the OT alone, I'm going to have to mess with that this afternoon.

Share this post


Link to post
Share on other sites

But yeah, thinking about it now I bet some basic Karplus-strong stuff could be done with the OT alone, I'm going to have to mess with that this afternoon.

Fuck yeah! Please record & share :)

Share this post


Link to post
Share on other sites

Well that was easy, but not as exciting as I thought.

 

EDIT: tl;dr it works and has some real sound design potential but isn't very flexible and can't really play melodies or anything because of the coarseness of the OT's delay time.

 

 

 

You definitely can make a simple karplus-strong algorithm with two tracks.  The down side is, the delay time isn't nearly short enough or fine enough to make it useful the way I set it up, although you might be able to get something a lot more interesting by making your own delay with two flex machines sharing a single record buffer, I haven't messed with recording and playing back from the same buffer at the same time yet so I don't know enough about it to say.

 

But at the very least, I was able to patch up a really simple KS proof of concept fast.  I used tracks 1 and 5 but you could use any of them obviously:

 

 

Cue cfg set to studio. Cue out left patched to input A

 

Track 1 = flex machine with a short white noise sample generated in the OT (did the whole "record an input with nothing patched to it and then normalize" trick).  Attack, hold and release set to minimum.  Balance hard left. No effects. Main output level set to 0, cue set to 127.

 

Track 5 = Thru machine, INAB set to A and volume turned all the way up. Balance set hard left. No effect in slot A, delay in slot B with time and feedback set to minimum, send and volume set to max, Tape off, DIR minimum, sync on (even the minimum delay time is too long with sync off), lock off, pass off.

 

That's it.  You can fine tune the width parameter in the delay and the volume parameter in the amp page for the track with the delay line to get it sounding how you want, but there's really no way to control the pitch.  You just get one note of a kind of basic K-S slap bass sort of sound.  If you turn the tempo all the way up to 300 you can get a few notes out of it, the useful range seems to be with delay times between 1-12 or so with the tempo on 300. Definitely not flexible but enough to play with and resample at least. I've been getting some mildly interesting stuff by using the LFO designer to modulate the delay time across the useful range.  The attack and release times on the AMP page of the flex track definitely have some potential, too, as does the BASE parameter on the delay.  Definitely interesting, if kind of a one-trick pony.

 

And obviously, instead of using the filter built into the delay you could pass that track to a neighbor track and use the actual filter on it for more control.

 

You also need to set up a third track as a thru machine listening to input A with the Cue level set to 0 for monitoring.

Edited by RSP

Share this post


Link to post
Share on other sites

Here's a completely nonmusical demo, just one 16 bar loop with a trig on the downbeat on track 1, at 300bpm.  Randomly tweaking the attack and release on track 1 and the delay parameters on track 5. Just a bit of LFO modulation of the delay time near the end, but not much.

 

https://www.dropbox.com/s/oewhveg31j8miys/OT%20Karplus-Strong%20test%20NORMALIZED%20CDDA.wav?dl=0

 

I think this DOES have some potential for synthesizing drum sounds completely in the OT for resampling.

 

EDIT: I recommend listening to this through headphones or monitors, because most of the action is in the deep bass. Through laptop speakers it mostly sounds like a metronome click.

Edited by RSP

Share this post


Link to post
Share on other sites

I'm going to rebuild it from scratch tomorrow, because in the process of messing around with it I seem to have unearthed some OT bug where one of the tracks I wasn't using was picking up the cue and routing it to the output even with all of its input sources turned off, so I had to just turn its volume all the way down and work around it.  I'll patch up a cleaner version tomorrow and share the actual project files with some more coherent documentation.

 

If anyone with more OT experience can come up with a more flexible approach to this it would be amazing.  The sticking point is the range and coarseness of the delay time in the OT's delay, and I think the answer is going to lie somewhere in creative misuse of record buffers and/or pickup machines to get much shorter delay times with finer resolution than what you can get from the delay effect.

