I’m trying to make a state azure/ tangerine dream style rack in the least possible space, but it feels like something is off. I don’t know what it is, but a gut feeling is telling me that something is hindering my progress, not helping it. Can someone at least try and help?


Do you already have all of these modules? What is it that you feel is missing? Are you having trouble getting the tones, sequences, or mixes that you want?


I hope it's an exercise in planning rather than you have all theses modules and you're stumped!

in addition to farkas' excellent questions - here are some thoughts:

sequencing/quantising: there is massive overkill even for using sinfonion - I recommend 4-5 sequencer channels - if you want to add more random then a single permutation/pachinko is sufficient - personally I would go for Marbles instead of either pachinko is a clone of marbles and less ergonomic - and marbles is like a triple turing machine - which permutations is derived from - one of the sequencer channels should be completely independent (ie run at a different speed/separately clockable) and programmable - this is for switching song parts in sinfonion

metron is a trigger sequencer - mostly used for programming drum patterns - I think this is completely superfluous - unless you are planning on using the bitbox for

I would try to think in terms of

sound sources < sound modifiers < modulation sources < utilities + sequencing and final mixing

I would also when trying to emulate tangerine dream (or any other music based around multiple movements within the same piece of music) think in terms of layers of modulation - how to alter volume of parts over time?

sound sources: I see 3 modules - plaits, t-sl v2 and chainsaw - plus the bitbox (which I know nothing about) which I am assuming that you are intending to use for another 3 voices - you really want 8 v/oct inputs for voices - so this seems about right in terms of number - not what I would have picked but fair enough - that's personal taste - although I would probably want 2 or 3 of the same analog vco

vcas: definitely very light on the vca front - 4 vcas (yes I know that plaits has a built in lpg) I would want at least another couple of quad vcas in this case - they are needed for much more than note shaping - automated volume control over time and modulation for 2, which brings us to -

envelopes/modulation sources: there might be just enough envelopes, but then there will be a lack of controllable lfos - I would be tempted to add zadar and batumi and a matrix mixer to make modulation more interesting - mix modulation sources - possibly might need a couple - I might remove o&c

sound modifiers: plaits has a built in filter/lpg and chainsaw is stereo - the filters you have are mono so would work with individual outputs from the bitbox and the t-sl - I would probably dump all of them in order to get 1 or 2 more ergonomic mono filters and add a stereo filter for the chainsaw - I'd also want more in the way of general effects - a single mimeophon and beads are not really going to cut it for that many sound sources - I'd probably add at least a couple of fx aid xls, probably more

utilities: not really enough in this size - I've covered vcas and to some extent mixers (matrix mixers) above, but you will almostt definitely want more mixers - for panning and cross fading I would learn to patch these from vcas, lfos etc, but there is not a lot in the way of other basic utility functions - I would spend a decent amount of time researching all the different sorts of utilities and how they are used and an equal amount of time thinking about how these can be incorporated in your patching of this synthesizer (I'm talking weeks or months here not minutes or hours)

end of chain: how are you planning on listening to this? personally I think the eoc mixing solution you have here is woefully inadequate - I would look at the equivalent of a battleship eoc mixer - either the wmd performance mixer or something that is expandable such as the tesseract modular tex-mix (I have this one) which is extremely flexible as it is expandable 4 channels of mono or stereo at a time (and better yet more cost effective and lower power consumption than the wmd perf mixer)

I'd also want some effects on the send bus of the mixer - probably some reverb and I personally like the lofi setting on the fx aid as an eoc effect - so possibly another couple of fx aids and/or a very nice stereo reverb - the new strymon starlab looks and sounds very good for this role

if starting from scratch go very slowly - build a single voice chain (sound source, modulation source, sound modifier, some utiltities) and work from there maybe planning to add another module or 2 every few weeks or months towards something similar to this over a period of a couple of years - otherwise you will probably be overwhelmed

additional after looking at the bitbox further - if you are intending on using this as a sample player for 3 melodic voices + some percussion - I would probably want a more sampler with more channels - or multiples of this - I'm not going to recommend anything for that though as I know nothing about samplers in eurorack

Utility modules are the inexpensive, dull polish that makes the expensive, shiny modules actually shine!!!


