At first, I didn’t really noticed a difference between cables when using a linear power supply/SD system based around Cadets. Later, I switched to a lighter case powered by switching supplies, and the cheap no-name cables I was using were picking a lot of noise from the PSU, so I went for Polar Noise Shielded cables, which helped greatly with the noise.
Then I started building a system around HD modules, and that’s where cables shielding/length started making a difference. Both parameters effect the capacitance of the cable, which basically creates a filter between the two modules connected, hence why it helps with high frequency noise coming from a switching PSU.
Here are some tests I already posted on the Discord a while ago, with a black and white checkerboard in 720p60 as a source, with direct connection from VU003B to VU007B.
Same cable length, different shielding:
Same shielding, different cable length, Polar Noise Shielded iirc:
I’ve also tested with alligator clips (basically cables with no shielding), it does improve the bandwidth but picks up a lot of noise (from Capsule PSU iirc). Second capture is with RG179 cable and Rean jacks, RG179 is specified at 64pF/m, jack sockets also adds a bit of capacitance, so didn’t really noticed a difference with RGB minisnakes. Last one is Tendrils, which are up to par with Polar Noise Shielded from what I’ve tried:
Haven’t tried other cable types, though seems like RGB Minisnake performs the best, then keeping connection between modules short does also help, but it gets tricky with dense patches and/or passive splitting. Also, all those captures are 720p60 which is the format that requires the most bandwidth, so going for a lower resolution/framerate could also help.
Seems like it will always be a tradeoff between bandwidth vs noise immunity, though I only tested with VU003B/VU007B with a Capsule power supply, I guess Gen3 modules might be less prone to noise since they all have their own integrated power supply.