Random consumer product idea for an MMI device

Modular synthesis makes heavy use of different kinds of randomness (stepped, smooth, etc) for MANY different reasons… Although the most popular modules only sell in the amount of maybe the low 5 figures, the modular synth fan base has disposable income and gets VERY excited about adding new modules to their racks, especially ones that are based on esoteric scientific research. This is just a fact of the market right now.

Because of this, if we had a psychoresponsive device with high bandwidth but a small power footprint, that could be smd soldered onto pcbs and turned into modules fairly easily, it might be possible to bring MMI to a niche consumer market…

I am just imagining a module powered by a QRNG that outputs triggers or control voltage based on the types of randomness provided by this hypothetical micro QRNG suitable for embedded projects. You could use control voltage or a knob to adjust the bias, and other variables that affect the randomness.

Unlike other applications of randomness, I truly believe that modular synth players would be more open minded towards the application of MMI, as many of them already feel connected to their rigs, and being that there is a large viral ecosystem on youtube of modular players, the virality of the module being so unique would get it to sell simply because most modular synthesists do whatever they have to do to put their hands on a new module.

Controlling the chaotic voltage of a module with intent would be cool because, you could possibly influence transposition changes or subtle changes in melody in key with a quantizer.

There are also a lot things that can be done with random gates ,and self patching to have the randomness affect the module itself, etc.

When I originally thought about the subject, scott had this to say:

To use MMI in place of the builtin random generator requires understanding the function of the random numbers in the module. In addition the MMI generator/processing should provide the maximum effect size so the result would be noticeable. This would likely be a continuous (not initiated) form of MMI, which is the hardest to get a large effect size from. Just providing random numbers from ANU’s online source would hardly seem sufficient to achieve noticeable, real-time effects.

As always with MMI systems, real-world testing of various designs is necessary to determine which provides the best (the most interesting, engaging or pleasing) results.
Let’s take a look at one of the most popular RNG modules, the

So, years later, I am starting to understand what he meant by that, and seeing that it indeed would be possible if enough work was put into the correct project.

Scott also had this to say:

The PrdCore.dll (not quite up to date yet) already uses the PCQNG software-enabled MMI generator (on Intel compatible PCs) that could provide a psycho-responsive (MindEnabled) output. Not the top of the line MMI generator, but readily available and could be at little to no cost. Data could be processed in a simple algorithm in the host computer and output an analog signal through a speaker output.

There are a few digital simulations of modular synths available for free, so this should actually be testable with some tinkering using VCV rack

Anyway I know scott has been focused on consumer products that use MMI, and I think the synthesis crowd (which is growing, and dedicated, but small) is a great test market for if users would employ MMI tech, as many of them are hacker and makers themselves.

I know someone already made a quantum random source of CV module using Zener diodes, so I am betting (especially with this zero energy switch type thing) this would be a viable product, if production volume was high enough.

To give an example, lets compare the Make Noise Wogglebug.

Integrating MMI means users could potentially influence the module’s parameters directly through their mental states or focus, such as adjusting randomness intensity or selecting modulation patterns without physical interaction. Imagine molding soundscapes or rhythm textures simply by shifting your concentration or relaxation levels. The incorporation of a QRNG would elevate the module’s capability to produce truly unpredictable and superior-quality random voltages. The inherent indeterminacy of quantum randomness provides a layer of complexity and uniqueness in the modulation signals that classical randomness can’t achieve, enriching the module’s ability to generate intricate, evolving soundscapes.

The module could also adapt its modulation patterns based on the user’s emotional or cognitive states detected through MMI, creating a direct link between the musician’s internal state and the sonic output. This adaptive behavior would allow for highly personalized performances and compositions, where, for instance, a calm state might induce ambient modulations, while a focused state could trigger rhythmic patterns. Moreover, by reading inputs from multiple users, an MMI-enhanced module could enable collaborative sound creation, turning audience participation into a direct influence on the sonic result, opening new avenues for performance art.

The introduction of machine learning could enable the module to adjust its response patterns over time, aligning closer with the user’s preferences and typical mental states, thus enhancing usability and reducing the learning curve. This would not only make the module more intuitive but also minimize the need for physical controls, leading to a cleaner, more streamlined design that could be more accessible to users with physical disabilities or those seeking a hands-free interaction with their synthesizer.