 

Even as-is there is a lot of potential here.  The example I posted is just proof of concept, but using p-locks and the crossfader, using different samples (or external audio) instead of plain white noise, using neighbor machines after the delay to do more processing before feeding it back, and actually having a rhythmic sequence of some kind would open up all sorts of possibilities.  It's also pretty limiting that right now you have to work at 300bpm just to get any kind of useful range out of the delay time.

Edited by RSP

Share this post


Link to post
Share on other sites
Guest Chesney

There are some KS waves in the blofeld, that is fun being controlled by octa or MnM. 

Share this post


Link to post
Share on other sites

When I first got the Axoloti I mostly used it for KS stuff, it's a lot of fun when you can really implement it correctly, although I think I ended up with around 20 parameters and I'd usually end up spending a bunch of time tweaking them and getting different sounds and not actually ending up recording anything.  This was mainly an experiment to see if it would even work, but I think I will actually use it for something sooner or later.

Share this post


Link to post
Share on other sites

That sounds pretty cool to me! Definitely hearing some potential here and this gives me some other thoughts. As far as calendar months I have a decent amount of OT experience (got mine in late 2012) but I'm also a lazy dilettante wanker so take my ideas with a grain of salt. Anyway here they are:

- Use chorus, flange, or comb filter instead of the delay itself, and set at least one of them to 100% wet. This should give you a bit more precision with delay times, with the tradeoff of having shorter delay times.

- Layer multiple delay/chorus/flange/comb on one track, or spread across neighbors, to add delay times.

- Use a synth (or whatever) another one of the inputs as the source instead of a noise pulse. I wonder how tricky it would be to set this up, though. If you don't mind working in mono, you could use source as one stereo channel and feedback as the other side, and use the spatializer's mid/side feature to fade between them, so you get a wet/feedback control. I'm probably not explaining this well but this video might clarify a bit:



Do you use the OT with any external controllers? If so, or you're interested, PM me and let's talk, because I have some ideas that might make this more usable...

Share this post


Link to post
Share on other sites

That sounds pretty cool to me! Definitely hearing some potential here and this gives me some other thoughts. As far as calendar months I have a decent amount of OT experience (got mine in late 2012) but I'm also a lazy dilettante wanker so take my ideas with a grain of salt. Anyway here they are:

- Use chorus, flange, or comb filter instead of the delay itself, and set at least one of them to 100% wet. This should give you a bit more precision with delay times, with the tradeoff of having shorter delay times.

- Layer multiple delay/chorus/flange/comb on one track, or spread across neighbors, to add delay times.

- Use a synth (or whatever) another one of the inputs as the source instead of a noise pulse. I wonder how tricky it would be to set this up, though. If you don't mind working in mono, you could use source as one stereo channel and feedback as the other side, and use the spatializer's mid/side feature to fade between them, so you get a wet/feedback control. I'm probably not explaining this well but this video might clarify a bit:

 

Do you use the OT with any external controllers? If so, or you're interested, PM me and let's talk, because I have some ideas that might make this more usable...

 

Yeah, I thought of using chorus and flanger about an hour after my last post yesterday, going to try it later today.  I don't actually know if 100% wed on the OT means "delayed signal only" (in which case the strongest chorus/flange effect would be at 50% wet and anything beyond that would push it into vibrato with a couple MS of delay) or 100% wet means "deepest chorus/flange effect" in which case it would be a 50/50 balance of the dry signal and the delay signal and wouldn't be useful here' I've only been using this thing for a few months and I rarely use the chorus an flange (and when It do it's subtle). But I'll try it and find out.

 

I'm definitely going to mess with other impulses and processing external signals and stuff at some point, even filtering the white noise makes a big difference int he timbre you end up with (I didn't play with that on the OT yet but in software and on the Axoloti it was always one of the biggest changes you could make - maybe bigger than the filter settings on the delay line, even).  For now I want to keep things simple and see if the basic concept is really going to be flexible enough to be useful beyond making some weird sounds and sampling them. But yeah, definitely using external signals is one of the most interesting possibilities.

