Just as December 2020 was coming to a close, and the year’s surprises, both good and horrible, were seemingly behind us, a new surprise — quite the former — popped up for modular-synthesizer enthusiasts. The hardware manufacturer Instruō, based in Glasgow, Scotland, announced that it was making almost all of its modules available in software form, 17 total, and better yet: entirely for free. The modules run on the free VCV Rack software platform, which is available for macOS, Windows, and Linux. (Visit Instruo at instruomodular.com, and VCV Rack at vcvrack.com.)
The response was immediately enthusiastic. At cdm.link, writer Peter Kirn said, “It’s got just enough of the sorts of tools that let you get adventurous with sound design, while remaining accessible and balanced.” The gearnews.com website praised the originals for their “beauty, depth and innovation,” and pondered whether the free versions would disincentivize hardware sales (its verdict: “I don’t think so”). Discussion boards quickly started chiming in, paying particular attention to the Instruō hardware modules that weren’t ported over, including Arbhar and Lúbadh.
I reached out to Instruō founder Jason Lim to learn more about the process and decision-making. Why weren’t those few modules included? How did the company manage what must have been a considerable undertaking? Why had they opted to make them free, since VCV Rack has introduce “premium” (aka paid) modules from a range of developers? As many of the original Instruō modules are analog or analog hybrids, what was the experience of porting them to the purely digital domain? How did the company approach the differences between the hands-on, knobs’n’sliders originals and the software versions? Lim graciously agreed to be interviewed via email, and below is the discussion, along with visual examples of both the original modules and their virtual offspring, as well as the process involved in bridging the gap.
The Instruō modules in hardware (above) and VCV Rack software (below)
Marc Weidenbaum: How did this project come to be, making the hardware modules available in digital form on Rack? Were you or anyone else at Instruō using VCV Rack before initiating the process of porting the modules to it?
Jason Lim: There were a few factors that led to this project, but it was very much a case of very fortunate timing. The planets aligned, so to speak, and allowed for this to become a reality. For a bit of background, I quite regularly have internship placements ongoing here at Instruō. Some positions have been coordinated as more formal partnerships with various educational institutions. In many cases meetings are more personal, friend of friend introductions and the like. I’ve been extremely fortunate to have Instruō develop so rapidly. At first it was me alone in the spare bedroom. Since those days I’ve been able to grow the team quite quickly with a really creative group.
I am Glasgow-based, and there are a number of very good universities and colleges here. My friend and collaborator, Dr Sebastian Lexer (co-dev of the Arbhar and Aithēr modules with myself), teaches at Glasgow University in their Sonic Arts department. He splits his time between education and working at Instruō. He gave recommendation to me for a couple of Master’s students last year who were approaching their final years of study. Their program requires an industry work placement in which they would work on, document, and develop a project within the field of music technology. This was of course all put in place pre-COVID and their start dates were planned for mid-June 2020. As the time approached, we had to rather quickly shift gears and figure out a new project that would be better suited for remote working.
I’ve followed VCV as a platform since its release, with great interest. What Andrew Belt has brought to the community is huge! It’s the sort of thing you’d read in forums as speculation and pipe dreams. It’s an immense undertaking and I think he’s built something truly groundbreaking. I occasionally teach a synthesis/sound design course here in Glasgow at subSine | Academy of Electronic Music (subsineacademy.com). (This is my friend’s music school, and we operate from the same premises in the Southside of Glasgow.) Several years ago I developed a curriculum for a 10-week course, which I have taught there semi regularly. I have since reformatted it as a shortened weekend-long intensive workshop. Modular synths are the primary tool I use, but it is a broader look at synthesis in general. Working from the modular building blocks really helps build a strong foundational knowledge.
I’ve used various virtual modular environments and plugins over the years as my main teaching platform. There are many great virtual instruments and VSTs out there, but I have always had in mind to explore the viability of creating an Instruō inspired range of fundamental modules within VCV Rack. It is such a good pairing to, and emulation of, the Eurorack format. Ideal for education.
Enter Murdo Graham and Finn Quicke! Their internship placements were able to go ahead despite the global pandemic. We had some good Zoom chats with Sebastian and myself. Exploring the possibilities of producing analogue models of maybe a couple of the Instruō modules seemed like an ideal task. We now had a new plan and everyone was onboard. I’d like to add that everything here was new to these guys. Not just digging into a new development environment, tool chains and getting acquainted with systems like GitHub, etc. I believe many elements of audio synthesis were new to them as well. They got started and I pretty much left them to it while maintaining my usual day to day operations in-house.
