When we’re connecting together other audio gear, we know we need to pay attention to whether a given signal is at microphone, line, or professional level – and even then, we may need to tweak the levels to get the signal in the sweet spot between noise and distortion. (Or maybe push it into the “sour”spot where we get noise or distortion on purpose.)
That might be true in Buchla, Serge, Moog, or even Dotcom systems…but the Eurorack world is a bit more like the “wild west” with literally hundreds of manufacturers making their own decisions about what is the right way to do things. This is compounded by the Eurorack format having less headroom (distance between a typical signal and the power supply “rails”) than the other formats. Although few people talk about this issue, I’m convinced that signal level mismatches – not size, components, or even the power supply – is what’s really the problem when people complain Eurorack “doesn’t sound as good” as the other formats.
I’ve been spending some time tackling this issue over the past month, including in an interview for the new Rakt newsletter, and in posts for my Patreon subscribers. In the main article in this newsletter, I’m going to explain what the basic problems are, and how to set up your patch to work around them. Once you learn to manage the signal levels in your modular, you will be able to get a wider range of sounds out of it, and overall it will behave more like you expect.
- featured article: How to manage signal levels in your Eurorack modular system.
- Alias Zone updates: I’ve updated We Only Came to Dream with a new track: Devotion. Here’s the details, as well as how the track was created, and a video of it being performed.
- Learning Modular updates: I had a fun, wide-ranging interview with Arch Delaro of the new Rakt newsletter. It was mostly beginner-focused, but there were also good tips for intermediate-level users as well. Although originally intended for print, he recorded video of the interview as well…and that’s included below.
- Patreon updates: How much envelope does it take properly drive a VCA? How much signal does it take to properly drive a VCF? These are two topics I covered in detail in recent posts for my Patreon subscribers.
- upcoming events: I’m staging a mini-tour for April 2023, leading up to this year’s NAMM show. I’m also starting to think about the rest of the year…and you can be part of that.
A Patching Strategy for Eurorack
When I decided to get back into modular, I chose the Eurorack format for a few reasons. One is that it has the greatest variety of manufacturers and modules, feeding my love of finding new sounds. Another is that it put control voltage and audios signals on equal footing, making it easier to experiment with all sorts of cross-modulation and other ideas.
However, the Eurorack format has a couple of inherent problems, as well. One is that there is very little room between the standard modulation or signal level and the limit of what its specified power supply can deliver, making it too easy to clip audio and flatten modulation. Another is that different manufacturers have chosen different levels for those audio and modulation signals, creating incompatibilities between different modules (sometime even from the same manufacturer!).
At this point, it is way too late to fix either of the problems outlined above: literally hundreds of thousands of modules have already been manufactured, and they’re not going to change.
Therefore, it has fallen on the user to adjust for those differences themselves. The problem is, hardly anyone has told the user that. And manufacturers have not made it easy for users to figure this information out for themselves.
It is possible – and actually, not that hard – for the user to “patch defensively” to take these issues into account. The added bonus is that this patching strategy can also be used offensively to coax new sounds out of your modules.
The patching strategy itself is simple: In short, make sure there is an attenuator (bonus points if it can boost the signal as well) between every source and destination for audio and control. Then tweak those attenuators to find the “sweet spot” (or salty spot, or sour spot, or…) for the sound you’re trying to create.
I’m going to cover what the problem is in the first place, followed by more details on this recommended patch. Then I’ll conclude with a check list of what I wish manufacturers would do to make the life of their users easier. Even if no manufacturer reads this and follows it, it will give users an idea of what features and specifications to look for when choosing modules.
What’s the Problem?
There are three problems, actually.
One: There’s a lack of agreement between manufacturers about what is the “correct” or “expected” voltage level for both modulation and audio.
For example, most VCAs I’ve tested need a control voltage of +5v to pass a signal at “unity gain” where the output level equals the input. However, different envelope generators typically send from +5v to +10v at their peak, with +8v being common. This means that envelope shapes or the audio itself can get distorted.
Another example is that most VCOs output a signal with a ±5v range. And about 2/3 of the filters I’ve tested expect to see ±5v for optimal operation. However, as soon as you mix two VCOs together, the signal can be as high as ±10v – and now 2/3 of those filters are going to be distorting.
The original Doepfer A-100 specification suggested that VCOs output ±5v, LFOs output ±2.5v, and envelope generators output 0-8v. At a Superbooth meeting between major Eurorack manufacturers a few years ago, those levels were tweaked to ±5v for LFOs, and 0-10v for envelopes. This updated recommendation makes perfect sense: It puts the overall voltage range of VCOs, LFOs, and envelopes on the same level, with just a simple 5v offset between LFOs and envelopes.