Jim's advice is spot on. Ditch the Bitbox , STMix module, and Metron sequencer. Get a massive mixer like WMD Performance Mixer or the Befaco Hexmix. I have the Befaco Hexmixer and love it.


Whoops! Forgot to mention that I don’t and have never owned any modular synths before. This is future planning.


Excellent well I do also recommend download free software VCVRack and try building a similar rack. It will be great experience and helpful in your planning and learning process at no risk.


Thanks for all the tips jim! I do have a few things I would like to share tho.

1: I totally agree that metronome is complete overkill. It’s been a long battle between the tubbutec sixquencer and the metron and still no real winner. I like how compact the sixquencer is, but i want something more ergonomic like the metron. The algorithm by grayscale seems appealing, but i do like the idea of ≥16 steps in case i want to add a cool drum transition at the end. (After thinking about it, I’m just going to go with the algorithm.)

2: the mixer situation as been another long and arduous journey as well. I had the hexmix and hexpander in the blueprint a while ago, but that took up an assload of space and i just decided to scrap it, I’m thinking about diy’ing an 8 channel stereo mixer in the same form factor as the mutable instruments frames / blades, but I don’t understand breadboard schematic and have never been able to find a blueprint for a mixer online.

Thanks for all the help!


Well personally for a complex sequencer, if I were you then I would go with an Eloquencer, Vector 512 or Erica Synths Black Sequencer over the Metron for what you are looking to achieve. I have the Metron and the other sequencers except for the Black Sequencer. The problem with Metron is that it is not great for sequencing pitch or cv even with the add on Voltera module. It’s great as a gate drum sequencer. You want the ability to create complex ratcheting and random evolving patterns which are tough to do on the Metron.


To add to what Jim mentioned, I would really recommend revisiting what kind of gear bands like Tangerine Dream were working with. It was often very simple analog gear, at least in the early days. A few analog oscillators and filters (Moog, Roland, Arp, etc.), simple 8 step sequencers like the Doepfer or Behringer (though I would recommend the Winter Modular Eloquencer for your purposes), dotted eight note delay, etc.
I would go back to the drawing board on your plan. You can likely achieve what you want to do in a much less expensive manner.
As a follow-up question, what gear do you already have?


As far as mixers go, it kind of depends, while the WMD performance mixer is great you could get away with something like the cosmotronic cosmix, you definitely can get more value with an external mixer and save the HP.

I agree this is a bit weak on VCA's.

I find sequencers are a very personal item and really depends on the workflow you are looking for. I have several and each has a purpose, one I don't think is mentioned a ton but is my favorite so far is the NerdSeq, as far as step based sequencing I don't think there is a more powerful sequencer 10 tracks (6 modular and 4 sample/audio including fm voices). While I wouldn't consider the sample tracks a comparison to a dedicated sampler (limited memory) they are a nice to have for some added spice, or you can use these as basic oscillators including 2 or 4 operator fm.

I try not to emulate the gear choices of artists as this rarely works out, what workflow works for one artist won't likely work for another. I remember when I was young I brought my drum kit to a party and another drummer used my kit and pretty much showed me I had a lot to learn because I never heard my kit sound so good, I never was able to reproduce his "tone" on my kit. Same kit different artist, completely different result. Its your instrument to develop your workflow and your sound.

If you really want polyphony you are better off getting a polyphonic synth, it can get really expensive really fast in euro and still be a struggle to dial in a sound as quickly as a good poly synth will. State Azure is a good example of a mixed mode artist that incorporates lots of different gear and is not strictly modular and that gives a lot of flexibility and can save a ton in costs.

As others have mentioned, start small. Sound source (plaits is great because it is so versatile with built in lpg/vca), a way to sequence and a modulator, Maths is a great choice. Start here the perhaps add some fx, Beads or an FX Aid, and away you go.

Pip Slope, I think there are better options Maths has great envelopes. I think Acid Rain Maestro is a great clocked modulation source when you start reaching for more modulation, and if you want more function generation/envelopes you can always add a make noise function or a joranalogue contour 1.

