#321

I like this idea @mzero! You could try to use an analog switch ic like the MT8816AE (8x16 matrix used in the SSSR Matrixarchate) or go for a MAX11300 to also have 12 bit adc/dacs? Not sure how noiseless tho.

#322

They make a lot of different parts… I’m currently eyeing:

  • MAX4571 or MAX4573 - 11 clickless SPST switches controllable via I2C or SPI - $6.50 in quanties. Need 6 so - $39
  • MAX14662 - 8 SPST switches via I2C or SPI - $3 in qty. Need 8 so $24 - but not clickless, so that needs some additional components (extra cap and res. I think?)
  • MAX5410 - 2 digital potentiometers with zero crossing detection and mute - $2 in qty… Need 32 so $64 - and could have volume control per matrix point

So yeah - as I said, BOM cost would be high. This device, even if made maker friendly, is going to be a bit spendy.

2 Likes
#323

I’d be down for this as you’ve specced. I like the MIDI on 3.5 (I have several MIDI sequencers that I prefer over computer) and love the LPF-instead-of-trim. One or two CV in and assigned via the X key would be sweet but I wouldn’t want to burden your design which is nice.

1 Like
#324

I don’t know if we’re getting anywhere but can I just say how wonderful and inspiring and just ACE this thread is? Thank you all for your insight and openness and sharing and imagination.

25 Likes
#325

This is one of the most fascinating threads I have engaged with on Lines, although I can’t say why exactly…

2 Likes
#326

Absolutely agree, this has been a source of daily inspiration for me.

I believe one point we seem to be close to consensus on is digital control over analog mixing/filter circuits. Is that correct? Like it does seem like we are at an impasse on the specifics of the layout and function. But maybe we could agree on some specific details and see what emerges from that?

I may be totally off base on this. It would simplify some aspects, like much of that could be controlled via Teensy or STM32 development boards. I might start experimenting with this soon.

What analog mixing/filtering circuits seem nice for this application?

2 Likes
#327

this is most interesting to me. i’m already using my er-301 + 16n as a flexible mixer. while digital control over analog circuits sounds good to me (is it cheaper/easier/more efficient?)… i’d say that a fully digital small box with a lot of 3.5mm jacks and a usb port is also just fine by me. maybe there’s a prefs file that lives on the box and in that file we specify our preferred routing/midi mapping. sorta how on the motu interfaces you can save a cuemix preset to the unit and run it standalone. the difference here would be midi (and sure, i2c) control over the selected parameters.

5 Likes
#328

Hey loadsa crazy stuff being channeled in this thread! Just had an idea to throw in the mix here…

Since many high end consoles have a fancy stereo compressor/tranformer section thought it would be hella cool for our mixer to have a master bus that could incorporate some Colour Format analog cards with some wet/dry and gain controls…just so our mixer can have some customisable analog tricks and flavours.

8 Likes
#329

okay, more quickly cobbled together drawings to spur conversation

here’s some 1/4” jacks in a grid, heavily inspired by the crow design. 8-in/8-out, but other counts and configuations are maybe more ideal.

no usb jacks on the sides shown, but the idea is this would need external controllers and some script/text file configuration. could obviously be more robust (crow-like, running lua scripts locally) or could be simpler (text file controls audio routing, nothing more).

image
image2732×2048 361 KB

also maybe this is the right way to approach modularly adding channels, since each vertical ‘strip’ is so straightforward.

personally, I like this. either this or some of the more ‘mixer-like’ pure analog designs are what get me excited. stuff that sits in-between leaves me feeling a little empty, but I might be alone in that.

10 Likes
#330

I love this idea and the crow-esque design, however, i would likely want some sort of visual feedback/ led situation. maybe even decoupled led’s like grid/arc :slight_smile:

4 Likes
#331

I recently learned about this the other day; http://midimuso.co.uk/index.php/cv-12/

I wonder if we could take inspiration from it, and design a board which has a bunch of headers on it, where users could attach whichever control surfaces they wanted (faders, buttons, encoders/potentiometers (?)), which would be assignable in software? I imagine certain things would have to remain fixed, like the number of inputs/output channels.

I’m trying to think about ways a mixer could be designed to be as open as possible, but leaving as much room for flexibility, while also remaining in the realm of hobbyist DIY-able. Maybe not having a fixed layout could help with that?

#332

this seems pretty perfect
I don’t think I would want LEDs as I’d be using a controller which would reflect the state of the mixer but I suppose some people might be controlling it via software or something else

#333

I think having LEDs that indicate levels would probably be really helpful (even just a single one to show clipping), but I’m unsure about other assignable LEDs.

if each input and output had a single led (which could do metering, but also other stuff), would that be enough?

might be worth thinking through some basic routing/apps/scripts for this kind of assignable mixer and see what we think would be required.

1 Like
#334

that’s enough for me. :slight_smile:

just enough feedback to operate without an external device, should i want to :slight_smile:

2 Likes
#335

This would be huge for me… not a deal breaker per se, but definitely nothing out there that does this.

In that same vein, one more thing: is it possible/feasible to implement a high-quality A/D monitoring the input channels? This might open possibilities for scriptable dynamics engines, triggers, etc.

#336

I like this design. It’s exactly what I had in mind when I mentioned a simple minimalistic box earlier in this thread, it’s really cool to see a nice render :slight_smile:

When you say it has no USB port, is it that it’s not visible in this image or it has no port at all and communicates via OSC ?

FWIW, I second the idea of having some discrete programmable LEDs for a visual feedback.

1 Like
#337

I didn’t draw any USB ports, but it would certainly need them.

2 Likes
#338

I’m trying to understand this direction.

If what people want is n-in / m-out digital audio into some processing element that can do arbitrary mixing and signal processing, controlled from MIDI and/or OSC… Why aren’t the numerous existing ways of doing this working for you?

One can easily find the hardware to do this. Two quick examples:

  • MOTU ultralight mk4 (8 in + 10 out) and an Intel NUC
  • Audio Injector Octo (6 in + 8 out) and a Raspberry Pi 4

Either of these will run SuperCollider - and have more than enough power to mix the inputs with EQ. SuperCollider is easily scriptable (in sclang) and natively supports both MIDI and OSC. Once set up - either can boot and run without display, keyboard, or mouse - and respond via MIDI and/or OSC as you see fit.

Since I don’t see a lot of people doing this with the existing solutions… I’m wondering what about the “just ins & outs” designs in this thread excite people. Is it?

  • Small, single enclosure?
  • Precoded / scripted for mixing?
  • Anticipation of cheaper price?
7 Likes
#339

I think (not sure) part of it is the idea of the mixing remaining analog (under digital control)?

#340

I think all the potential benefits you list are valid, although the single enclosure bit is the most compelling to me.

the more interesting bit to me would be finding better ways to make routing and customization of the behavior accessible. maybe lua is the right path for that since it dovetails so nicely with the existing ecosystem (and the skills people have been learning), maybe there’s another, better path.

my perspective on everything in this thread (which I fully acknowledge is limited) is trying to find the right place where the interface (direct and indirect) feels like it will present unique creative interaction.

this is also why I am completely unmoved by every objection in this thread that insists that the thing being talked about already exists if you work out the spec sheet. I don’t think this effort is about bringing the right spec sheet to bear at all, nor do I think that we’re trying to find unique ‘market fit’, this is about figuring out what the community would be excited to use.

2 Likes