So the MCU on the NerdSeq appears to be https://www.microchip.com/wwwproducts/en...2048EFH144 .
Notably, this chip has several fast (18msps, for a Nyquist cutoff at 9,000KHz) ADCs built in.
Looking briefly at the PCB, it seems at least the first analog input is connected (didn't look for the others) to something - it disappears down a via, but there are ground planes adjacent to it so it shouldn't just be grounded.
Are the ADCs connected to the CV inputs? Would make sense.
If so, can the input filtering / level shifting circuitry operate at audio rate?
If it can, then does anything else prevent a NerdSeq from live-recording 12 bit samples? (Presumably, RAM would have to be pre-allocated before recording, but other than that...)
Being able to do sound design with my modular, record into a sample, and then use that sample seems like a pretty amazing workflow- right now that seems to require either getting a PC in the loop (yuck!), using an external, pricy module to record and then swap the SD card from the recorder into the NerdSeq, or an even pricier module like the BitBox + the More Triggers expander.
(On a related hacking note - looking at the datasheet for that MCU, there's an external bus interface that supports pretty hefty amounts of SRAM. Might have to see if I can't break out the ol' bodge wires and make a More Sample Memory expander...)
Notably, this chip has several fast (18msps, for a Nyquist cutoff at 9,000KHz) ADCs built in.
Looking briefly at the PCB, it seems at least the first analog input is connected (didn't look for the others) to something - it disappears down a via, but there are ground planes adjacent to it so it shouldn't just be grounded.
Are the ADCs connected to the CV inputs? Would make sense.
If so, can the input filtering / level shifting circuitry operate at audio rate?
If it can, then does anything else prevent a NerdSeq from live-recording 12 bit samples? (Presumably, RAM would have to be pre-allocated before recording, but other than that...)
Being able to do sound design with my modular, record into a sample, and then use that sample seems like a pretty amazing workflow- right now that seems to require either getting a PC in the loop (yuck!), using an external, pricy module to record and then swap the SD card from the recorder into the NerdSeq, or an even pricier module like the BitBox + the More Triggers expander.
(On a related hacking note - looking at the datasheet for that MCU, there's an external bus interface that supports pretty hefty amounts of SRAM. Might have to see if I can't break out the ol' bodge wires and make a More Sample Memory expander...)