However, not all manufacturers follow either one of those suggested approaches. Testing modules in my own system, I’ve found VCOs with ±1v to ±8v output, LFOs and other modulation sources that output ±2 to ±10v, and envelopes that output a maximum of +3v to +10v. My VCAs expect anywhere from 4.75 to 7.5v to reach unity gain, and filters clip at anywhere from ±4.5v to ±10v. What a mess.
Two: With a Eurorack specified power supply of ±12v, most typical op amp chips can handle a maximum signal of about ±10v. (There are “rail to rail” chips that can handle closer to ±12v, but those are rarely used – most often in output modules.)
If you mix two typical oscillators together at full level, you’re already at the ±10v limit; if you mix in a third, or pass the signal through a resonating filter that boosts the signal when the corner frequency and harmonics are aligned, you can easily go over that limit. Likewise, if you mix together two modulation sources, you can easily reach or surpass that same limit. The result is the audio and/or control signals getting “clipped” as they are unable to go beyond ±10v. Sometimes the result is intentional, or a happy accident; a lot of the time, it degrades the sound, and causes your modules not to behave the way you expect.
(At least virtually everyone seems to agree on the 1 volt per octave pitch control standard. However, there is little to no agreement over what range to use for pitch voltages: some use 0v as their lowest point, and/or 5v as their high point; others go well below or above that. I discussed some of the consequences of this awhile back.)
Three: Most users are not aware of the issues above: you rarely see it mentioned, even in user manuals. And, many users are in the habit of turning signal level attenuators up all the way as their starting point – louder is better, right?
What’s the Solution?
To work around these problems, it’s time to stock up on utility mixers – preferably ones that can boost a signal in addition to attenuate it. If you can boost/attenuate a signal as well as optionally offset its voltage (for turning LFOs into envelopes etc.), you can solve every problem mentioned above.
My personal favorite module for this is the Frap Tools 321. Each of its three channels can boost, attenuate, invert, and/or offset a signal. It can also mix signals together, of course. And as a bonus, those mix outputs have a “-6dB” (in other words, attenuate by 50%) switch, making it very easy to mix together a pair of sources and cut them back down to the level of a single source. The Erogenous Tones LEVIT8 is another utility mixer that can boost signals (plus invert four of its eight channels). If you know of any other mixers that can also boost, please list them in the comments below – they are far too rare. If your mixer can only attenuate signals, there’s a cheap workaround: use a splitter cable or a mult, and send the same signal to two inputs, allowing you double its level if needed.
Now when you patch, make sure there is at least an attenuator – built into the inputs and sometimes the outputs of many modules – if not a full-blown utility mixer between every modulation or audio source and destination. The image at the top of this post shows a suggested patch for a typical “East Coast” approach using the universal patch language from Patch & Tweak; I’ll repeat it here:
After you have the basic patch built, experiment with different amounts of boost or cut of the audio signal into the filter or VCA (or waveshaper, or distortion, or effect, or whatever) – this will allow you to move it between clean and overdriven, with different results depending on the module. (For bonus points, if you’re using a utility mixer, you can also try offsetting it, which can cause asymmetric clipping in the module downstream for a different sound in some cases.)
Similarly, experiment with different amounts of modulation depth going into the control voltage inputs of your modules. You’re probably already used to doing this with LFO or envelope depths to, say, the filter cutoff, using the attenuator built into most filter modules.
Also play with control levels such as how hard your envelope is driving your VCA. With a typical 8-10v peak envelope output, if you turn the CV attenuator up all the way on your VCA, most of the time you will either be boosting the signal level (which can cause unexpected distortion), or clipping the top of your envelopes (adding an unintended “hold” segment between the attack and decay). In particular, if your transients are sounding too “fat” (lingering at their peak level too long), try turning down the CV attenuator on the VCA until it sounds the way hoped. In the figure below, the left side is an example of a “flat & fat” output from a VCA when the envelope’s depth has been turned up too much; the right side shows the result when the CV amount has been attenuated.
As a bonus, I highly recommending adding an oscilloscope module to your system – preferably one that has at least two simultaneous inputs or “traces.” This will allow you to see what is the output level of your various modules, and to compare the signal you’re sending into a module to the signal coming out of it. This is what I do with virtually every new module I add to my system.
What Should Manufacturers Do?
At a very minimum, I would love to see manufacturers add more information to their manual or online specs that spell out what voltage levels their modules output, and what levels they expect on input for “optimal” operation (i.e. unity gain on a VCA, the point before clipping on a VCF, etc.). That won’t cost the manufacturers anything extra aside from some time, and will save users from having to buy oscilloscopes and measuring it themselves.
