Wireless DMX

[Richard Bunting]

We are wanting to investigate Wireless DMX, as a there are now 3 or 4

manufacturers who produce units using the WiFi standard. My question is this

- should there be agreement between all depts as to who uses which

channel (of the 13 available) so there is no interference. Lighting

could be using up to 8 (for various universes of DMX), but will sound

be using any in the future? What about remote control set trucks, or

general comms (though this would normally be on the PMR frequency range).We are looking at this from a rental point of view, where tours will

be in differing venues each night and be subject to different conditions each day.

 

What are peoples experience of Wireless DMX so far (if any), and their

thoughts for the future.

 

Regards,

Richard Bunting

 

VLPS Lighting Services

[Wilf dLampy]

It may also be necessary in a touring situation to fit alongside any WiFi data network run in the venue (or possibly the office next door).

 

A system with a re-adressable channel (as long as peeps don't confuse it with a DMX address ) might be useful. Then the LX guys can chat to the sound guys (it has been known to occasionally happen) on the first date of the tour or the prep rehearsal and find out what channels each is planning to use. At each venue, a quick sweep with a Wifi laptop, or a handheld tool (I think RS/CPC stock them) will tell the production wether any additional local networks are in place that they might wish to work around.

[peter]

Lighting could be using up to 8 (for various universes of DMX)

 

As I understood it, you could run multipul lines of DMX via one ethernet cable, and thus via one band of WiFi. I might be misunderstand this, but assuming you can run 4 lines on one band, are you implying a run of 32 lines of DMX?

 

Its also worth mentioning that there are certain interference issues with using all 13 channels, similar to using the deregulated radio mic frequencies - they all work on their own, but if you try and use all at once, you get unexpected issues. I'm sure someone more experienced will correct me on all of the above, but that is my understanding.

 

P

[Neil Frazer]

With the right kit you can run 64 lines of DMX over a TCP/IP network. Artistic licence, Howard Eaton Lighting, Pathway Connectivity etc, all make the kit to do this. One thing to consider is security. Make sure that any wireless link uses WEP or any geek with a WiFi card can hack your show! You should also consider the reliability issue. What happens if you loose the link for some reason? Fair enough if it not feasable to run a cable then use wireless but if it is just a case of saving time then would it not be safer to just run an ethernet cable?

[TomLyall]

just by 2p worth... but, unless you have a really big spread out rig, or some stuff on barges on a river or somethign else like that, would it not just be easier to run a cable... I cant see that this would be any more expensive and it would certainly be more relyable

[Ike]

in some situations, for example when your a line short and running another would mean digging up a field, when wireless DMX is a god send

[Tomo]

I would think the main use for wireless DMX would be the run between the desk and the stage/major lighting trussing.

Once you're at the base of each truss you'd drop back to cable.

 

One feature I would want in any wireless DMX system would be a 'link good' light for every receiver on the transmitter so I can tell from my end whether anything has lost the signal.

Also a signal strength meter, but I'd expect that on any professional RF system anyway.

 

WiFi is full-duplex so that shouldn't be too hard for manufacturers to implement.

 

The reason for this is that wired DMX will keep the signal if it's got it to start with (well, unless the cable is *really* knackered), but any wireless system has a significant likelihood of loss-of-signal, so I want to *know* if that's happened.

[David Buffham]

any wireless system has a significant likelihood of loss-of-signal

We've been playing with transmitting lighting control data over a wireless link, though I am concerned about the reliability of wireless comms in a performance situation, particulary after reading comments like this, so I was hoping to find out more about what kind of problems can arise with wireless links and how they can be guarded against. Some specific concerns I have are:

 

- what happens to an established network link if another wireless network starts up on the same channel? Do both networks go down, is there loss of bandwidth, what happens?

 

- what is the relationship between signal quality and bandwidth? i.e. do you only get the advertised bandwidth across the link with 100% signal quality? In reality would you ever have a link with a theoretical bandwidth capacity of 54Mb/s? (before taking into account factors as max utilisation, overhead etc..)

 

- what factors will affect signal quality? Is the wireless signal as susceptible to drop-out and interference as radio mics typically are?

 

- whats a realistic range for a wireless network, and to what extent does distance between transceivers affect bandwidth, signal quality, etc. etc..