The integration of Mind-Machine Interaction (MMI) with modular synthesis offers a fascinating new way to create music, directly linking thought to sound. This approach not only captivates users by demonstrating how our mental states can shape music but also educates them about MMI’s potential. Seeing thoughts and emotions translate into sound in real-time can spark curiosity and inspire further exploration into the science behind MMI. This immersive experience demystifies MMI technology, making its complex principles accessible and engaging, and showcases its practical applications in an artistic context.

Hi Joshua, thanks for your detailed post. The most responsive MMI system is currently the MED100Kx8 generator with the output processed by advanced processing methods (fully described here and elsewhere). The MED100Kx8 is the size of a thumb drive and plugs into a USB connector. Then the 100KHz output is processed to produce a usable result, which is currently done in the host computer. This is doesn’t sound exactly like the miniature module you suggest, but it would be a starting point. Theoretically the same FPGA that produces the random numbers could also be programmed to output fully processed MMI results. These results can be in whatever form is desired: analog via DAC, or binary bits at desired intervals, etc.

The new hardware I am working on includes a microcontroller, which might be able to perform all the processing to provide the MMI results. However, the current version is 2.5 x 4.5 inches (PCB) – not huge, but still not a miniature module. In addition, the sensors are so sensitive, they could not be placed within a highly noisy digital environment. Decades of research suggest that shielding would not disrupt the responsiveness of an MMI generator. However, as the designs become even more responsive, the effect of shielding may be different than what is currently believed.

For practical reasons, the device should not be targeted to be highly miniaturized, because that is very difficult and expensive, and is a stage of development that comes after fully demonstrating functionality and market demand. Eventually, when MMI is used everywhere, custom integrated circuits will make them much smaller, less expensive and even more responsive.

One of the secrets to better responsivity is the use of a large array of sensors, which would be much easier with small, inexpensive custom circuits. That type of – perhaps massive – detector array is probably years away, but an array of my new device, though it would be physically large, could demonstrate the desired functionality.

Thanks for the response Scott, this cleared up a lot of mystery for me. You generously loaned me an MED so I am aware of its format. There are modules that are rather large, even some containing raspberry pi sized computers, but for obvious reasons most of those are “0hp” or zero horizontal pitch, and kept outside of the rack. So while right now the idea is outside of Dieter’s Eurorack specification, I don’t believe the consumers interested in buying and playing one of these instruments would be put off by having it enclosed and shielded in its own skiff, which is quite a common thing to do, I own a couple myself for certain modules.

Further, someone with 4+ racks of modules is going to make room for an interesting module, regardless of size. I guarantee beta testers will try and find away to rack the shielded device, c’est la vie.

Basically this does not sound like a bad limitation for a 1.0, cand indeed there is a lot of noise in a modular rack, so shielding may be necessary for certain components. Creatiing a 0 horizontal pitch (thing outside of the rack that interacts with the rack)

Dieter’s Eurorack specification:

The physical specification is based on the Eurorack) standard of:

  • 3U (5.06 inches or 128.5 mm), where height “U” is measured in rack units, rounded for a lip (nb: 3U in standard rack units would be 5.25 inches or 133.3 mm).
  • 1HP (0.2 inches or 5.08 mm), where width “HP” is measured in horizontal pitch units. A card width is generally integer multiples of 1HP, although some manufactures work in multiples of 0.5HP.[21]

This rule is of course often broken, and manty single use computer type modules are around that size 12.5hp. Point being if it were rackable, the people with the rigs worth houses would make room for these, modular people love new ways to interact with their gear.

Miniaturization was a big trend in modular for a while, but now there are cheap cases made in Asia that compared to the case the modules may as well be made of gold…

I’m very interested to see your new invention, and keep studying the classic modulars for chaos, uncertainty, and randomness, and probably start experimenting with single features digitallly like in reverb and delay, As well as looking at random shift registers that act as sequencers.

People with large cases that could house 13hp modules are usually very invested, and often participate in funding campaigns for open source (or at least somewhat open) module companies. It would be a pretty easy sell because there is a community of “Synthfluencers” globally that highly manipulate the market by giving good reviews on gear in return for free gear.
Anyway, am getting ahead of myself. This sort of thing has been on my mind for a long time, and I am now realizing the potential of some of the different things randomness does for synths, and the potential MMI could have on creating new ways to interface with their instruments.

