Synth Thoughts: Lessons from the Eurorack Universe

It was in September of 2017 when I got my first two Eurorack modules: a dual VCA and a MIDI-to-CV/gate interface. My goal: make analogue monosynths velocity-sensitive.

A lot has happened since then. In June of 2018, after a lot of planning, I got a full-fleged MonsterBase case together with a bunch of modules to fill it. Some semimodular synths were acquired before, and so today I look at a nice setup – and a few valuable experiences.

In no specific order, let’s go through them!

Modulargrid is a good friend

modulargrid.net is an online resource containing a module database, a system planner, and some other goodies (like an active forum).

This one allows you to search for modules, assemble them into a setup, share screenshots of those, and much more.

modulargrid.png
You want to fill that blank space with a distortion module? Here’s your choices!

There’s a free account – and the paid-for (affordable) unicorn account which gives you a lot of benefits.

Really, if you have more than, say, ten modules, you’ll want to use this.

Modules for every need – sadly, that’s not true for cases

Modulargrid currently has 8249 modules in its database. That’s a lot.

modcount.png

It stands to reason that there’s something for every need in that universe. Sadly, the choice of cases is rather poor in comparison, and I’ve yet to find a flexible (may I say “modular”) case solution. What’s more, due to the sheer volume of these devices, international availability is an issue. Prepare to spend some time looking – and don’t forget to mind installation depth (see below)!

It’s not that beautiful behind the scenes

The good news: seen from the front, everything in the Eurorack world is really and truly interoperable and even manages to look cool! Modules have fixed height (3HU), quantified width (in HP/TE, starting from 2), and exactly one type of connector (1/8”) for all your needs. You can patch everything into anything without any risk of damage, and if you remember to plug outputs to inputs, something will happen.

This is not true for the world behind the modules, i.e. what happens inside of the cases. Let’s start with the maximum module depth: the original Doepfer A-100 specification set that to be 100mm, and most (but not all) Doepfer cases allow for installation of modules with such depth – and that for a reason, as there are modules with a depth exceeding 70mm.

Sadly, that’s an exception. Most cases have a maximum installation depth in the 50-60mm range, and there’s cases going as shallow as 34mm. What’s more, module manufacturers are not exactly very well-organized in publishing the installation depth of their modules (and might forget to consider cables mounted to the back in their specification) so unless you have one of those really deep cases, it’s often some hit and miss.

doepfer_a108.jpg
Doepfer’s A-108 a.k.a. best LPF in the world is an example of a deep module.

Next is the wiring inside of the cases, which mainly takes care of power supply. This is done with a combo of ribbon cables and (female) headers, which are usually supplied with the modules. They’re connected to the module on one side, and to a backplane on the other – the backplane either being an actual backplane, or another ribbon cable also referred to as a “flying bus board”. Pin connectors are either unkeyed (defined by spec for whichever reason) or keyed – if they’re the former, you might reverse polarity, which can ruin modules, if it’s the latter, you might run into situations where you can’t fit it properly (and still are able to ruin modules). The supplied cables also have a tendency to be too short, and most of my experiences come from combo of cases and modules both by Doepfer. They build the case, they build the modules, they defined the standard, so – why?

There is a workaround at least to the short-cable issue (you can DIY them pretty easily with COTS components), but it’s still not nice.

Size matters – and there isn’t a perfect one.

We’ve already talked about the installation depth (which can be an issue) and height (always 3HU), which leaves the width: modules go from 2TE (which is about 1cm) to 80TE (Behringer Neutron) , maybe beyond that, and definitely pretty much everything in between (including uneven widths, which violate specs).

Historical context: in Doepfer’s original A-100 design, there seems to have been an unwritten rule that minimum width is 4TE for any module, minimum width for a module with two rows of jacks is 6TE and minimum width for a module with both jacks and knobs is 8TE. Then came manufacturers like Bastl, Erica (with their Pico series), and 2hp who only make components which are – you guessed it – 2TE wide. Recently, Doepfer has even embraced the downsizing trend, while still keeping it on a sensible level. This is a living world!

First thing you need to understand is that size costs money. How much? In a recent calculation, I found that one TE of rack space costs roughly €2 in the proportional cost of the case. Which means that if you have eight individual VCAs in your system (not an unusual amount for a larger setup) and you opt to use components which use 2TE for a pair instead of the oldskool ones which use 8TE for a pair, you’re saving 48 bucks in rackspace cost alone.

However, 2TE components rely on very slim knobs and jacks that extend right to the border of the front panel – meaning you might run into problems if you put a lot of them close to each other.