Share this post


Link to post
Share on other sites

So, chorus wasn't that useful but flanger works really well.  Here's some only slightly more musical messing around with a simple 16 step loop (as before, a lot of low bass action in this that doesn't come through laptop speakers at all):

 

https://www.dropbox.com/s/orx6w7n941rtb4v/OT%20Karplus-Strong%20test%202.wav?dl=0

 

Track 1 is another noise impulse made by sampling the OTs self-noise just like last time, loaded into a flex machine with similar paramaters to last time (attack, hold and release are all at miminum as a starting point).  Only difference this time is that I added a filter. Track 1 is routed to the left cue output only (main level is zero, cue is around 110, amp balance is hard left)

 

Track 2 is a thru machine listening to the cue L being fed back into input A with he volume all the way up on the playback page.    This time instead of using a delay, I used a flanger with depth, speed and width and feedback set to zero, and mix set to maximum, followed by a filter. On the Amp page, everything is left at the defaults except I used the volume together with the filter to dial in the point where it rings but doesn't feed back uncontrollably - for me this was volume around +13 and filter width around 105 with the LP set to 24db.

 

I also realized that by using a third track to monitor the audio before I was getting the impulse blended in with the actual KS sound, so this time I only used two tracks, with track 1 only routed to cue and panned hard left, track 2 routed to cue and main and panned center, and the headphone mix panned to main only in the mixer (obviously that only mattered because I was working on headphones but it would be easy to overlook). The down side of this method is that you have to pan the resonator track to the center if you want to hear the sound centered, and that means it's going to both cue outputs, so if you wanted to patch a second K-S algorithm and use both at the same time you would have to hard pan one left and the other right, otherwise you would have crosstalk between them (which opens up a whole other level of stuff to mess with - using two different KS algorithms in parallel and playing with the balance of the resonator tracks in both of them to get cross-feedback between the two resonators is probably going to sound ridiculous in a good way, and I'm going to try it later this week, but isn't something you'd always want), but it's still better than monitoring the cue and having the dry impulse blended in all the time like in the first recording I uploaded.

 

As far as creating the loop, I used trigs an plocks on both tracks to trigger the impulse and adjust parameters - there's a LOT of possibility here since you have independent sequencer tracks for the impulse (track 1, the flex machine + filter) and the resonator (track 2, the thru machine with the flanger + filter in a feedback loop).  I hardly scratched the surface.  Using different sequence lengths for the two tracks is going to be a lot of fun.  The only real time tweaking I'm doing here is that the LFOs on the resonator track are set to modulate the delay time of the flanger and the base and width of the filter, and I've got a few scenes set up on the crossfader with different speed and depth values and am messing with them.  All of the other control is plocked and static.

 

 

Anyhow, this is WAY more useful than doing it with the delay like I did before.  I've got some errands to do, but when I get back I'm going to patch myself up a couple of preset K-S projects, one exactly like what I sued here and another with a second K-S algorithm for experimenting with cross-feedback like I mentioned above  I'll upload them this evening.

Edited by RSP

Share this post


Link to post
Share on other sites

Ok, here are two basic K-S projects:

 

https://www.dropbox.com/s/b3k8w7bot4irokc/KS_Template01.zip?dl=0

 

Bare-bones K-S algorithm on tracks 1 and 2.  Load a short sample of some kind into the flex machine on track 1, patch the left cue output straight to input 1 and you should be good to go.

 

 

https://www.dropbox.com/s/6ik54n8suqezm8b/KS_Template_Xfbk01.zip?dl=0

 

Dual K-S with cross-feedback.  Same as the above, but with a second K-S algorithm on tracks 5 and 6.  Patch cue L to input 1 and cue R to input 2. The balance control on the amp pages of tracks 2 and 6 control the stereo width and cross-feedback (which are inseparable in this setup).  Bey default, 2 is panned hard left and 6 is panned hard right.  As long as they're both hard-panned to opposite sides, they're completely independent, but as soon as you start panning one toward the center, its feedback loop starts to also be fed into the other, and that opens up a lot o possibilities but also gets out of control easily an will turn into howling feedback, so its better suited for plocks than for controlling with the crossfader.  In general, the more centered they are the more likely it is to get unstable and start howling, but you can keep it under control with the volume on the amp page.  I think a good enhancement would be to patch the feedback loops through some kid  of analog limiter to clamp down on it when it starts to feed back, and in general you could patch pretty much any outboard gear you wanted into the feedback loop.  Also, neighbor machines could be used to add more processing inside the feedback loop but remember, anything that works in stereo is going to be fed back into both K-S patches!  I think ping pong delays have a lot of potential here.