You can guess the rest! It really didn’t take them long at all to get things going, and they began churning out the module builds. Once I got things in place with social distancing safeguards, we were able to all work in-house eventually. More real world comparisons and testing could take place, and it was so much better with everyone in the team getting to know each other. Finn and Murdo were able to start branching out and trained up on the assembly stations and managed their own time on the project while clocking in some hardware building. I took that opportunity to dig into the VCV back end myself, picking their brains as I went. I rebuilt the Harmonàig for the collection with some help from Finn optimising things. It was an ideal project for me to get my head round the VCV development process. It is a very well considered platform.
The announcement video for the modules being available free in software form on VCV Rack:
Weidenbaum: Many if not all of your modules are analog, such as the Cš-L and Saïch. What did you do to make the Rack ones sound as much like the originals as possible? What if any were some of the difficulties you faced in the porting process?
Lim: Many of my designs are fully analogue. Some are hybrid, using microcontrollers to handle UI, controls, routing, and even calibration in some instances. The oscillator cores and all signal paths are analogue, other than the obvious DSP-based modules. I didn’t really know too much about the process or practicalities of analogue modelling so it’s something we explored as development got underway.
Much of my circuit work is derived from classic reference design schematics and from classic synth DIY publications. My background is in music. I began on classical violin and expanded to a range of different instruments. It was during my studies at Berklee College of Music that I started teaching myself circuit design and C programming, following a few introductory courses that were offered in my chosen major. I actually ended up minoring in acoustics and electronics. Eurorack in general is forever building upon a legacy of iconic analogue musical instruments. When developing hardware I learn by doing. I’ll study a circuit from many sources and then build prototypes. Much of the time it’s a case of finding replacement components for older obsolete variants that were used in originals. I then tweak and tune things by ear until it does something I particularly like or it leads me to unexpected territories. I rarely start with a solid goal in mind when it’s something mostly analogue. Digital designs are always a bit different, and there are inevitably some “chicken and egg” moments. The Harmonàig is a good example of this: In designing it I had more of a plan based on my own long developed software implementation of modal harmonisation. The interface came before some of the final feature implementations. Digital often leads itself down the dreaded “feature creep” path!
Once the foundation models were up and running in VCV, we started to consider how best to model the real things more accurately. I believe some of the first modules were the Athrú, Cèis, vinca, øchd, and tanh. The majority of character matching has come through recreation of response curves. There was a lot of graph plotting and mapping of the characteristic responses for particular parameters and behaviours, such as envelope segment timings and VCA gain response ranges. It’s been a fascinating exploration throughout the project as I’ve had to get hyper analytical over very specific details of each module as decisions were made on how best to realise a software model. There are many aspects of the project that I myself am still not hugely well versed in. I understand that the VCV engine calculates things on a sample by sample basis for example, which makes a lot of sense in terms for the emulation of cross modulation behaviours in modular signal processing. A big hurdle is implementation of antialiasing, as there will always be a limit to frequency range at the Nyquist point. Considerably higher sample rates are more readily required, compared to basic digital audio recording/playback. This topic in particular played a role in the research side of Finn and Murdo’s projects. Antialiasing algorithms for the oscillator cores, etc. I pretty much left them to it and the characteristic behaviour modelling came more once we had foundation waveform production in the case of the oscillators. Some aspects of the waveshaping ended up very close to my analogue process, such as sine shaping from a triangle wave. The wavefolding is modelled very truly to the analogue summing network and the filters use software models based on the core circuitry.
The important aspects I wanted to retain were the many imperfections and quirks that come with analogue. We matched many characteristics of the analogue behaviours very closely. Some specific examples being:
▰ The tracking responses for 1V/octave control over pitch: In digital it is very easy to perfectly map a linear control range to exponential frequency response. Perfect doubling of frequency over linear voltage integers is what the original analogue circuits strive for, but digital perfection is too clean musically. Summing parallel voices in analogue/acoustic sounds great because of the complex interaction of minute dissonances. Analogue voices do this naturally by being imperfect from each other. Play two perfect digital sawtooth waves together at the same pitch and you end up with only one louder sawtooth wave. In short, we made it more loose.