PICT_20190721_164357.JPG
Things can get pretty crowded. Which is an issue if you have knobs between those cables.

So what is the sensible middle ground? While I find that having a compact (as in lots of functions in a certain width) setup desirable, planning sensibly with regard to alternating between slimmer and wider modules if possible and maybe going for a 4TE variant instead of a 2TE one if in doubt makes sense.

This world embraces Kaizen

With normal synths, you buy a complete, ready-to-go system. If you later find you’d have wanted a different filter, or maybe a third envelope – well, congratulations, you’re screwed.

The modular world is different. You typically are able to have it change and grow with your experiences, your changing demands.

It’s a world of Kaizen – a world of continuous improvements and PDCA cycles. You might plan a system based on your understanding of individual modules and your needs. Your needs may change, as may your understanding (and let’s not forget the very dynamic market of available modules), so it will make sense at some point to assess where you’re standing and where you might want to go. And this is typically possible by adding and replacing a few modules, rather than getting a whole new system.

Set limits (a.k.a “it’s addictive”)

Yep, this stuff is addictive, and you run a risk to get so many modules that it becomes next to impossible to manage it.

So how do you set your limits? In my case, it opted to center it around a somewhat large case that was large enough to allow for patches which include many voices and sounds, yet build it in a way that made carrying by one person possible, all that by using COTS components and cases/bags.

Due to the repeated use of this abbreviation: COTS stands for “commercial off the shelf” and means you use things that you can buy in a normal store.

The centre component is a Doepfer Monster Base – a large, two-row 168TE case that stretches the limit of the “desktop case” variety. Kudos to @triumcirculorum for the recommendation – a guy who next to said advice releases interesting stuff, remastered two of my albums and is a modular synth artist in his own right.

Built around this is a “board in front” holding three skiff cases (two of them with semi-modulars), and another board with a minilogue, an Analog Rytm, various stompboxes and a mixer. And yes, those two boards fit into a standard keyboard bag together, so you can carry it together with the Monster Base and the keyboard they sit on in one go. Although it’s not an easy one-person lift.

Spielsachen.JPG
Yup, can be carried in one go. Surprisingly.

The total Eurorack space of this? 580TE – which may seem like a lot if you’re just starting, but believe me, you’ll hit that limit rather quickly, and might be happy you decided on such a limit beforehand. Better still if you set a lower limit, like only a Monster Base. Or a Mantis (a fantastic, affordable case btw).

Experimentation is encouraged. Or required?

What happens if you patch an oscillator into a VCA’s CV in? Or a series of pulse into both the audio and frequency CV in of a WASP filter? Something will happen. And you can find out.

In fact (to phrase it in a simplified form), any connection of exactly one output to (at least) one input will do something.

You’re not required to do this. You can always go standard oscillator into standard filter into standard VCA, envelope out into filter cv in and other envelope out into VCA cv in, and keyboard cv out into oscillator 1V/oct in, and gate out into envelope gate in. But most probably, this isn’t why you got a modular thing in the first place.

0-Coast
Makenoise’s Patching Rules: there are no rules. (from 0-Coast manual)

So experimentation is not required. But definitely encouraged. Maybe even a raison d’etre.

Supply Chain Management just became a lot trickier

This section is from a German perspective. I doubt that your mileage will vary.

I’ve been used to be able to get all of my music-making-related stuff from one of the large German retailers, and practically everything I got with a maximum of one week of lead time. A lot of things are even available on amazon, if you absolutely want to shop there.

I’ve been buying modules and accessories from five German and a Netherlands-based stores. I didn’t do that because I wanted to, but because you can’t get everything everywhere. Want some examples? Sektor Sieben cases from justmusic. Synthrotek from Musicstore. Thomann has the shortest lead times for the things he has in stock (which, among other things, are a lot of doepfer modules). Schneidersladen for everything else. Except for error modular, which only seem to be available from Error Instruments.

Sometimes lead times can be a guess at best. I’m still waiting for a module that was ordered two months ago with a projected lead time of two weeks.

This doesn’t mean that you can’t work with modular synths. You just need to know that, well, a shopping trip, if only an online one, will take just slightly longer.

Modules for every need revisited

I already mentioned this: there’s a huge number of modules. And then started bitching about cases. Now I’m revisiting that statement to shed some light on the actual choice of modules.

The Eurorack world is a big one. Part of that is caused that the entry hurdle for new designs and designers is really low. Some people publish designs (including gerber files and BOMs) and release them with some open license. Some sell DIY modules where they supply the faceplate and a PCB and leave you with a BOM of COTS elements and assembly instructions. Some designs are made in truly low quantities, some only fulfill a very specific purpose. Others are pretty generic, are sold in large numbers, and have been available for almost 25 years.

