Which MI modules are purely analogue?


Buttery misguided intentions aside, I love my digital (CPU) software tools just fine, but I did get into Eurorack because I thought it was analogue. So even though I have no prejudice against chips, I do think I will try to keep my modular experience as analogue as possible. (guilty keeping to myself that I just bought a used Elements at a great price).

We shall see. It’s as much about the choices we make, and the subsequent creativity the resulting limitations reap, as it is about the music that comes out the other end.


Well the patch cables are definitely analogue


Thing is, even though many modules are digital they still accept analogue audio in and output analogue audio. So you can use digital where it works best and analogue where that works best. No compromises.


That is definitely a thing I enjoy with digital in modular, the modules aren’t adding latency in my DAW by using up CPU. If they add aliasing I can pinpoint where the sound is coming from and tweak it easily, and work hand in glove with analogue modules.


DSP describes the mechanics of making sound through computation. (It literally stands for digital signal processing.) This means everything you described after it all fall under it. (Assuming the digital circuit you reference has a processor of some type in it, otherwise I don’t see why it would be a digital circuit instead of just a circuit in general.)

It isn’t a flavor, it is the main flavor all others are derived from. Just like how Ford is the manufacturer and they make escort, mustangs, Taurus, etc. car models. Those cars are all very different but they will all get you from point a to b.

That said, there are many different takes on DSP in action, but the processor used in those endeavors is not really the main point of distinction. The DSP code (methods) involved are first and foremost, and then the hardware matters, as it is was enables that code, and also limits it.

Lastly the whole is it analog or digital question is still an important question, but it is one that is becoming more and more irrelevant thanks to advancements in both hardware and DSP theory/code.


DSP also refers to a specialised processor (Digital Signal Processor) that has an instruction set and design which is more suited to processing signals (concurrent execution of some instructions), be it audio or radio or otherwise.

Sharc, Motorola 56000 series and so on.

Using DSP to refer to software only solutions is a more recent phenomenon, or at least it is in my world :slight_smile:


Certainly can be. I think for the purposes of this discussion focusing on DSP as the theory/software behind the modules makes more sense.

If I buy an fpga based module it doesn’t imply any type of synthesis or the underlying dsp mechanics. The inverse is equally true and wouldn’t imply which processor the module is using either (minus very specific code bases designed for that specific processor).

I just didn’t want someone to worry unnecessarily about the processor being used in this context when it is the synthesis method being employed that matters. :grinning:


Well, trying to bee not too picky, you might have to take into strong consideration the platform on which you will deploy your code, when talking about sound processing.
Power stability and references, overall design of your “analog path” in all, expecially inside fpga, precision and accuracy of ad and da sections, distortions on vca, both on digital and in analog. And capabilities of your cpu are fundamental also. Think about stability of clocks, interrupt management, availability of a floating point unit, ecc.
We are not talking about desktop pc with all the processing power in cpu and soundcards and layers over layers of software.
An embedded platform is always a tradeoff between manageability, affordability, availability.
And, of course, with all these constraints, the design of both code and analog platform matters a lot. Look at mi modules: many are digital in the processing and/or generation but each one is different, because of specialized functions. Even if you can see some “engineered blocks” reused. One does not have to reinvent wheels all the time.
So,do not think dsp code can run nowhere with the same results, if it has to became “analog” at some points.


Oh absolutely!

You can see I’ve already agreed with you. :slight_smile: Still if you as a customer go to a module, you are not looking at the processor in the module (ask the numerous MI customers what chips are in their modules and very few will know) but rather what the module is capable of doing. For this reason while the underpinnings are interesting and exciting, I find that it really just confuses people when you talk about more than you need to. Of course if the OP finds this all useful, then nevermind me.


The OP is fascinated by this conversation, though it went over my head about 5 posts ago. But I’m glad I asked. This is a lively and bright community. Cool.

What I’ve been upto in the 3 weeks of my modular adventure so far:

This piece is all analogue with the exception of MI Elements. All sequencing is just simple clock dividers strung together and self referencing. No sequencer per se.

All rhythm and melody is generative. Only some of the changes in timbre are done manually.


SuppleWhat, your tracks are very interesting, I like Oh no, OH NO a lot.
You make me feel like I’m wasting my time because I’ more into buying stuff that making music, or noises, or sounds or whatever.
I’ll try to change this starting today.
So, thank you.
@2disbetter, I apologize, I’ve been too much involved in the subtle technical aspects than the musical one. I was in search of 2 seconds celebrity… sorry…


There is no need to apologize, and I apologize if I made you feel that way.


Thank you. Happy to know I’ve offered some inspiration.



An FPGA is different to a DSP chip though. A DSP can only really do what its instruction set allows, it is a processor. An FPGA can become a video chip, a sound chip, a processor, anything you can fit on it that works. Many consoles and computers have been recreated with an FPGA, not software emulations but real recreations of the hardware.


I never said they where the same just that you are speaking about the hardware aspect of dsp which in the context of this conversation doesn’t really matter as the hardware doesn’t specify the type of dsp theory employed.


Hey, do not think fpga is free from constraints that plague software… The analog path the signal has to travel inside the fpga can be, and often is, worst than dsp and microcontroller with integrated specific devices… A good and wise design is needed independently. Nobody has a magic wand by choosing “” “the right” “” ® device.


A little chorus can help too,with plugins or guitar pedals…no?


i think a very good solution to your problem is to get the modules you desire and have some analogue ones to impart a bit of that sound on the rest.
very good choices for this application are: any analogue filter with a bit of character, a serge resonant eq, a low pass gate, an overdriving vca, ears used as a distortion (!!!), analogue delays, spring reverbs and so on.
trust me, in the end a bit of saturation can warm up most things.

also tbh, MI modules tend to sound quite clean but definitely not what you would call sterile in my opinion.
especially the resonators (rings and elements) can sound very organic. yesterday i did a patch with rings which convinced my girlfriend i was playing field recordings of metal drums and bells.