 

The second efect slot on the impulse tracks is also free, and there's no reason not to use that, too, or to swap out the filter in slot 1 for something else, or to replace the flex machine with another thru machine and use an external signal as the impulse source.

 

That's all I have time for right now, but in late August, once I'm done moving and getting my studio set up again, I'll experiment some more and then do an in-depth video tutorial about it.

Edited by RSP

Share this post


Link to post
Share on other sites

Sweet! I like what I'm hearing, I bet you could get some Autechre-ish stuff out of this...

Share this post


Link to post
Share on other sites

Sweet! I like what I'm hearing, I bet you could get some Autechre-ish stuff out of this...

Yeah, I think this has some real potential.  Once I've moved I'm going to really dive in to it, and play around with stuff like putting other stuff in the feedback loop.  I've got an Akai MFC42 that would probably work really well in there, especially in the stereo project I uploaded.

 

The big thing is probably going to be just really spending time on the sequencing, though, there's a whole lot of stuff I was getting out of this that I didn't record, but it's not the kind of thing that's very good for tweaking live because even the simpler two track setup has parameters spread across 4-5 different pages on two tracks that all interact with each other, so it's really more suited to deliberately composed sequences than random tweaking like I was doing.

Edited by RSP

Share this post


Link to post
Share on other sites

Interesting ideas here, I'm tempted to get something similar going on the nord. I've been getting into using interesting feedback loops for filter resonance lately, but never thought to use outputs and inputs to introduce external processing..

Share this post


Link to post
Share on other sites

I mentioned this to a friend of mine this afternoon and he said he used to do K-S stuff with two or three Electribes feeding in to each other before he got rid of all of his hardware and went 100% Live/Max, but didn't elaborate.

Edited by RSP

Share this post


Link to post
Share on other sites

Oooh, I just thought of something I'll have to try as soon as I can:  even though there are only two OUTPUTS for feedback loops possible on the OT, I could definitely run them both in to Headrush pedals (I've managed to get two of them over the years for $40 total), send the dry signals to inputs A and B and the delayed signals to C and D, so two impulse tracks and feedback loops could actually be fed into four resonator tracks.

Share this post


Link to post
Share on other sites

I mentioned this to a friend of mine this afternoon and he said he used to do K-S stuff with two or three Electribes feeding in to each other before he got rid of all of his hardware and went 100% Live/Max, but didn't elaborate.

ha yeah, I've seen a few friends do this, I think they get to a point where they want more control over everything. A friend of mine recently sold his entire modular setup and went back to puredata running on a Surface Pro + a bunch of USB controllers, and has succeeded in replicating the sound he used to get with the modular, but with much more flexibility. I think of doing this as well, but tbh I make better music within restrictions than when I can just get carried away and do whatever I want. The nord modular is close to being too much flexibility, but its limited processing power (and lack of MIDI out modules in the G1) forces me to be a bit more creative.

Share this post


Link to post
Share on other sites

 

I mentioned this to a friend of mine this afternoon and he said he used to do K-S stuff with two or three Electribes feeding in to each other before he got rid of all of his hardware and went 100% Live/Max, but didn't elaborate.

ha yeah, I've seen a few friends do this, I think they get to a point where they want more control over everything. A friend of mine recently sold his entire modular setup and went back to puredata running on a Surface Pro + a bunch of USB controllers, and has succeeded in replicating the sound he used to get with the modular, but with much more flexibility. I think of doing this as well, but tbh I make better music within restrictions than when I can just get carried away and do whatever I want. The nord modular is close to being too much flexibility, but its limited processing power (and lack of MIDI out modules in the G1) forces me to be a bit more creative.