I have several Instruo modules they are great, but on the expensive side. I think there are better options here as well. The Dixie II+ is a great analog voice.

As far as samplers go, I think the Assimil8or is the best eurorack sampler hands down, it is a bit expensive but worth it and I have a 1010 blackbox standalone unit and am a fan of 1010 in general.

You definitely have a lot of quality modules but I wouldn't just march down this list because where you start after you have a few modules and a better idea of what you can and want to achieve I assure you will change where you end up.


Well, I think I can see what the problem is here...overall, this is a pretty good example of Sexy Module Syndrome, and the signal flow in there is going to be really convoluted if this all remains the same. Plus, there's A LOT of utilities that ought to be in there (VCAs!!!!!!).

I can see the point of the Metron. But at the same time, I think you're going to need to scale a lot of this build...Metron included...way back in terms of these huge modules. For example, it might look like a downgrade to go from the Metron to, say, this: https://www.modulargrid.net/e/tubbutec-6equencer-3u However, that Tubbutech module is pretty much the sequencer from the Roland TR-606, very workable on the fly and live, and it's got enough outputs to justify using it with the Bitbox as a dedicated trigger sequencer. If you want to go down this lane, those are the sort of choices you need to tackle...either some big module that's pretty spendy, or one that fits but isn't as glamorous.

I'm going to get back to this later tonight. I can see what the main idea is, but this will need a LOT of smackin' around to get it to make more sense...

[A few hours later...]
ModularGrid Rack
OK! Now, that's what I came up with. Much of it remained from your build above, but as I noted earlier, I did wind up having to pull some modules outright. Even so, this was a sizable PITA to work out because not all of those big modules could be dropped. However, most of those were retooled with smaller yet about as functional modules.

TOP: I put the main sequencers plus the Bitbox up top. The Bitbox needed to be located by the Tubbutech 606 sequencer for ease of use. Then we get into the oscillators; you'll notice a second Plaits, for starters. This is because while 1 VCO might be fat, 2 VCOs becomes downright obese! And that's all thanks to detuning one of a pair of VCOs slightly...ergo, two Plaits. Then the supersaw oscillator, and a Veils so that you can have amplitude control over those oscillators. Since the Veils has a "breakable" mixbus, you can mix the Plaits on VCAs 1 and 2 and sum that down to mono at output 2, while the Chainsaw would use VCAs 3 and 4, outputted directly so that you can maintain the Chainsaw's stereo field.

SECOND: Modulation and other things. You have the O&c first, then an ADDAC Intuitive Quantizer, which is a four-voice quantizer with user-definable scales and some rather comprehensive capabilities. This, of course, replaces the Sinfonion which frankly had a number of features that weren't needed here, such as a chordal engine, etc. It's also WAY smaller and WAY cheaper. After that, there's a Xaoc Batumi + Poti for 4 LFOs, then the soon-to-come Buchla (by Tiptop) 281t quad function generator. Think "Quadrax" but with a number of extra tricks, such as the quadrature functions and some useful internal signal mixing. Cheaper than a Quadrax, too. I went with a Lowgain CVP1 for the modulation mixing/processing and a Happy Nerding 3xVCA for modulation amplitude control. Last in the modulation is a Xaoc Zadar + Nin with four EGs, then Mutable's Blades dual VCF, which is actually part of the voicing...but the Metropolix forced that onto this row.

THIRD: First up is a little Doepfer multimodule, which in this case contains your noise source (with fixed filtering), random source, and sample or track-and-hold. This also gives Marbles some noise to chew on. Wogglebug's after that, and then a full-on Permutation completes the random-source section. After that, you'll notice a little 4-step CV sequencer...this is to deal with the loss of transposition capability that the Sinfonion had. By setting up your changes on this little sequencer, you can globally alter the key of the Metropolix or most anything else. Next is Maths, then your Mimeophon and Beads, located above the mixer for convenience.

