1. Dante requires Dante enable components. If one has an AV setup, then in order to use Dante on to setup an AV network one must change out any device that doesn’t have a Dante chip. If one is planning on replacing their mics, mixer, computer, cameras, etc, with a Dante capable one, then the cost are less of an impact. However, if one prefers not to replace these components then using Dante can be cost prohibitive (because one would have to buy at least one interface box for each component), or impossible. A switch that allows one to use any AV component seems more acceptable than being limited to an AV component because it doesn’t have Dante. I would be very interested in feedback regarding integration of Dante with legacy equipment in a cost effective way.

2. It seems that all discussions across the internet on Dante is audio focused. So it is unclear whether or not Dante is really up to the task of managing Video recording across a network. To do this would one need a Dante network switch? If so, then the generic switch argument drops off the table. I would enjoy getting more insight on this.

I know these comments are focused on Dante, but I imagine they are the same with any similar technology requiring embedded chips in the source.

I’ll also concede the cost and complexity statements as presented, but only to a point.

In my experience the cost differences aren’t as great as you’ve implied when factored into the cost of a complete system. Is it more for the hardware? Yes. Is it break-the-bank more? Not usually and I’ve generally been able to at least come in equivalent if not a little cheaper for Dante systems over the alternatives. First, I personally have found AVB enabled systems to be fairly comparable in cost to Dante enabled systems when compared to a non-network enabled version of the same product so not sure where your data is coming from. You do make a good point that AVB could or should be less being open source though and maybe you’re looking at different products than I have. Second and following off that, the reduced cost afforded by off-the-shelf network hardware usually more than makes up for any increase caused by the cost of the chipsets in my experience. Third, I find most cost increases to be just as much natural differences in quality levels as it is the addition of network connectivity regardless of format…lower end products aren’t as often network enabled to start with and higher end products cost more to start with. And of course forth, the cost savings in Cat5/6 cable vs standard audio cable factors in when looking at the complete system. As an effectively proprietary protocol even though it is open source (Music Group being the primary implementer as you’ve mentioned) and due to it not being natively network based in this context limiting it’s usefulness in the systems I typically deal with, I haven’t used AES50 as often and can’t speak to it’s relative costs.

As for complexity, it can be complicated and maybe it isn’t as intuitive “out of the box” as you may want it to be or it could be, but I have found a lot has to do with getting used to how it works with the receive and transmit channels (some people just have a problem wrapping their heads around it for some reason) and actually taking the training courses. I’ve never really had a problem with it myself but I know lots of people that have. In almost all cases the people that have the most trouble with it are also the people that have the most trouble with networking in general and/or haven’t taken the courses. You may argue that it should be intuitive enough to not need a training course, and to some degree you have a point, but the training courses Audinate put together actually address the network as well to address the issue of people having issues setting up the network to start with so they aren’t just Dante and are actually pretty useful on their own. (Having taken all 3 levels of the Dante certification courses, I’d recommend anyone working in Pro Audio in today’s environment take them.) On top of that I’d argue that any system coupled with a user base that is unfamiliar with the basic concepts of that system will instantly turn unintuitive. (I do a lot of trainings for schools for example where the part time drama teacher can barely turn the audio console on. For that audience, it doesn’t matter what Audinate did with the UI design, it would be very complicated for them.) That said, “out of the box” for a basic system, you’re generally talking straight crosspoint matrix routing with the Dante Controller software…pretty easy once you figure out the terminology. It isn’t till you get into systems which are inherently complicated to start with such as the university example you gave that the complexity really starts coming through and you have to “stop and think” about what you’re doing and get into the guts of the software, and any other network system I’ve used has had the same types of complexities in the configuration environment in those situations. It’s a new world out there where you almost need to be a network technician in addition to an AV technician. It’s going to take more and continuing training to keep up with technology advances and a lot less of the analog days of “I know the basic layout of this console and that console is just bigger/newer/etc.” In short, technically the complexity is a con, but not much of one in my opinion.

For my own rig, I was looking into Dante but with the recent announcement of the Pro Tools Carbon, I may go AVB with hopes that Avid will be expanding on the platform.

Hey Liam, thanks for your reply. let me try and address your points in turn:

COSTS
As mentioned in the article, AVB requires certified network switches which are more expensive than regular Ethernet switches as used by Dante.

The flipside is that Dante hardware is typically more expensive than AVB systems and especially more so than AES50 systems. This is an avoidable factor caused by Audinate’s monopoly on the chipsets implemented by manufacturers. That cost has to be passed on to the customer.

LATENCY
Theoretically, AVB should offer lower latency than Dante due to time-sensitive networking. However I am happy to concede that if in practice you’re seeing latency as low as 125u seconds that’s lower than AVBs typically 2ms.

AES50 still offers the lowest latency of all however at 63us. In reality would any of this be noticeable in a live venue? It’s debatable!

COMPLEXITY
I have yet to meet a fellow live engineer or systems engineer that hasn’t had a gripe about how difficult Dante is to work with. I’ve also spoken with academic institutions with large Dante networks complain about how unintuitive Dantes networking software is. I’m happy to stand my ground on it being a complex system to use, if not to implement physically.