Відео

Indeed, it is. The VCV Rack module is an interface to the circuit board shown. The synth uses VCV Rack for control voltages and the voice cards provide an analog audio path for modulation & output. This makes for a pretty cool hybrid since you can configure any sort of control options with VCV Rack. This is the fourth voice card I've done, you can find examples of the other voice cards (oscillators, filters) in my other videos.

Thanks! Still a work in progress but it's getting there.

Latency is set by the application under Linux/ALSA. And Pi's do support I2S. I don't know much about the XMOS products. I'd expect they have full audio bandwidth whereas what I'm using has a 4 kHz sampling rate, as that's more than sufficient for the use case. Can you describe your use case?

Dang, no joy there. The setup works for me with MacOS and Linux; Windows is a bit of a blind spot as I don't have a Windows machine to test with. I'd assumed it would be compatible but that could be a poor assumption. Can you get Windows to recognize the Gadget if you set it to two channels at 44.1 kHz? If that works then try increasing the channel count and lowering the sampling rate til something breaks.

It fits really nice, doesn't it! The samples are randomly selected from a bunch of (mostly) 80s video games & movies. I need to do a newer demo of the entire machine, it's pretty fun.

Thanks for the feedback and you're right on the lack of hardware detail here. It's really tough to put these videos together for a generic-ish audience. With this first video I kept many lower level details out of the picture and focused on the frontend UI which I felt is the most accessible part. The second video discussed aspects of the Raspberry Pi. I see one more video in the series that goes into the hardware layer, that might be what you're looking for. If I started the first video with "the Pi's 12 MHz SPI is level shifted by the backplane with an 74HCT245 buffer to go from the Pi's 3.3V to 5V and routed to voice cards 8-channel DAC where analog signals go through a reconstruction filter at 1.5 kHz" I think I'd lose everyone pretty quick. Look at the schematics on github for that :) Really though if there are specifics you'd like to see let me know, I'm thinking about what should go into this third video as it should cover the hardware side of things. Not sure about the latency comment. Latency is quite good as it's basically the latency of a soundcard with a small buffer.

@@UltraTroninator thanks for the reply! what I missed was the"functional block diagram" In fact I was surprised you delved into quite some detail before the full picture. About latency, I think it is bad for realtime human-triggered note events. Generated and/or sequenced is fine. The fact that samplerate is down to 4ksps may mean that if you have a fixed minimum buffer (byte)size at any stage, it'll mean more latency(12 times at worst!) than usual audiorate.

Glad you found it! The form factor is definitely a thing. I'd like to get towards eurorack form factor, that ought to be possible and would likely be the easiest for users to adopt. Additionally with eurorack it could leverage an existing power supply, so no need to include that in the solution. I like the idea of making voice card changes easy, that said the voice cards are not designed to be hot swappable. While I don't have Ableton, there shouldn't be any reason why it wouldn't work with Ableton. VCV Rack would have access to the same hardware interfaces as it does in standalone mode, allowing the Zoxnoxious interfaces to be found.

Definitely an ear bleeder. It's the sound from an exorcism of buzzing robotic bees: ua-cam.com/video/RDfod9Jk-CQ/v-deo.html

I'm wrapping up deep dive video two, it should be out tomorrow so stay tuned for that. Anything in particular you're looking for on the Pi? I've got all the Pi config stuff in the github link in the video description.

That'd be great! If you get around to it let me know. The Z5524 dual voice card is really good. The other boards (3340 VCO, 3372 VCF/VCA) need a few design updates. And most critically the backplane board, that really needs a usable form factor. I keep thinking eurorack. Or at least some project box so it's not so janky.

@@UltraTroninatorI'm not a big fan of eurorack because of it's cost, is this currently set up to be powered by euro 12v dc? If the current backplane boards doesn't have functional issues then I'd probably just design a case to be 3d printed around it.

Power for the backplane is a eurorack compatible 16-pin, +12, -12, +5. Any decent eurorack power supply would work. If you were in the Seattle area I'd give you one of the spare boards I've got from JLCPCB.

Everything is on github: github.com/brer-rabbit/zoxnoxious/ Modulation sources can be from Rack and/or from the other boards. That was a big thing for me: designing signal routing such that one VCO could modulate another VCO (for example).