▰ Hard sync of the Tòna and Troika sawtooth oscillator cores: Theoretically, hard sync is a forced reset of a waveform cycle. In these modules’ sawtooth cores there is actually a much more loose locking of waveform reset, which is dependent on the charge amplitude level of the core waveform. This is a “weakness” of the circuit I worked around for the Cš-L (which had to cleanly hard sync even at LFO ranges). Sonically the hard sync is very different between these oscillators. The VCV models emulate these variations and behaviours very well! The Tòna/Troika hard sync is a very particular musical effect which has translated beautifully.
▰ System noise floor: The first side by side comparison we tested was the Athrú wavefolder. Wavefolding was a big milestone as so many of my modules feature different flavours of it. The pure folding equation perfectly models the circuit stages. I couldn’t believe how accurate it was! But when A/B-ing the analogue Athrú to its VCV model sonically it was very easy to tell which was which. The signals on a scope were really accurate, though, so it had us scratching heads for a while. My theory was a lack of stability. I added in minute amplitudes of pink and low frequency noise to all the parameters of the VCV Athrú. Immediately it became close to indistinguishable to the analogue!
▰ Frequency ranges of things: My analogue oscillators have a slightly less typical accessible range via the interface controls. Certain cores will run upwards of 35KHz (the Tòna will actually hit >50KHz). They are oscillators for dogs as well as humans. These are my own design choices which give the option of pushing things above the threshold of audibility as well as below. Sample rate effects using S&H is one example of using these high frequencies. In digital, internal processing sample rates are required to be as high as possible for this behaviour to translate. Certain limits had to be implemented to retain a close match in parameter position and behaviour. Muting/stalling things at particular thresholds is required before aliasing becomes severe.
The process of faking a noise floor and instability to the models became a more prominent element to the VCV development as the range grew. The guys put in place virtual tolerance and deviation ranges which we modelled on the analogue originals. These are in place over almost all parameters and functionality. I very much wanted to mimic enough “analogue-ness” in the collection so that if someone were to become familiar with the modules in VCV, getting hands-on with the hardware would be familiar and a more seamless transition. To this end we actually went as far as adding characteristic randomisation between instances of duplicate modules. For example, the first Cš-L added to a new rack will be the same every time, but a 2nd, 3rd, 4th … would have a slightly different characteristic “fingerprint.” This is modelled to the same degree that two of the same analogue module would differ from one another when parameter controls are physically matched. When saving a Rack, the characteristic fingerprints per module will be retained. That is a feature I am particularly happy about! I’m very glad VCV allows for this capability as it means building an Instruō rack in Rack holds some similarity to doing it physically. The collection of modules will be unique to that instance of Rack. The differences are very small of course. Quite possibly indiscernible for the most part, but keeping things feeling analogue was a strong goal from the start and subtlety is key. Considering that every virtual hardware control has a finite resolution, it means that an exact floating point value can be typed in for any knob and it will recall that exact position. If this were a modulator/carrier operator pair of analogue oscillators for example, returning to a specific index ratio is not really a possibility via analogue potentiometers. This behaviour should behave in a similar way, at least between different patches and pairings of the virtual Instruō modules.
Waveforms compared in hardware (above) and software (below)
Weidenbaum: Were you concerned, from a business standpoint, about the use on Rack adversely impacting your sales of hardware modules? Was this digital porting process a costly undertaking for Instruō?
Lim: No real concerns. To be honest, it’s not really an aspect I was considering until close to the launch, as there was never really much of a plan to begin with! I had no starting expectations in terms of scale for the VCV project. It’s Murdo and Finn who have gone above and beyond, and completely destroyed all expectations and targets repeatedly! It was such a massive task. Early on I thought it would be cool to end up with maybe one or two of the simpler modules modelled close enough to then contribute to VCV Rack as a fun little bonus.
Sebastian had actually spent a little time with VCV many months before. He was able to build the very first instances of tanh and Vinca. It was enough of an introduction to the platform for us to know it would be a considerable amount of work. Being able to put Murdo and Finn on the task allowed it to just keep growing and growing. We kept pushing the goal up and up until it was looking viable to actually have pretty much the full product range built in VCV. It was never a premeditated plan to virtualise the modules to then use as a marketing tool. The project more naturally came about through fortunate circumstances so we were able to have a little bit of fun with it.