I think my first task wiill be to take one of the classic randomness generating modules in VCV rack, mod it it to connect to MED, and see what happens.

The modular synthesis community, with its open-mindedness towards new technologies and experimental approaches to music creation, presents an ideal testing ground for MMI technology. The ability to influence sound and modulation through cognitive and emotional states would not only offer a new dimension of musical expression but also foster a deeper connection between artists and their instruments.

Introducing MMI technology into modular synthesizers could also serve educational and inspirational roles, demystifying the principles behind MMI and inspiring further exploration into its applications in art and technology, and although it is not uncommon for makers to only create like 100 modules at a time, the last 10 years modular has become increasingly popular on social media platforms.

Another example of a single purpose computer is the norns shield norns shield | monome/docs

This way this company works is extremely interesting from an open source POV but thatts another topic. Basically they invented those grid based MIDI controllers with the LEDs to show states, and just gave it away basically. And if a large company uses the tech and tries to credit them, they ask to be removed from the promo. Anyway.

This device is basically a raspi music computer with a few knobs and buttons and with a 3.5mm to MIDI converter that can talk to other workstations via cv, midi, and I believe even i2c. There is also the newer module crow, which is a programmable module that exists inside the rack, allowing the norns shield to easily exchange voltage control with the rest of your rack, so the format is definitely there for something like this.

I used to think that an MMI light switch would be a killer app, but now I am thinking that something like an MMI based randomness synth module could clearly introduce people into many aspects of MMI and randomness. And the modular synth enthusiasts, many of whom already believe that there is some mystical connection between them and their rig, and religiously learn the function of their modules inside and out, would be lining up for a pre-order if I could come up with a working example.

I think this market would be one interesting place to expand consumer adoption and education about MMI

Just an update on this, I finished the easy part of seeing where MMI could enhance the top 3 randomness based modules, now I am working on integrating the MED into VCV rack so I can test to see things like, if putting an pseudorandomness through a quantizer sounds perceivably different than putting MMI cv through a quantizer. This is a primitive form of stochastic sequencing, and if MMI indeed really is a way to look for patterns, perhaps you could get a different kind of pattern from an MMI based stochastic sequencer, otherwise they sound pretty random.

There is also the possibility of comparing an analog shift register vs a quantum one. If you could run random cv through a quantizer and introduce a random note in key based on mood or intention, that would be pretty cool.

Much testing to be done, but I fell down the synth rabbit hole years ago so there is plenty of time to obsess.

One thing I realized when doing some writing regarding the subject is that the modules should probably have a way to teach the operator how to mentally influence them, using either an LED or a sound as feedback. The possibilities are endless here but it will take a long time to figure out what works. Luckily the ideas can be tested for free on VCV rack.

Once I have some experiments going I will approach the DIY synth maker scene to see if anyone is interested in collaborating, and that just may bring some interesting and knowledgeable people here.

It seems likely some variations in processing will be used depending on exactly how the randomness is used. Let me know when you ready to go with a specific application or applications for MMI randomness, and we can discuss how to get the highest responsivity. Two processing methods use continuous processing and provide an analog output, and binary results from initiated results (in response to a keypress or other trigger). These two examples illustrate how different the processing might be.

Note, it’s also possible to use some processed feedback to trigger or control the MMI section, including measurements of physiological parameters such as Electrodermal Response (EDR – formerly GSR), Heart Rate Variability (HRV), Electromyography (EMG), etc.

Thanks for the support Scott, your advice is invaluable. As I researched this week I had a newfound respect for the sheer effort you have put in to get MMI this far…

I did some market sentiment analysis on a synth DIY forum, where the audience was both interested in MMI and had some feedback on how they would use mental intention in their synthesizer modules, and in fact many who responded even said they had this kind of idea noodling in the back of their heads for a while. I tried to keep it hypothetical as not to cause an argument about the scientific validity of MMI, but of course there was some discussion about the reproducibility of MMI. Nonetheless the suggestions were fruitful, and here are some findings summarized:

  1. Exploration of Novel Controls: The notion of using MMI to add “virtual limbs” for musical expression suggests a desire for more immersive and expanded control interfaces in electronic music. This could lead to MMI modules that offer nuanced control beyond traditional physical inputs, potentially revolutionizing performance dynamics.

