tbDSP: a Eurorack DSP monster!

It is rumoured (mainly by Olivier himself) that Olivier is working on at least one new DSP-based Mutable Instruments module for release some time this year. This is eagerly anticipated by all Mutable fans, I’m sure.

But this thing just caught my eye: tbDSP

It is still under development, but looks to be far more than just vapourware, and the projected price is around the US$600 mark according to one of its developers. It seems to pack a great deal of DSP horsepower behind its panel, mounted on a 6 layer PCB… DIY it isn’t, and it seems that neither the code, nor the hardware design, will be open-source, and the dev toolchains are all proprietary too. And the only demo so far seems to be some chip tune stuff. Overall, a bit like a closed-source Monome Aleph (which IS all open-source, bless it) for Eurorack. Nonetheless, interesting. The spectrogram display looks pretty nifty.

Seems like a more powerful Z-DSP.

@6581punk …except you don’t need a stupid proprietary memory card to get new setups onto it, and a stupid board to hang out of the front of it to program it yourself, which is a massive plus, in my book…


There really aren’t that many Z-DSP programs around, so you don’t need that many cards.

The dev environment is pretty good in that what you change on the screen is then run on the device. Anyone who is bothered about how a dev board looks probably wouldn’t be able to code for it anyway :slight_smile:

@6581punk oh, I’m not opposed to the way a dev board looks- I just don’t think it’s necessary for a device like the Z-DSP to require one in order to program the thing- much better to just stick a USB socket on the thing. You can still monetize 3rd-party content development by selling the software to actually create programs for the device by selling the software to do so.

I’m really tempted by the Snazzy FZ Ardcore? (despite the company’s website looking like it was designed in the mid '90s). It’s essentially an Arduino stuck to a Eurorack module, so you can use the Arduino dev tools, and it’s all open-source, with an archive of ‘apps’ you can download for free, tweak and upload to the module directly from your computer. I’d love to see a more powerful device, with a similar open-source model.


Well, unless you know the architecture then you can’t really say that USB would work. I think it is a pretty optimal design, so it uses simple parts. Where’s the SD card storage on the Shruthi for instance? where’s the USB port on that? it’s a simple design that works.

The Z1 chip is actually used in a lot of guitar FX pedals from what I’ve read. So adding the expense of USB makes the chip and runtime too costly. The only reason the Z-DSP is so costly is the niche market and the fact they’re partly hand built.

The Ardcore isn’t very powerful, it’s a 16Mhz Atmega. So it has less sound generation power than the Shruthi which runs at 20Mhz. It’s a very simple board and it not worth the money they expect your to pay for it. Build one, the parts are around £50 or so tops

DIY boards available here:


You won’t get such a nice reverb as the Halls of Valhalla on the Ardcore.

Would you really want to have to load up the Arduino IDE on a laptop and plug your laptop into the Ardcore to change programs all the time?

There is also nw2s, which is DIY and about to go into production. It is based around an Arduino Due. Presumably it is open-source because it is hosted on GitHub, although I couldn’t find licensing details. There is at least an API to encourage third-party extensions and hacking (in C++). It seems to be targeting mostly modular synth control, but it looks like there is enough horsepower to do some audio synthesis and audio production as well. I wonder if any of the MI code (e.g. Braids, Tides, Grids) could be ported to it? Might be feasible. Anyway, it looks rather interesting.

Looks good, but I believe less and less in general purpose programmable modules. Also, I wonder what would be the power consumption. 3 processors… wow!

The DSP modules I am working on implement a very specific function, with a choice of CV inputs and a UI layout tailored for it.

Yes, I am inclined to agree about polymorphic modules such as these. Actually, they’re not polymorphic, just multifunctional - that’s the problem. I can’t help thinking that a laptop and a pair of Expert Sleepers ES-3 and ES-6 modules to handle high fidelity, high precision, low latency CV and audio I/O between computer and modular is a much better idea. Maybe someone just needs to develop a laptop in Eurorack format…

That said, one nice thing about the tbDSP is that when you touch the knobs (capacitance sensing?), the current parameter to which that knob is mapped is displayed momentarily on the display panel.

The nw2s does look intriguing. I think I stumbled across that project a couple of weeks ago. I like the fact it has Bluetooth built in, so you don’t have to physically connect it to the computer to get data onto it. It also has an SD card socket, so definitely scores over the Ardcore on both those fronts. Unlike the Ardcore (which has too few), it has almost too many I/O connections. For me, it loses points for taking up so many HPs. I might be up for a full kit one day though, if I ever buy another couple of Happy Ending Kits…


Don’t forget that OWL euro module.

To second Pichenettes: i don’t like these “Blank Sheet of Paper” Modules, no matter how great the Paper is. Its like having a Schweizer Taschenmesser when all you need is a trustworthy 3mm Hex Wrech….

> if I ever buy another couple of Happy Ending Kits….
you confuse “if” with “when” !

I don’t agree with that analogy. They are like synthesizers, you can alter the sound source. It might not fit the traditional Euro model too well, but then neither do some of the MI modules as they’re digital.

I also think that there should be a Eurorack Laptop if we want the fullest versatility in one module.

Or maybe a Eurorack Raspberry Pi (EuroPi anyone?).

There is a big grey area between the traditional “one module per function, all analog” modular world and the more advanced digital/dsp modules. I guess the key is what Pichenettes said, keeping whatever is in the module focused on a defined function/use and tailor ins and outs to that function/use.

Well, Braids fits pretty well because when you switch it on it’s the same as when you switched it off. You could just set it to one synth model and cover the screen, you’d be none the wiser that there’s other sound models :slight_smile:

There’s purists and traditionalists in all areas, there’s also people who want immediacy and others who would program a Korg Wavestation SR for hours for fun.

i think the only time these things worked as viable products was with the nord modular and that was souly because the software was outstanding and useable. at the end though, even that didn’t have enough of a user base to keep it alive and assuming you will have hoards of programmers flocking to your product to make patches for it is a strech at best. A good example is the microdec, pretty well designed product and pitched as a make your effect deal but not a single user patch ever materialized. without a software front end that is easy enough to use for the basic, not power user, you’re asking for trouble

@altitude that’s why the Arduino-based modules have a headstart. Even if they are underpowered- having a mature, well-known IDE means they’re far more likely to have an active user base.


again, that’s kind of my point though. You still have to be musically orientated programmer to get into it. That pool of people isn’t too big and you would need a large number of people like that to have a viable product. A graphical interface ala nord modular is the only way to pull in the number of people to make a product like that work and that’s a tall order to develop. It’s a catch 22, make it cheap and technical, you only get a small number of user or make it expensive and easy which means you need a large number of users to recoup the costs…

I think the Ardcore being around £250 for an Arduino board with some pots is the big beef with it. £99 would be a more realistic price.