Initially I didn’t really know what to do with the collection when starting to consider how best to share it. Instruō is first and foremost a creative hardware company. I’ve learned to design within this niche hardware format and have just kept doing it because I love it, and enjoy making music with modular synths. My vague foundation guideline for designing a module is to create a thing that I want to use in my own rig. Invariably I have several ideas and designs ongoing at any one time. At this moment in time, I count eight that have at least the first round of physical prototypes, not including more modules that are collaborative designs. We’re well into double figures all included as we go into 2021.
For a while we were trying to figure out if there was a price point that made sense for the VCV collection. Ultimately I decided to release it free of charge. I don’t believe it’ll take away from sales of the hardware. Any exclusive VCV users who have no interest in buying hardware wouldn’t have been customers anyway, and more and more I read comments from people starting their journey in modular with VCV. Many get the bug and take the plunge and start a physical rig. Even if there’s a slim chance of someone who started their hardware Eurorack journey following VCV, and if they consider Instruō modules as an option due to familiarity with them, then it is totally a win!
I am also very much aware of the year that we as a planet have been going though and how negatively affected so many industries have been. I’ve been extremely fortunate that the niche industry that I develop in has not been too negatively affected by the pandemic. In terms of demand and sales there have actually been some increases, and besides a few hiccups with supply chain delays, I’ve been very lucky. I happened to restock components and materials at timings that didn’t cause any severe production holdups so I’ve been able to continue business as usual.
With Murdo and Finn joining the team on placements as part of their Masters program I’ve essentially been able to assign two full timers to a project that would otherwise have required significant resources and cost for R&D. They were also able to join the team in-house eventually and clock in some assembly time and learn the various tasks in the workshop during the project. I am very excited to say that they are actually both going to be joining the Instruō team in a full time capacity once they graduate! So, due to a ridiculously favourable perfect storm this project came about and grew into something amazing without financial investment. All these aspects hugely influenced my decision. Releasing the collection in this way felt like the right way to do it. It’s a contribution that me and the team can make to say thanks for being able to do what we do.
Hardware module development underway in CAD (computer-aided design) software
Weidenbaum: Did the process of porting the modules result in any realizations on your part about them. That is, did you learn, especially with the older modules, anything by revisiting them this long after their release?
Lim: That was quite a fun process for me. I had to get hypercritical on a lot of fine details of all the modules. Even my earlier modules are still very relevant to my ongoing design and development process. I’m essentially constructing an ever-growing library of circuits, components, layout tricks, and algorithms that I’ll refer to and build upon as I prototype. So it was much more a case of considering interface interpretations.
Details that I focussed on were things like oscillator sync behaviour, inaccuracies in parameter ranges and scalings, and the little characteristic differences between features such as wavefolders and filters. I currently have four instances of my wavefolder circuit implementation, three being built into oscillators. I purposefully tweaked them in each iteration of hardware, giving each their own quirks that suited the instrument.
The linear FM behaviour was a big deep dive that was very precisely modelled. I use the same matched transistor pair and temperature compensating resistor configuration in the majority of my oscillator designs. This allows for the linear to exponential conversion that translates to exponential control over frequency (i.e., 1V/octave).
The linear frequency modulation signal affects the core in parallel to this converter and has some unique behaviours under certain conditions. Pitch will deviate around the fundamental when modulation CV is applied to the circuit. Rising in pitch when positive voltage is applied. If this CV is negative however, the frequency will decrease from wherever the fundamental is all the way down into LFO territories. This occurs over the range of 0V to around -3.4V. The core will actually get down to its stalling point, but will then flip. At this flip point the frequency will jump up to somewhere around 80Hz or so. The interesting thing is that the coarse/fine frequency controls will affect this flipped frequency spanning around 100Hz in range but inverted from low to high! It’s simply the nature of this circuitry when negatively biased in that way. It is a characteristic quirk which translates to its unique timbre when modulated with a bipolar or negative signal with amplitudes reaching/passing the -3.4V break point.
If, for example, an oscillator is patched as an operator pair and the depth is increased modulating the carrier. Every negative swing of the FM will cause a jump to this alternate frequency range. This appears to occur even when modulated at audio rate. We spent a good amount of time exploring a digital recreation of this quirk. The result is a very true recreation of the analogue oscillators’ behaviours under linear FM.
One of the bigger considerations was the fundamental difference in user interface. In VCV you can only point and click “monophonically” to use a relevant term! I personally find mouse controls to be very limiting as a musical instrument interface. In hardware you get really hands-on and can mess with many things at once. The feedback of interaction is much more immediate. Button combinations are also a big thing when it comes to physical control interface design. The Harmonàig in particular uses a good few button combinations for navigating features and functionality. This interface had to be adapted to VCV’s right click, drop down menu which is an elegant solution given the context.
