I spent some time tonight doing work on the project. Roughed out (almost all) the I/O modules' analog signal flow, power, and a start at I2C. This now leads to the point of using ADCs and DACs.
My research skills are somewhat failing me. I am not sure at this point how to get a digital signal from an ADC to a processor point, particularly over ethernet. This is my current snag. The signal has to travel over a distance before it hits a non-preamp processor, either a computer or a dedicated card.
So how do we take the signal from multiple ADCs, get them *to* the ethernet line, through it, and to the processor? Ethernet should just be what it always is: packets. Large MTU packets probably but still packets.
So is the ADC just a stream of binary or what??
Help!
@opensourcestudio Using plain Unix sockets, you could just stream data over UDP somewhat network-agnostically. You’d want some kind of sequence value to ensure you don’t log data in the wrong order when packets arrive out of order, but TCPs Ack/retransmit overhead is probably not worth the added complexity. Unless you’re dealing with really high sample rates, you’re probably not going to run into any big issues. Also, allow configuring over Ethernet from the main control point.
Streaming ACN is a UDP-based protocol for streaming data for theatrical lighting, that has many bulletproof implementations on different platforms and languages. It is multicast-based, where a receiver subscribes to whatever ‘universes’ (groups of channels) they want to receive. So it’s really easy to add extra listeners or transmitters to a network without worrying about IP settings. It has sequencing built-in, and holds 512 bytes per packet. It might be easy enough to just it, and pack your samples into the 512 bytes.