@@UltraTroninator Sweet, looking forward to checking it out.

Lots of modules in VCV would never work in hardware for ergonomic reasons (eg 8vert). Others rely on right click options. I started with VCV and now own hardware euro. Patching in either is pretty different, with VCV leading you to lots of "single purpose" modules, multiple copies of the same thing, and just more modules in general. The Surge collection in particular relies on a modulation scheme that is digitally native. I've used max, chuck, and super collider, and VCV is just the right level of abstraction for me. Downside is the graphics are computationally expensive. YMMV, to each their own, may many flowers bloom, my way is best and if you don't like it I'll send you chocolates.

Spot on correct about patching being different between physical and virtual. In the physical space you want stuff like Maths where you've got a ton of functions in a small space. Then go to VCV Rack where you get more single purpose stuff. And stuff like the patch in this video which is blatantly poor planning on physical layout and limited use of some modules/abuse of others. And it works just fine. Save, new, start afresh for the next patch. Another thing on patching differences between real & virtual. What's odd with the Zoxnoxious synth: it has access to all those VCV Rack control elements, but then it's only got six physical voice cards in the audio path. So the Zoxnoxious hardware voice cards tend to pack a lot of functions. For example, the VCF has a white noise generator on card, just so you don't need a separate voice card for that. Another example being the dual VCO voice uses two completely different synth chips that one probably normally think of combining. This gives a bit of a different patch paradigm than purely VCV Rack, or hardware eurorack. And flowers can bloom from each one.

hardware synth with a software frontend. or whatever you want to call it :)

Thanks for dropping a comment! Certainly if I thought there was a market for this, it's a very niche thing. That and a number of things need to happen first. Get it to a reasonable form factor, eurorack most likely. A lot more documentation needs to be written too. I'm slowly getting there. The gerbers / JLCPCB BOMs, and all the source code is available on github for someone adventurous.

Triangle will give similar result; the main thing is where the signal crosses the switching boundary. Pulse has an advantage in that you can vary the width so it can be more dynamic. That said: saw/triangle ought to be interesting since, with a greater modulation level, one should be able to switch between tri / tri+saw / saw instead of just two waveforms I demo'd here. Where the pulse jumps between any two (tri / tri+saw here), a non-pulse function could transition between all three tri / tri+saw / saw. I've not tried noise but I suspect it'd be noise in / noise out. Hear how you get a fair bit of the fundamental of the modulator (second half of video)? Noise would likely be noise. Now, if you put noise through a sample + hold and had the clock on that be something interesting that may lead somewhere.

Everything is a perpetual 6 months out :) All boards need some tweaks and revisions. Then I'd like to get this to a eurorack form factor. Or something similar where one would already have a power supply on-hand. If you want *just* a card I can get that for you much sooner. The cards interface with most microcontrollers over standard protocols, and you could code up whatever frontend interface you want

@@UltraTroninator oh I just mean it would be cool to see a West Coast Buchla style voice card in your format. It came to mind since you made the Krell patch!

That'd be something to get to! i picked up a bunch of Yamahs FM chips awhile back so that may be on the radar as well.

Thanks for dropping a comment! I love to play around with this stuff, and yeah it'd be great to hear what others could do with it. I've offered on the VCV Rack forum to port other's patches to a Zoxnoxious patch. "Remix" may be a more appropriate term - as it's certainly going to be tweaked. Nobody has taken me up on that but I'd be more than willing to try with a Rack patch. It's a tough sell. The synth is a strong monosynth, or even 2-3 monosynths, one can do a bit of polyphony if you limit it to a single VCO per voice. That and it's a completely dry audio path- no Rack effects involved since the audio path never goes through Rack itself; and I'm not adding any. If you (or anyone!) are interested let me know and I can give some pointers to how one would design a patch.

thanks! Once I finish up the auto-tune code and the dual VCO synth voice I'll get demos for each of those out, so stay tuned for those. And if I could find/port a VCV Rack patch that shows this thing off better I'll do a recording for that. My musical ability is near zero; I can do the circuit design & coding but it's a stretch to go beyond that.

Thanks! Plenty other blippity bloops on the channel here :)

Thanks! Hoping to get this accessible to others and in a workable form factor. Just more dev cycles ;)