Weidenbaum: Many modules, Instruō’s among them, are hands-on. Presumably many people will use MIDI controllers, in addition to CV control. Did that use case play any role in your planning? Also, did you test out the modules with any touchscreens, like Windows 10 machines, or the iPad as a second screen?
Lim: It’s not something I really considered too much until after we launched. My own use of VCV had been fairly minimal until we got well under way on the port. When I teach using virtual modular platforms I am demonstrating specific patch examples and signal flows. Very often I build soundalikes as examples during classes. After the initial sessions, during which I use a range of my own go-to example patches for teaching signal flow, I integrate more the concepts of analysis/recreation and ear training. I teach techniques and methods for analysing existing voices and timbres and share synthesis techniques and approaches to recreating them. In all instances I tend to have a very goal-orientated approach to using virtual modular environments. I can imagine many people use controllers to get hands on with parameters to use it in a similar way as I would use hardware modular. VCV is more akin to a DAW than other soft synths in my mind, although with a very different workflow.
As I got farther into testing the VCV modules at the later stages of the project, I used a physical system utilising an Expert Sleepers ES-9 module to interface with a virtual duplicate of the system. Fortunately the HP total of a physical ES-9 and a Mordax DATA scope matches that of the VCV scope and 16 channel audio interface! Most of the time I was building patches in parallel on the hardware and in software. I’d then use many channels on the ES-9 to share a single control voltage source to both instances. It was never the most musical endeavour (as beta testing often isn’t!), but it was extremely exciting to hear and see the software models compared directly on a single oscilloscope. I am extremely happy with how true the VCV collection is to the real thing. I believe there will be many instances where it could be indiscernible audibly when compared.
Weidenbaum: Could you explain which license you employed, and how you came to decide that was the right one for Instruō?
Lim: The collection is available as freeware under a closed-source proprietary licence. We did discuss the idea of open-source at various stages but I ultimately made the decision to go with a free give away but closed-source. It’s an immense amount of code that Murdo and Finn have developed. There are things that they have figured out, created, and developed through their own research and perseverance. It didn’t feel right taking their hard work and publishing it open when my own contribution of code to the project was only that of a single module!
There is always the option of opening things up at a later date but the priority goal was to get a powerful and versatile musical instrument functional and available for people to make music with.
Stages of panel design for a module
Weidenbaum: Are there any plans for the modules to appear on miRack, the iOS app?
Lim: No plans at this time. I only recently became aware of the iOS branch of VCV and don’t know a whole lot about it at this stage. I have received some enquiries about the Instruō collection being ported to it but I will need to look into things more. I believe it is a paid for app which might make things more complicated. I really like the free to use nature of VCV, and importantly, its creator’s philosophy. It being cross platform Mac, Windows, and Linux is also a big deal for accessibility!
Weidenbaum: It’s notable that not all the Instruō modules are part of this set, such as Arbhar and Lúbadh. Might they become available later, and would it be part of the “premium” Rack offering?
Lim: It is something we are exploring but it’s very much a question of logistics and practicality. Due to the very unexpected series of circumstances that allowed for the VCV Instruō range to become a thing, it hasn’t cost Instruō anything for development. That window of opportunity is essentially closed now.
In retrospect, it is very hard to even conceptualise what the project’s (meaning the whole Instruō VCV collection) requirements and scale would actually be when starting from scratch. It’s not the case that I started with a plan to produce the virtual collection as a means of marketing. Now that it’s done and they are out there, I think it’s going to be a hugely positive thing for showing off the Instruō modules, and I hope it will ultimately translate to sales at some stage. But if I were to suggest it to myself a year ago… It’s really a project that would require the full-time employed position of a developer for at least a year to bring it to fruition, not even counting Sebastian and my own time contributions in project advising and management. It would fundamentally have been a big gamble on the resulting port being something I’d want to share at the end of it, with or without a paywall.
The Arbhar and Lúbadh are very different animals in comparison to the rest of the range. I have spotted some comments speculating as to why they aren’t there, especially being digital in nature themselves. They are indeed digital, but porting would be far from simple. Both modules run on quad core arm processors. They actually use the Raspberry pi CM3+ embedded platform. These things are a force of power! 1.2GHz quad core with a gig of RAM. There is a lot of stuff going on under the hood on both Arbhar and Lúbadh. I am skeptical as to how efficiently any form of port would run as a module within VCV rack. Not without a considerable rewrite of the fundamental granular and looping engines would it be possible. These modules are also continually under development and improvement. At the time of writing this, there are some significant firmware updates in the works for both modules that are really going to do some amazing things!
