MPE/touch controller options?

Equipment
61

I would go at least as deep (the smaller dimension) as the grid, or else your manta pads are going to be really little.

One thing you might consider is the Bela Mux capelet. Gives you 64 individual analog inputs. And then you’re building on top of Bela, which is an amazing platform. Then again, I don’t really know what your budget is, and Arduino is cheaper.

A project like this is pretty much what I intend to do with my Bela and Mux capelet. And since it’s built on Bela, I’ll also be able to include on-board DSP. Should also be possible to build in MIDI/MPE output for integration with other stuff. Hadn’t really considered HID output until now…

1 Like
62

Hmm, I hadn’t thought of incorporating some ‘heavier’ embedded brains into these kinds of projects. I’ll have to tally up to see how much IO I need, as maybe I can manage it with a mega or something, as the beagle+bela+cape might take up too much of a physical footprint for it to be as ‘short’ as the grid/arc.

Cost isn’t super crucial as I only plan on making the one, but cheaper is generally better if the same will do.

That is a good point about on-board DSP too, as I can handle whatever low(er) level signal conditioning that needs to happen on there.

I definitely want to go general purpose with it, otherwise it would be great to build a proper bela “instrument” out of it.

1 Like
63

Yeah, I guess I’m less concerned about size than you are.

64

Heh, my “it can be smaller” knows few bounds…

(I have some kind of bag/container fetish that seemingly requires me to have a container that’s exactly the right size for what’s going into it)

4 Likes
65

I think I probably spend a lot less time on the road than you do. Fetishes usually have their roots in some kind of experience. :wink:

66

The LinnStrument runs on an Arduino Due (the code is open source and very clean if you want to reference it - https://github.com/rogerlinndesign/linnstrument-firmware).

Edit: You could also try using a Bela, I have one and they’re pretty awesome. They have a lot of CPU and run a whole tweaked Linux OS. Having lots of I/O without having to roll your own circuits/boards around that is nice. But you’re right in that the size might be too big. But then again will doing it with an Arduino be significantly smaller? (unless you want to roll your own pcb as well). Either way I think you’re going to need some sort of multiplexing going on to handle all the I/O you want.

1 Like
67

Darwin interviews Geert.

2 Likes
68

I really need to try Cubase.

69

That’s really interesting and insightful. Aside from the MPE-centric part of the interview I really like how immediately shot down the separation of “writing firmware” from “being on stage”, in both the literal sense of him talking about and sharing his code publicly, but also in the sense that that is part of the music making too. It’s not a distinct thing.

1 Like
70

A technical question for those of you that know (more) about this stuff.

When talking to a friend today he pointed out that stuff like the René and Pressure Points are “monophonic” in terms of continuous control, which is something I hadn’t considered before (vs something more MPE-y like Manta/Soundplane/etc…).

Does anyone know how the capacitive multitouch is achieved on a Manta? (or similar polyphonic touch controllers).

71

It’s different for each device and involves a combination of hardware, firmware, and software.

I don’t believe the Manta is an MPE device.

If René or Pressure Points were going to be polyphonic, they’d need individual output jacks for each voice. And if you want separate jacks for different components of a voice (pitch, volume, timbre, etc) the number of jacks multiplies. You start to see why CV-based systems are usually monophonic.

That is not to say that a monophonic instrument cannot be quite expressive. In the interview with Geert he talks about how they always start with monophonic interactions and get those right before they layer in polyphony.

73

Yeah that makes sense. I guess I meant whether it was having 5v (or ground) on one strip and having an analog pin on another strip (though I guess that would be monophonic too with the voltage being shared across the connections?)

I don’t think the Manta is “MPE”, but I think it reads/sends different touch values per pad (I could be wrong here).

Can definitely do a lot with monophonic control, just mainly trying to brainstorm what the hardware side of a controller like this would be (i.e. could it be a Teensy 3.6, or will it have to be something with more oomph like a Bela)

74

This might help you:
http://www.cgs.synth.net/modules/cgs86_trk.html

1 Like
75

That’s a very minimal circuit for a monophonic touchplate keyboard. I once ran across a fairly technical description of how the Linnstrument works, and after I get back from the grocery store I’ll dig that up.

1 Like
76

There is a very minimal description of the Linnstrument hardware in this talk. Mostly Geert talks about the Arduino firmware, which is his baby. I could have sworn I have seen more technical description of the hardware somewhere, but maybe I dreamed it.

2 Likes
77

@jasonw22 perhaps this is the link you were thinking of : http://electronics.stackexchange.com/a/155690/10810 , I posted it a while back. (its something I stumbled on when the linnstrument was very first released)

1 Like
78

@TheTechnobear how open source is the Eigenharp these days? (I noticed that you’ve got an active branch of the EigenD software.)

I found the following statement on their developer website:

The Pico Agent has some binary components that contain important DSP code connected with our sensor design that we have patent applications running on and do not intend to release. They represent a lot of R&D and investment for us and their absence makes little difference to the utility of the open source release.

Is that true?

Particularly:

their absence makes little difference to the utility of the open source release.

For me the utility would be that I could keep the software working even if Eigenlabs no longer exists… (see Pebble…)

79

Yes, perfect, thank you! And Discourse is telling me I’ve posted more than 26% of the replies here, so I’ll hush now. :wink:

1 Like
80

@sam yeah, so the pico ‘key decoding’ is closed, though Ive been given access (under NDA) to the code, which has meant I’ve been able to build and release binary libraries for arm / 64bits variants etc… so Eigenlabs have been flexible and helpful - perhaps one day it will be open sourced.

(having seen the code, I can assure its not something you would want to tweak, its down to how the sensor data is converted into useable stream of key x.y.z data, only useful if you want to manufacture the sensor/keys, hence why Eigenlabs are sensitive about it)

it should also be noted, this is only true for the pico, the alpha and tau do not have this ‘issue’, as this code is in the firmware of the harp - so the usb packets come to the computer already decoded, so all code for them is open source.

though, Im (obviously) an advocate for open source, so would prefer it be open source, I think its true this is not really an impediment.

ALSO… Im actually working on an ‘embedded’ type solution, Ive got a Beagebone black running the Eigenharp (and soundplane ;)) ‘drivers’ to midi/osc, so I can if I wish ditch the computer, and view them as controllers spitting out (mpe) midi / osc.
(this is then pretty similar to , say, a Linnstrument, which just has the microcontroller internally)

anyway, I still love both my Pico and Alpha, and whilst I use other controllers , I’ll always keep these - they are so unique, I suspect its unlikely we will see anything similar in the future, which is a real shame :frowning:
but the upshot is, I’ll be here to support them … moving the software forward!

4 Likes
81

Thanks for the detailed response @TheTechnobear

It’s shame that it’s just the Pico that’s got the closed bit, as that’s what I’d be interested in, they seem very affordable secondhand. I’d been thinking if could get a usable event stream out of the device that it would be nice to output that (monophonically) to an Expert Sleepers ES-3 in my modular, it’s conceivably something that might have worked on an RPi too.

You’re right I wouldn’t want to fiddle with it, but I could absolutely see the need to recompile it in the future. In the end these devices seem to require a large investment in time, and for me that means needing to feel confident that I can maintain the software going forward.

Just out of interest where are these binary blobs in the GitHub repo?