@m5mpy microcode is very low level, I like how these MCU bring back old memories and old hacks like I used to do on Amiga. And how in general they force into better programming habits.
So what I'm trying to do is general purpose signal camera. At the moment, a sound camera will do. From what I was told on eevblog, I need one FFT per microphone to calculate inter-mic delay. So far Arduino FFT switched to uint32 (faster on esp32 than uin8 I'm told) I'm getting 512 bins at 12 ms, which is't very zippy as I'll need 4 cameras at least. Not that big a deal because I don't need 60fps refresh but ... well, that would be nice :D
As for the frequency, at the moment I'm interested in detecting the origin of low frequency sound as they're notoriously difficult to locate by ear. Eventually I'd like to expand into an EMF camera.
You may have already read this, and i didn't have time to ready it properly but the info at the link below discusses the ESP32 sleep modes in some detail... Nothing jumped out as a clear solution in a quick skim read.