Sebastian and Kian (co-dev of the Lúbadh module) have their hands full expanding and enhancing these existing modules’ capabilities. All while we are continually in hardware production of these things in house. Building virtual versions isn’t high on the priority list at present, but I don’t want to say it is completely out of the question either!
Weidenbaum: What’s next from Instruō on Rack? Might you make any Rack-only modules?
Lim: I’d very much like to continue development/expansion of the Instruō VCV range. I will maintain support of the collection as the VCV platform grows and develops of course! There are several new modules (physical) that I have in the works that will add a lot to the range, and they would be great additions to the VCV lineup. It will all come down to logistics and whether or not it’s feasible and makes sense to assign resources to the task.
As I’ve described, the window of opportunity has closed now, so any more VCV development would require investment, and would really need to be a justifiable endeavour in financial terms. The Instruō team is expanding and I think we’re really getting into a nice flow with things. I’m trying to build up a place of work and environment that inspires creativity and collaboration. The job description for us all at the end of the day is creative hardware development and production. I can see VCV becoming a very useful prototyping tool as I go forward. I draw renders of faceplates and interfaces as I design modules. It’s actually these exact renders that became the VCV module aesthetics! So it may not be too much of a jump to go from illustrator render to VCV mockup for seeing things visually in context. It will very much depend on the module’s nature whether any functionality can be developed or tested virtually or not. Once I have a visual mockup I tend to proceed quite quickly to the first revision hardware prototype. There is nothing like holding the actual thing in your hands for feeling how the interface translates physically.
Weidenbaum: Do you have any other digital music-making tools in the works?
Lim: Not software-only. Many of my module designs are hybrid using combinations of analogue cores with digital interfaces and some being digital in nature but in a hardware context. Of the multitude of modules on the development white board, there are some very interesting concepts which will be digital at their core but the interface will remain fundamentally analogue.
Weidenbaum: The Instruō website has a lot of detail about the modules, and not much about the people who make them. Could you give some background on who works at Instruō, what their roles are, and what background experience they bring to the company?
Lim: At the moment we are an extended team of 18. Some of us are full time, others freelance, part time and/or remote working. We’re all friends and friends of friends. I started Instruō solo, of course, working out of my spare bedroom building individual modules by hand. Over time, I met people who I clicked with and we would just start bouncing ideas around over drinks. It’s always been a very natural progression. I’ve pretty much been making it up as I go since the beginning, and found what I think works through practice. I rely heavily on feedback from the team and like to think we operate more as a collective.
I’ve essentially stumbled into a situation where I now manage a company. This is all new to me at every stage. I do actively try to keep things as “non-corporate” as possible. We all come from music, arts and generally creative backgrounds and we now have a platform to work within and make our own. I’ve been steadily renovating and expanding our workshop premises. As we’ve needed more hands on deck, people have come and gone naturally. I keep things very freeform. In non-COVID times people will often drop by to catch up and those with prior experience are always welcome to muck in.
Everyone makes their own schedule. I stock good coffee in house (shout out to papercupcoffee.co.uk) and we’ll occasionally chip in on a keg for the beer tap. Then there is the standard in-house arcade machine, games consoles, and lots of music gear, as one would expect! I turned one of the rooms upstairs in the building into a jam room for everyone to share as a creative space. We ended up staying late for some after work drinks on the last Friday before taking a break for Christmas. It descended musically into some long form jams. I’ve realised now that the modular synth aspect of Instruō is really just a front. I’m actually just piecing together a massive music ensemble!
For the most part in day to day operations we are all in one big room in which assembly and design is going on simultaneously. I like it this way where everyone can chime in and give suggestions when there is problem solving going on. It’s a fun way to design where we can turn round and pitch an idea and have people know what we’re rambling about and give suggestions/ask further questions.
If anyone was at Superbooth19 in Berlin, most of the crew at that time was there with me helping out at the booth. Everyone has an interesting background, some are active DJs, electronic artists, musicians, makers, visual artists, educators. I’ll put myself in front of the camera and take the spotlight when it’s required at trade events, etc. It’s something that is slowly becoming less uncomfortable to me fortunately! I much prefer playing a gig than talking but I’ll put on the “leader” hat when it’s needed. I encourage any co-developers of modules/projects to get as involved publicly when things launch as they are comfortable doing. There are a lot of us and we’ll occasionally chime in on discussion threads and forums.
