Replacing an inch diamter monster cable with a cat5 cable - one of the most significant benefits of fully digital desks.
OK - A&H called their connection dSnake, and it is ethernet Layer 2 compatible (which is good in many ways, but could result in confusion). What does that mean in terms of latency, in terms of how you can connect things etc…
The easiest way to connect a stagebox and a mixer is with: direct ethernet cable
The stagebox and desk both use ethercon shells to protect the ethernet cable, and use of those is recommended - but not essential. The RJ45 port in the centre of an ethercon is a perfectly standard RJ45 port, so any patch cable can be used.
NB - I’ve not used a crossover cable - has anyone else ended up using one accidentally?
If you don’t have a long enough ethernet cable, or you need a bit more than the 100m limit, or (and I think this is important) you want to able to ‘plug in’ a desk at a number of different points in a building without having to repatch anything then you can: Use a simple ethernet switch
Any ‘fast ethernet’ switch (that is 100Mbps) should be fine if it is dedicated to the dSnake network - I would advise against using a hub - and would check to ensure that the switch is full duplex (they all should be).
A GigE switch will function just as well.
Of course this opens up other options. Longer distances
I have had to provide a remote link to a stagebox in a different building - we installed a couple of fibres between the buildings and a switch at either end to convert between the Cat5 and the Fibre. This is fine - it works exactly as it should…
BUT: Warnings
Almost any switch of this type will try to be clever - ours was trying to sniff traffic to prioritise it. it took us a little while in testing (this is key) to identify the “features” of the switch which we needed to disable.
We could not trunk the fibres - We experienced frequent dropouts and bad data. I put that down to the jitter introduced by the change in routes (despite both fibres being identical in length)
Good news
VLANs are fine - we actually ran multiple video feeds over the same fibre as the dSnake, with the switches at either end providing port based VLAN between them. This also works over copper, and you can therefore run the dSnake AND the control network over the same cable.
question
Has anyone tried WiFi or Powerline connectivity?
Latency etc.
Andreas posted:
Ok, for sake of curiosity I did a quick sniff and the “protocol” seems to be as simple as a protocol could be: It’s just a bunch of 24 Bit audio samples within a Layer2 frame (LLC) sent as broadcast packets.
The raw payload (about 217 Bytes) could hold up to 72 channels (we already have 40 monitoring channels from Qu to Stagebox along with the regular returns), so there simply is no room for a second sample but probably some additional protocol information (like box type, expander connected etc.).
Consequently frames are sent in about ~20µSec intervals, which nicely fit to our 48kHz samplingrate.
It seems that samples are being sent basically as they are generated - there isn’t an appreciable difference in latency between local and dSnake connected devices.
Things to avoid
Plugging anything else into the dSnake network. (Note the exception around VLANs above)
Nice one, Bob!
From my experiments I noticed that a hub does not work at all, probably because of collisions caused by half-duplex communication.
If someone insist to try it, don’t be worried if you can’t get a dSnake connection after removing the hub. Just reboot both stagebox and desk and you’re up again. I guess the failed transmissions (due to ethernet collision) are interpreted as a bad connection and shut down the transmitter. There’s still a blinking yellow light, though.
So: Do not use a Hub for dSnake!
At this point wouldn’t it seem like a good idea to just write a driver to read the raw Layer2 audio stream of the dSnake.
Instead of USB, hell go Ethernet…why not?
Quick fix, maybe just change the order…USB for MIDI, dSnake for Audio??
In the sense that I believe you can extend the range, or convert over fibre, of an ME-1 by running it through a switch you can’t do active/active splits using network switches.
They allow you to do things like plug it in one of a few places, but you can’t plug into multiple places at the same time.
Note that the ME-1 is a possible exception - it might be possible to present a star topology of ME-1s with a network switch, because I suspect it is UDP broadcast traffic…
You won’t be able to mix and match.
My first point was that the Ethernet Layer2 compatibility was potentially confusing - and here is that confusion.
I’m very interested in this “sniff” of the dSnake protocol…
Do you happen to have any of the “initialization bits” that are exchanged when a StageBox or Expander first connect to the mixer? I noticed that my QU32 didn’t “register” a dSnake simply by plugging in the Ethernet to the dSnake port, so I’m assuming there has to be some kind of initailization / handshake.
I’d like to write a GStreamer pipeline to parse these raw layer-2 frames into an audio stream on the device. I’m already using GStreamer to parse the raw USB audio (Thank you A&H for fixing the noise issues with the 1.95 firmware), but I’d really like to give this a shot too! For anyone interested in how to capture USB raw audio into GStreamer (and Open Broadcast Studio), check this out: Examples · fzwoch/obs-gstreamer Wiki · GitHub
@Crosan007 I haven’t seen anything that looks like initialization packets getting exchanged during the first connection – however, if you’re just looking at getting the audio stream for recording on a computer then you might not need to send anything. If you just connect the dSnake port to an Ethernet port on your computer you’ll receive packets with audio information. I recorded a sample, but there are some weird things (maybe DAC related) with the audio signal. It sounds recognizable, but the samples sent create a sort of “stepped” wave instead of a smooth waveform.