Spin Semiconductor

Hi,

I thought about getting some randomness in the future algorithms I'll try out with the FV-1.

First I have to ask a simple question. As I understand the FV-1 has two analog inputs (=adc).
It has one delay-memory for the delay based algorithms ... at program start it has to be choosen, (from) which input (=adc) values are written into the delay memory, right ?

So, most of the algorithms can't be in a "true" stereo mode (example: different delay for adc1 , adc2) cause
the adcs can't share delay memory. This is only a hypothesis originated from the knowledge about the FV-1 I I have NOW, correct me please if I'm wrong.
Coming back to the topic. My idea is, if most of the delay-based algorithms will be mono (= one adc is source of signal that will be transformed) one adc seems to be free for (I call it) sidechain algorithms. I could send a very noisy random signal to this adc and inside the "signal-algorithm" I could use my noisy signal (= read the signal , is it below or above value x ?) to make random decisions ...
Besides the problem how to create/get a noisy/random signal for one of the adc, I'd like to know if my idea is a practicable idea.
Best rewgards

No, you can do stereo. You can read from both ADCs and write to both DACs. You can use the memory for both, the limit is simply that the delay memory is 32768 long so the total cannot be more than that. So you can split it equally (16384 for each side) or unequally (24K for one side and 8K for the other for example) or how ever you want.

To generate pseudorandom noise try an LFSR, see this viewtopic.php?t=528

frank wrote: ↑Sat Aug 16, 2025 2:13 pm No, you can do stereo. You can read from both ADCs and write to both DACs. You can use the memory for both, the limit is simply that the delay memory is 32768 long so the total cannot be more than that. So you can split it equally (16384 for each side) or unequally (24K for one side and 8K for the other for example) or how ever you want.

To generate pseudorandom noise try an LFSR, see this viewtopic.php?t=528

For my understanding, the delay-memory is shared between the two "channels" ? Stereo - meaning two completely different sources - would mean I have to split the delay memory in two parts by software ?"

Basically yes. Use the memory how you want to. You could read the right input into a delay then read out from the delay and write to the left output. The "mem" statement defines memory blocks, just define a block then read/write it as you desire.