BOTTOM: Added a PWRchekr at left, then the Mutant Brain and Pam's. And then the pile of little modules is a full-on timing and event extraction setup, which is going to be pretty necessary for making all of the sequencers jump around nicely...it works like this: First module is a dual trig/gate delay. A second pulse divider (the main ones are in the Pam's) then allows for some strange numerical series in division. The compliment to this, the Multiplier/Ratchet Gen, is next, then a small module from Adventure Audio lets you pick off per-count trig/gates. And the key to this entire section is next, a Frequency Central Deep Thought, which is a Boolean logic module. Boolean logic is essential to getting some rather complex time alterations, since they take in two pulse signals and, depending on the gate type, can output a gate based on the coincidence of pulses...or not. It all depends on which gate you use and what you feed it with. After the logic, there's a Xaoc pulse summer, then a module to calculate the minimum and maximum across up to four inputted voltages, followed by a Derivator that outputs several gates, all dependent on voltage motion. The ANA then carries out some math operations on incoming voltages, and lastly there's a Joranalogue Dual Window Comparator...these output a gate when certain voltage thresholds are crossed, and since this is a window comparator, you get more than just the one threshold crossing. After all of the timing modules, the Bastl ABC works as an extra dual submixer to sum signals before passing them on to the mixer, a Toppobrillo Stereomix2. This mixer has your final VCAs for level, but also can pan and set AUX send levels via CV as well. This also has a CUE function for the headphone preamp, which is a godsend if you have to retune on the fly. Mono AUX send, stereo return...but only one of the latter, which is why I put the Bastl Ciao! in, as it has a second stereo in that can be mixed with the Stereomix2's output, letting you use both the Mimeophon AND Beads in parallel routing...which gives you a lot more control over the mix between your dry and wet signals. This also gives you a second headphone preamp, a few more metering LEDs, and your 1/4" stereo output pair.

This wasn't exactly fun, however...with much of the blame going to the case itself. It could be possible to get all of those big modules back in there along with much of the changes you see here, and then some. But sticking with this 4 x 104 cab is a definite drawback; I'd rather see a bit LESS density here, tbh, as I think this might have some ergonomics issues. But the main thrust here was to provide an alternate example, with errors corrected and some new concepts tossed in. However, if you took this to 120 or 126 x 4, it would be possible to reduce the number of tiny modules, which would make this a bit easier to program. Still, it's not THAT far off the mark.


a few more thoughts - depending on what other gear you have - a poly synth for example...

I would swap the chainsaw for a general cv - it's a eurorack implementation of general midi - so has loads of samples of real instruments - plus it can be expanded to be used as a polyphonic cv-> midi converter (possibly the best one out there for interfacing sinfonion with external poly synths) - & general cv will also double up as a percussion source!

agree with farkas - get vcv rack (although personally I dislike it) and consider a dc-coupled audio interface to interface the odular with it - these can be had for cheap - used RME interfaces or a bit more expensive and in rack es8/9

are you thinking of using the bitbox for drums? if not how are you thinking of using the metronome to 'add a cool drum transition at the end'?

building a mixer from scratch is a real pita - seriously check out the tex-mix - I really like mine!

I have something similar in terms of generative (marbles) and programmable (step fader, BSP - & thinking hard about a black sequencer) feeding a sinfonion (plus a pico seq - which I'm not at all keen on - for swapping song parts) for sequencing and maybe a couple more voices than necessary

you can check out my rack
ModularGrid Rack
which is something similar to what you are trying to do (NB left most section is mostly video synthesizer - some of the doepfer modules are used for both though) although this is slightly out of date - it's now spread over 8 cases not 6 as depicted

I agree with @farkas about going back to the drawing board too... step back as far as possible and start with what you requirements are - how many voices and what type - how do you sequence that, do you want programmatic or random or both - do you need percussion - much better in terms of price/features to go outside the box...

I'm looking forward to @Lugia's take they are often very interesting!

Utility modules are the inexpensive, dull polish that makes the expensive, shiny modules actually shine!!!


If I could like all your comments I would. All the help has been great!


Utility modules are the inexpensive, dull polish that makes the expensive, shiny modules actually shine!!!


Lugia, i copied your replan and modified it a little bit. Feedback welcome.