I think like myself, many of my colleagues lean towards the introvert character trait. Inclusion of the signatures of the developers on the final product is something I think is important. Like billing in a film or credits on an album. Many Instruō modules are my own but there are going to be more and more that will have a range of names visible on the backs.
A few of us have performed together at modular performances and many have their own music already released. Several work regularly in the creative industry independent of anything Instruō related. I like the idea one day of expanding to some form of record label branch of Instruō as a means to share work beyond just the modules.
The website is a hub for info on the products primarily. With the new Learning page that went up with the launch of the VCV collection, I have some big plans for an expansive resource for learning the modules and synthesis in general. The user manuals for each module were a big undertaking. My good friend (and modular maestro) Collin Russell is the author. We go way back! Met and started collaborating while both studying at Berklee College of Music. He’s now Portland, Oregon-based and continues to work with me remotely while being well-integrated into the modular and electronic music community over there. For a while I had little to no documentation for my modules. I made long form videos to serve as feature guides/resources and had some unformatted write ups that I would share as needed. Eventually I was able to invest a bit into doing it right. Collin has done an incredible job! The manuals go beyond just informative. There are musical patch examples included at the end of all of them. I’m planning to expand these to functioning VCV patches for people to try and learn with in addition to reading the manuals.
Weidenbaum: Instruō modules have a very specific look to them. Could you talk a bit about the design that inspired the work Instruō does? You being based in Glasgow, I guess it’s natural for an outsider like me to wonder if Charles Rennie Mackintosh’s design legacy plays any role in Instruō’s faceplates.
Lim: I wish I could claim some impactful inspiration like that and sound sophisticated! I can’t say there is anything like that actively going on behind the scenes, although I do live fairly close to Mackintosh’s House for an Art Lover, so maybe passive inspiration? I have only ever visited it once, though!
The truth is much cruder. More of a trial and error and improvised process. I play around with a hardware component layout first, once I have finalised the control set. This is sometimes before I have schematics finalised, but sometimes after. It very much depends on the nature of the module. I iterate and use past designs as reference. I’ve become very familiar with how spacing will eventually feel and look in practice in terms of spacing and placement, and I have a vague rule-of-thumb for spacing between particular parameters. There isn’t really a fixed grid that I stick to which is evident when seeing a full system of Instruō! This familiarity has built up over time. There are some older prototypes I’ve built in where I’ve quickly realised spacing that doesn’t work! It’s much more like doodling. Certain placements of knobs and jacks will insinuate a line or a shape that I then sketch around that. This first stage all takes place in EagleCAD on a faceplate/PCB dimension template. I end up with many sketch layouts in a single .brd file. Once I have something I like I’ll add annotations and then make the jump over to Illustrator. I map out all the drill holes for the hardware components and then layer up, adding graphics for knobs, jacks, switches, etc. I then play with the graphics and labelling to fill in gaps and define things. It’s very much a case of just knowing when it looks right, but not knowing what that is ’til I find it. I very often don’t have a specific plan in mind. It’s quite similar to how I approach working on a mix. Similar to ear fatigue, it’s easy to get stuck in a loop overthinking a small detail and redrawing things needlessly. I’ll sometimes put days in between sessions working on something. Fresh eyes will sometimes inspire a completely different direction within minutes of restarting. The same as going back to a mix with fresh ears and spotting the most “obvious” detail that you were previously deaf to.
I think over time I’ve dialed in some techniques and solidified the aesthetic of the Instruō designs. Each module is unique but I find myself seeing patterns sooner and preempting the order of events a bit. I think the most important aspect of my design procedure is actually the broad range of tasks I have at hand. As I described earlier, there are always many designs and ideas in my sketchbook at all times. I’ll spend the odd hour or day working at something and then shift gears onto a day of assembly, or programming, or boxing, or oscillator calibration (that one in particular is more and more regular these days). Being hands on doing repetitive work plays a surprisingly important role. It is thinking time. Whenever I feel a creative task is dragging a bit I can be equally as productive on a range of different jobs, and it inevitably helps things get unstuck.
Here the rest of the team, and links to learn more about them:
These are some videos showing the modules in use.