As for signal levels, I think the “new” semi-standard of ±5v for bipolar signals (VCOs, LFOs, and most other modulation sources) and 0-10v for unipolar signals (envelope outputs, and those few VCOs that output unipolar waveforms) is a good starting point. (I also think 0v for middle C with pitch voltages – and a minimum range of -3v to +7v as output for sequencers, keyboards, etc. – is also worth adopting, but I suspect I’ll get fewer manufacturers to agree with that.)
Then, as much as practical, add some form of level control to each input (or output – for example, the WMD Javelin’s envelope generator has a switch that toggles between 3v, 5v, and 8v as its peak). If you can’t put a level control on each input (for cost, space, etc.), then mention somewhere in your documentation that you recommend a user patches through an attenuator or utility mixer before patching into those inputs, to be able to get the most out of your modules and make them compatible with whatever other modules the user may have.
As for users: Look for the above in modules you are considering buying. In particular, some digital modules are limited to a 0-5v range as that’s what convenient for many microcontroller chips (although those ranges could easily be extended with the addition of an op amp and a few parts if the manufacturer chose to; the cost or parts count isn’t that bad).
That’s my advice/plea/manifesto when it comes to Eurorack voltage levels. I hope it helps some users, and that it encourages more manufacturers to make life easier on their users. And maybe as a result, we’ll get fewer people claiming Eurorack “doesn’t sound as good.”
Alias Zone Updates
I’ve updated my “comeback” album We Only Came to Dream, replacing the middle track (Ash Tree Window) with a previously unreleased one – Devotion – that both I and my mastering engineer/mentor Howard Givens feel greatly improves its overall flow. I’m also creating a limited edition physical CD of this version that also includes my own art; it is currently available for pre-sale on Bandcamp.
Although Devotion had not appeared on an album before now, I did previously release a video of it (above), and also wrote a detailed “track breakdown” post for my Patreon subscribers which is publicly available for others to read.
Ash Tree Window – the track that got cut from the original version of the album – remains one of my favorite tracks overall; I just wanted to present it in a better context. A lightly updated version of it will be included in on a new album later this year.
By the way: Friday February 5 is “Bandcamp Friday” where they don’t take the normal commission, meaning more money for the artists. If you were thinking of buying some new music, that’s the best way to also support the artists. I’ve been building my own “wish list” of albums I’ll be buying then.
Learning Modular Updates
Arch Delaro has started Rakt: a newsletter aimed in particular at those starting out in Eurorack. When I learned about Rakt, I agreed to be interviewed, and we arranged a Zoom session that Arch recorded for note-taking purposes as he asked questions. We ended up talking for twice as long as Arch had planned. Rather than edit the result down into a print interview, Arch decided it would be best to share the actual video, as we hit a wide range of topics. That’s what you see above.
The Learning Modular Patreon Channel is an archive of literally hundreds of articles and videos I’ve created for modular users trying to take their skills to the next level. I’m constantly adding to this archive with new posts based on what I’m working on currently. In January I wrote two technical posts related to the “managing levels” theme in the main article above:
- Envelopes versus VCAs: Mismatched by Design where I explain the mismatch between most envelope levels and what most VCAs want, with examples from my own module collection, and suggestions on better ways to work with them.
- Overdriving Filters where I go through 14 of the filter modules in my Monster studio system, detailing in video and sound when and how each of them enter an overdrive state. I was surprised myself by how some of the modules responded.
Both of those new posts are visible for +5v and above subscribers. (For those who are not already subscribers, the +5v level – $5/month – gets you immediate access to a library of literally hundreds of posts as well as new ones each month, while $12/month adds access to my Eurorack-focused online modular synthesis courses.)
As I’ve mentioned before, I have already booked a mini-tour in southern Arizona & California in the first half of April, leading up to the 2023 NAMM show. Stops will include:
April 4: I will be part of Steve Roach’s Ambient Lounge series at the Century Room in Tucson, Arizona.
April 8: Jill Fraser and I will be playing at the very funky FurstWurld in Joshua Tree, California.
April 12: I will be giving a performance plus talk/Q&A at 17th Street Recording Studio in Costa Mesa, California the evening before the NAMM show (the studio is only 15 minutes away from the convention center). Basek will also be there talking about some of the gear he uses, plus there is a promise of something special from the Vintage Synthesizer Museum to play with. Pizza and drinks will be available.
I have no solid plans after NAMM other than to drive back home. However, if someone wanted to put together an interesting gig (even a house concert) somewhere between Southern California and New Mexico early the week of April 17…let’s talk.
I’m also talking to the Moogseum about an event in Asheville, North Carolina some time later this year. If you know of something interesting going on in that part of the country later this year – or were interested in hosting a house concert or similar event – again, let’s talk, and I’ll see if I can string together another mini-tour out there.
January has been a month of re-grouping and finishing some long-delayed projects. February should see a return to music-making; stay tuned.
staring at too many oscilloscopes lately –