  2. Accessibility: MMI technology could serve as an invaluable tool for individuals with physical limitations, offering them a new avenue for musical expression. This aligns with broader trends in technology aimed at increasing accessibility and inclusivity.

  3. Scientific Skepticism and Curiosity: There’s obviously skepticism regarding the scientific validity of MMI, with calls for replicable, falsifiable research to substantiate its claims. Yet, there’s also a strong interest in the potential applications of MMI, even if currently considered within the realm of pseudo-science by some. This underscores the importance of rigorous scientific inquiry and validation in the development of MMI technologies, as well as the importance of the device to not only work as intended but to educate and familiarize people with the subject and train them in the art of focused mental influence.

  4. Emotional and Biofeedback Integration: The idea of translating emotional energy or biofeedback into control voltages for musical expression illustrates a desire for deeper interaction between human emotional states and musical output. This suggests potential for MMI devices that harness biofeedback mechanisms, to create responsive musical interfaces.

  5. Practical Implementation and Interest: The mention of existing modules like the Soundmachines BL1brainterface and various EEG-based projects indicates a tangible interest and existing groundwork for MMI in modular synthesis. This suggests that while MMI technologies might seem futuristic, elements of these concepts have already been explored and implemented and offer a glimpse at what is currently out there and for what products may have consumer interest.

6… Creative Possibilities: Ideas like controlling chord progressions, and even whimsical applications like sonically materializing emotions, illustrate the broad range of creative possibilities envisioned for MMI technology in modular synthesis. This diversity of thought highlights the potential of MMI to open new avenues for creativity and expression within the modular synth community.

This opens the possibility of many future modules, both simulated in VCV rack (which can be free or cost $) and produced as physical modules, like:

a Mind-Modulated Oscillator which alters oscillator frequencies or waveforms based on the user’s concentration levels, measured through MMI. Continuous interaction is facilitated by real-time signal processing, with LED feedback indicating successful modulation (e.g., green for high concentration, red for low)

or a Mind-enabled sequencer utilizing MMI to capture mental influence , this sequencer could change patterns or tempo, quantized or unquantized to the overall key of the rest of the accompaniment. Timed trials challenge users to experiment with mental intention to unlock new sequences, with color-changing LEDs providing feedback on state changes.

Mind-Enabled Loop Sampler : A sampler that loops audio based on MMI signals, with parameters like loop length and pitch affected by different mental states. It would incorporate timed trials for users to achieve specific mental states to unlock new sounds, with multicolored LED feedback reflecting success.

Overall the research was very positive, with many users requesting that I return with updates as the prototype develops. I would say the modular synth community would be very receptive to these experiments, and if a working device were ever made they are a community with expendable income to make a purchase or donate towards a gofundme for this kind of thing, and I would be overjoyed to do the marketing, fulfillment, customer service, etc as a boutique modular shop as I already run a business selling modular synths and accessories.

The name escapes me but a famous modular artist once said “Control voltage can do three things, go up, go down, or stay the same.”

So I think a good direction for a prototype would set it up to initially manage a control voltage. Then you could patch it in anywhere you wanted to try, a filter cutoff knob or pitch of an oscillator probably providing the best feedback.



Hi Joshua, Thanks for your kind words.

We may define the task: to produce a continuous output (control voltage) responsive to mental influence that affects sound generation. One or more advanced processing methods may be used with the output of a MED100Kx8 MMI generator to make it into a continuous number/voltage to be used as a control voltage. User feedback, which is essential for effective results, would be provided by the sound output. For this reason, the control voltage should be used effectively real time (latency < 1 second, but preferably about 0.2 s). The output can be uniformly distributed so there are equal odds of producing any number between 0-1 (or -1 to +1). Or, the output can be in the form of a z-score, which gives a nonlinear value above and below 0, making it harder to produce larger deviations. I think the uniform output would be the better choice, but only based on imagining how it will affect the results.

sounds good, it’ll take me a while to process this information, hopefully around a week. thank you for your feedback.