I have been talking to the nice folks at Audinate about allowing Dante to be a bit more flexible.
At present, Dante is channel dense and therefore only works on extremely low latency networks due to the fact that it needs to have RAM in the module to buffer each channel. The higher the channel density, the more RAM you need to buffer.
What I’m interested in is also allowing Dante to be used (with a lower channel density) in higher latency networks. Perhaps over a long range WiFi link or even over a high quality internet link or VPN tunnel.
This would allow Dante to be used for…
Multi-channel outside broadcasts
Remote Monitoring of audio (the remote mix)
Wide area audio distribution
Studio links / backbones
Multi-site events
In much the same way as the Virtual Sound Card allows you to choose 2 x 2 channels up to 64 x 64, depending on the latency of your network you could potentially have a lower channel count operating over a high latency link.
Now of course, Audinate is not going to do any development work to enable a feature like this unless they know that it is going to sell more modules and DVS licenses so I thought I would ask here whether this would be useful to anyone with an iLive, and if so whether it would cause you to buy more Dante cards if you don’t have them or perhaps more DVS licenses? or perhaps more iLives and therefore more Dante cards for them.
If the feature is useful to enough people, we might just be able to get them to build it.
Personally, it would cut a lot of cost out of doing large scale events and it would mean I could standardise on iLives and Dante for both concert style events and public address / broadcast events.
In fairness, bringing WiFi into the Dante equation brings in a number of questionably controllable WiFi issues with it. Many issues of 802.11 around the concepts of delivering reliable multimedia services are currently in the process of developing standards.
It’s not a trivial problem to solve in the unlicensed WiFi bands.
Dante and AVB have both required working groups to develop standards on WIRED networks. There are a number of additional issues on WiFi to address.
Intrigueing idea! The wireless part of it would scare me as well thought I could see playing with a circularly polarized system across a room or longer link outdoors to see how well the rejection factor helps with WiFi. Toby, i know I’ve seen a lot written from the FPV folks on how well CP works and I’m sure you’ve well versed in that as well.
Dante has actually almost nothing to do with audio. It’s a data neutral protocol, there are plenty of people using Dante for digital video distribution as well. It’s more useful to think of the Dante protocol as Bit Torrent or FTP, you can shlep anything you want over it.
That being said, there are tons of wireless systems you can use for Dante. Ubiquity Networks makes an “AirFiber” 1.24 gbps link and there’s another forum member who used it a few months ago to transmit audio across a runway during an airshow and it worked beautifully.
I agree with more ram you’d be able to do longer buffering, but this becomes the ‘airplay paradox’ which means there’d be a substantial delay, a delay high enough that standard network dropouts due to wifi would have no impact on audio. That’s a lot of delay, if you don’t believe me set your laptop next to your microwave while it’s on and try streaming something
It was me to was planning to transmit for the airshow.
The problem is that it works fine with the laptop running DVS so I ordered a whole pile of Dante cards for all my iLives on the back of that testing.
The problem is that the Brooklyn modules have less RAM than the laptop and also the DVS has a high latency setting which the Brooklyn modules don’t have.
This means, what works with DVS, does not necessarily work with the Brooklyn modules.
I’m using Ubiquiti carrier grade point to point links (these things look like satellite dish’s and are good for 20km+). This is not consumer grade WiFi we are talking about here. These things have an internal link capacity of 350Mbps and are rock solid. I can transmit 7 channels of high def video over these things without missing a beat.
It seems frustrating that I can do that, but I can’t pipe my audio over them.
The Brooklyn modules refuse to work over these links, but DVS works fine. The latency is just slightly above what the Brooklyn modules can deal with.
Seeing as I only need a handful of channels, what would be great is to be able to reduce the channel count on the Brooklyn modules and increase the latency buffer using the extra RAM freed up by the reduced channel count so that they work in a slightly higher latency environment.
At the moment the max latency for a brooklyn module seems to be about 5ms. My link sits about 5ms to 6ms when I tune it to the max.
It is fairly frustrating, its right within grasp but just beyond reach and what is the most annoying is all the money spent on 3 DANTE cards based on the testing of the DVS software. I (falsely) assumed that DVS and the Brooklyn modules would be the same.
