Crow: Bipolar voltage ranges

[rycolos]

Per the specs, it sounds like Crow can output +10 but can only go as low as -5. Am I understanding correctly? In that case, what happens in this case:

output[1](lfo(60,10))

Would this be a bipolar LFO that goes up to +10v but only down to -5v?

[Justmat]

@rycolos

output[1](lfo(60,10))

Will get you an lfo. -5v to +10v with a bit of time spent hanging out at -5v . Just checked.

[rycolos]

Interesting! Thanks for checking. I really, really need to get an oscilloscope.

[dansimco]

In theory, could crow + a max patch be an oscilloscope? Not sure if self patching here would yield the most trustworthy results but a crow scope could be handy.

[rycolos]

That’s a great idea. Would be simple enough with sig~ and scope~.

[zebra]

er… quick caveat here: crow I/O is relatively low-bandwidth. think control rate. for looking at audio signals, a DC-coupled audio interface would be way more effective.

you can get a “proper” scope that can handle high-speed digital logic signals (and perform protocol/signal analysis, &c) for pretty cheap these days. “PC scopes” like picotech’s offerings optimize the cost/capture quality ratio by foregoing a dedicated screen and offloading analysis tasks to the host

[tehn]

also, would be trivial to make a norns scope with crow

low bandwidth but likely helpful for most euro-grade inspection?

[Oootini]

How low is low? A stereo input l/r lissajous/basic voltages/frequency measurement app would be neato for sure.

[Oootini]

Ok, so reading voltage is done with the stream input mode, but what about reading hz frequency? Would that be the change input mode and just counting the 0 crossings?

1.5khz sample rate should be fine for basic tuning purposes. Does that mean freq measurements of up to 1.5khz are possible?

[zebra]

well… it depends what kind of accuracy you expect

firstly, the serial stream needs to keep up and there will be jitter &c. so this is better done on the crow side, which is planned: https://github.com/monome/crow/issues/277

secondly, getting halfway decent frequency from zero crossings close to or exceeding nyquist frequency is not at all trivial. since there is a higher-samplerate loop going within crow, i believe a software PLL could be brought to bear.

up to 750hz things will be pretty smooth sailing and you can always fall back on the FFT with the usual tradeoffs.

above the nyquist freq of the input, it’s not so clear to me. i think you can do pretty well with some kind of smart PLL if you know the approximate range of the target freq (like within an octave) but i am kind of just wondering out loud here. if you dont mind long latency then various filters can help (hysteresis, kalman filter.)

anyways it’s not not a simple ask and it’s something trent will have to spend some enjoyable puzzle time on, and has certainly thought more about it than i have already.

for the moment, it seems an easy enough experiment to see how fast you can get change reports back to a host, and how well they correlate with the input, if this is something you’re very interested in… but i would let go of any expectations of being able to make phase plots or anything like that.

crow+norns as a voltage meter, for sure that’s an easy win.

[Oootini]

Thank you for the thoughtful and deep reply! Looks to be some very cool stuff in the v2 roadmap. I guess I’ll just wait for that freq input update, in the meantime I might experiment and poke around as you suggest above.