This is both a curse and a blessing. So how to deal with it? One way is to simply make use of the information that’s available. One of the reasons that it took the Eurorack/A-100 so long to really achieve widespread use might just be that it required readily available loads of performance/tutorial//product presentation videos as an enabler for mass acceptance. And there’s dedicated forums, and folks on networks like twitter, where people will help you out.

Another idea is to start generic and then diverge into the stranger realms. Starting with one of the basic systems that manufacturers like Doepfer offer can be a good idea.

Basissystem
The Doepfer Bassissystem 2 – a good starting point, which also offers 84TE row for growth.

With those, you can start to learn the ropes and discover which modules you like, for which functions you might want more, and what you want to change or expand.

Another really good starting point is a voice module/semimodular synth conforming to the Eurorack standard. Looking at devices I reviewed already, this could be for example a 0-Coast, a Neutron, a DFAM or a Nanozwerg Pro (the latter with a fitting case). With those, you can also gather experience quickly and then move on to larger, individualized systems.

Patch memory?

Why is the set of settings for one sound on a synthesizer called a patch? Exactly!

It’s obvious that there’s no patch memory on a modular system. Well, unless you get a combo of a six-axis robot, a camera and a smart computer system. Which means that storing patches means writing down/photographing cable connections and knob/switch positions. And even if you seemingly recreate a patch, it might just sound slightly different. Or you simply decide to not write anything down after all.

How does this affect your work as a synthesist and musician? It just means that storing your MIDI parts and synth patches isn’t enough to allow for total recall, to work on a different piece in between and then continue hours, days or years later as if nothing happened. Most of the time, it’s a disadvantage. Sometimes, it’s an advantage, as it forces a more immediate workflow on you. All the time, it’s good to keep that in mind.

Module placement is important

It doesn’t matter if your system is a small lunchbox or a full-blown monster case solution: the way you place your modules is important.

You could argue that signal integrity issues play a role here. However, my understanding (both from theoretical analysis and practical experience) is that this won’t drive your decisions: modular systems are noisy and distorted enough already, so the reduced cable length here and there hardly plays any role.

Which makes it a question of human factors engineering.

The surface of the system is shared between a multitude of knobs, buttons, switches, LEDs, displays and jacks. You’ll want to look at the displays and manipulate the knobs, buttons and switches as you play the thing. Unfortunately, cables are connected to the jacks and result in obstacles in the space before your system.

Devine.jpg
“Ascension” by Richard Devine –  an example of a patch that’s hard to grasp

Your goal might be to keep the cables to a minimum and, while you’re doing that, keep them out of the way of the knobs and such. Which is a bit tricky, because in the end you could patch anything to everything, as we discussed before.

However, depending on the music and the sounds you prefer, there could also be preferred ways to patch. As an example, if you typically do subtractive synth voices, you’ll go from oscillator to maybe mixer to filter the VCA to output with your audio signal, control the oscillator with pitch CV and both filter and VCA with an envelope, which in turn gets triggered by a gate.

So if you had a setup with two rows, you could fill the top row from right to left with oscillators, mixers, filters, VCAs, and then something for output. Then, simply put all of the usual control sources (sequencers, envelopes, and the occasional LFO) into the lower row. That way, cables in the top row always go over a short distance from right to left, and control signals always go from bottom to top.

In my personal setup, the Monster Base is organized in a way similar to this: grouping modules into sources (oscillators, noise, microphones), modifiers (filters, effects, VCAs) and controls (LFOs, sequencers, envelopes etc.), you find a layout like this.

Monsterbase_Overlay.jpg
A setup ordered in sources (green), modifiers (red), controls (blue) and sinks (yellow)

This not only helps in keeping your cabling in an orderly fashion, it also makes patching more intuitive – especially important if you swap modules a lot. Because if you’re looking for an oscillator, you always can look to the right.

Once you’re done with that, you might work on the details in some situations where the patches are almost always identical. As an example in the aforementioned setup, I generate a stereo mix using a pair of Doepfer A-138s (4–2 stereo mixers) and Intellijel Unity Mixer (2x 3-1 adder). In the way I had first arranged it, it looked cooler – but the cables from the A-138s’ outputs (which are on the right of the module) had to go over said modules to patch into the Unity Mixer. By swapping the bottom A-138s with the Unity Mixer, I could do with cables that didn’t get in the way of my use of the A-138s – even though it meant it didn’t look as cool as before.

before_after.JPG
Swapping just the adder makes all knobs and jacks on the two mixers fully accessible.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s