 

 

Yeah, same here.  I spent the last few years hoarding 90s digital rack gear but now that I'm moving my whole goal is to just choose three or four things (tops) and spend maybe a month just working with them and making a bunch of tracks and then moving to another combination of three or four things.  Maybe always using the OT as a kind of hub (if only because it would make it easy to take whatever I was working on and adapt it for playing live), but maybe not.  I feel like the best stuff I do is when I don't even do much sequencing or looping or anything and just overdub stuff with as few takes as I can manage.  I'm not a keyboardist by any stretch of the imagination, so if I have to track everything live it keeps me from getting overcomplicated.

 

I wish I liked working ITB more because it would save a lot of space and money. I tried for a long time but it just isn't for me

Edited by RSP

Share this post


Link to post
Share on other sites

Yeah, I'm starting to get into making tracks that are somewhat "automatically sequenced", mostly in the nord, by using LFOs for both timing and sequencing. It's an interesting thread of generative sequencing that I'm finding more exciting and inspiring than I inititally expected, plus it really suits minimalism very well :D

 

Ha, I've got a bunch of 90s rack gear as well. I rarely use any of it in my tracks, but can't bear to get rid of it either. I'm convinced that someday I'll make an entire album using just the JV1010 :P

Share this post


Link to post
Share on other sites

Yeah, I'm starting to get into making tracks that are somewhat "automatically sequenced", mostly in the nord, by using LFOs for both timing and sequencing. It's an interesting thread of generative sequencing that I'm finding more exciting and inspiring than I inititally expected, plus it really suits minimalism very well :D

 

Ha, I've got a bunch of 90s rack gear as well. I rarely use any of it in my tracks, but can't bear to get rid of it either. I'm convinced that someday I'll make an entire album using just the JV1010 :P

 

My final acquisitions were a JV1080, a D-110 that a friend gave me, and a PG-10 for the D-110 that I sold my Bass Station Rack to buy.  With the PG-10, the D-110 is actually quickly becoming one of my favorite pieces of gear, but usually PG-10 prices are ludicrous, I just happened to nab one from a Japanese seller for a decent price (stillmore than a D-110 goes for but less than half what US sellers usually try to get, and in fantastic shape). OnceI'm settled in the new place I kind of want to try to do a whole EP worth of material just using it as a sound source and the Octatrack for sequencing and processing  it.  Maybe sampling but maybe not, I might just use it in poly mode with the direct outs to have 3 mono synth parts and a drum part, and just use thru machines to mess with them i real time.

 

 

I'm worried it's going to make me want to get a D-50 but I've sworn off buying gear for the next few years, if I want anything else I have to build it or trade for it.

Edited by RSP

Share this post


Link to post
Share on other sites

ahh yeah those Roland programmers are getting to ridiculous prices now. I sold my JX8P because I couldn't be bothered holding out for a PG800 that would end up costing me more than the synth.. they do have a really nice design to them though.

 

Ha, this is turning into the GAS thread. I'm also trying to hold off buying gear, but I'm pretty sure I'll end up getting an A4 this weekend :P

Share this post


Link to post
Share on other sites

ahh yeah those Roland programmers are getting to ridiculous prices now. I sold my JX8P because I couldn't be bothered holding out for a PG800 that would end up costing me more than the synth.. they do have a really nice design to them though.

 

Ha, this is turning into the GAS thread. I'm also trying to hold off buying gear, but I'm pretty sure I'll end up getting an A4 this weekend :P

 

My method has been to start helping friends of mine who are in bands but are getting more in to synths choose stuff (so I can shop vicariously through them; a friend of mine has had a Juno 60 sitting around his practice space for years and had no clue how much they're worth now, so he might sell it and use the money to buy like half a dozen less cool but more interesting things, and if he does I'm going to work on him to get an ESQ-1 since it's one of the last pieces of attainable old gear I really lust over but the last thing I need is another big keyboard), and also any time I want something I remind myself I have about a dozen bare PCBs I've collected over the years sitting in a box waiting to be built.

 

Yeah, this was a bit of a derail, I'll stop.  Back to OT talk !

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
Sign in to follow this  

×
×
  • Create New...