 

 

What we're doing is using a pair of Wireless Bridges on the Ethernet link from our Virtuoso console. The Ethernet from the console carries Virtuoso protocol to 'nodes' on the LAN, each of which generates 6 universes of DMX. So in this case its NOT actually DMX that is being transmitted over the wireless link but an internal control system protocol. This is a two-way protocol, which appears to work on the basis of only transmitting state changes to the nodes. So the bandwith requirement is extremely low compared to DMX where each of the 512 channels would be refreshed 40 or so times a second per universe. In fact there is sufficient bandwidth on a 100Mb/s LAN to connect over 100 nodes, giving more than 600 universes - so in theory we could run 300 universes over a 54Mb/s wireless link; should we ever need to!

 

 

Tomo mentioned a couple of of things in the last post about things he'd like to see in a wireless system. Because the Virtuoso protocol is duplex, the console knows whether or not each node is healthy - and if the link does drop out for any reason then the nodes hold last state and the system diagram on the Mac screen turns the affected nodes from green to red - so the operator can see if the link has failed. Although we haven't done this yet I also believe that the Mac can connect to either bridge to interrogate the signal strength at each end of the link, so again this can be monitored.

 

 

Anyway, the system is working fine in the office and we're going to try it out on the next couple of shows - at least during plotting & rehearsals - where it will be particularly useful to avoid trailing DMX lines through the auditorium - but before we use it in a live performance I'd like to understand more about the risks involved, particularly in relation to signal quality.

[bruce]

First a disclaimer - my knowledge and experience of DMX is that of an interested amateur. But wireless networking is what I do for a living....

 

- what happens to an established network link if another wireless network starts up on the same channel? Do both networks go down, is there loss of bandwidth, what happens?

 

Best case, things will continue to work, sort of, but performance will go down the toilet... Worst case, it breaks.

 

- what is the relationship between signal quality and bandwidth? i.e. do you only get the advertised bandwidth across the link with 100% signal quality? In reality would you ever have a link with a theoretical bandwidth capacity of 54Mb/s?  (before taking into account factors as max utilisation, overhead etc..)

 

You'll never get anywhere near the full bandwidth. In the field, if you get more than about 60%, you're doing well.

 

The nominal bandwidth for IEEE 802.11b is 11Mb/s, but this will fall back to 5,2 and then 1 Mb/s as the signal strength decreases. IEEE 802.11a and 802.11g, which both support up to 54Mb/s, behave in a similar way.

- what factors will affect signal quality? Is the wireless signal as susceptible to drop-out and interference as radio mics typically are?

- whats a realistic range for a wireless network, and to what extent does distance between transceivers affect bandwidth, signal quality, etc. etc..

The advertising blurb typically says a range of up to 100m. This is nonsense. You might get this in clear space, but in a typical office environment my rule of thumb is "1 or 2 ordinary walls".

 

Propagation depends on building construction. Things like reinforced concrete, metal ceiling tiles, metal shelves etc are bad. Also, reinforced glass anywhere nearby causes wierd interference effects.

 

Having said that, by adding a directional antenna, you can easily get a couple of km range, and 20km is not impossible. A directional antenna can be something as simple as an old Pringles can....

 

Also, directional antennae will reduce interference from\sources outwith the "line-of-sight".

 

802.11b and g both operate in the 2.4GHz range - an unlicensed band. They may experience interference from any other device which uses this range - that includes anything from Bluetooth phones to microwave ovens. As someone said before there are 13 channels available in the UK, but neighbouring "cells" need to be separated by at least 3 or 4 channels, so effectively there are only 4 available channels.

 

802.11a runs at 5GHz - again unlicensed, but at present there are fewer interference sources here. Also more channels available.

 

Comparing it to radio mic "drop out" is not really appropriate. You will get this sort of effect, especially in areas of marginal coverage, but since the apps we are talking about involve fixed transmitters and receivers it shouldn't cause problems. If I were setting up a system like this, I'd want to make sure I had a very strong signal at the receiver, perhaps using a directional aerial.

Anyway, the system is working fine in the office and we're going to try it out on the next couple of shows - at least during plotting & rehearsals - where it will be particularly useful to avoid trailing DMX lines through the auditorium - but before we use it in a live performance I'd like to understand more about the risks involved, particularly in relation to signal quality.

<{POST_SNAPBACK}>

 

I'd be very interested to hear how you get on.

 

To analyse the airwaves properly, you really need specialist tools. But if you just want to do a quick test to see what wifi channels are in use, install netstumbler on a windows wireless PC, or ministumbler on a wireless ipaq - both free downloads from www.netstumbler.com. Or better, get a linux wireless system and "kismet".

 

Bruce.

[SimonW]

One thing to consider is security. Make sure that any wireless link uses WEP or any geek with a WiFi card can hack your show!

<{POST_SNAPBACK}>

 

If you're really concerned about security you should realise two things

 

*) Once you've gone wireless, its relatively easy for anyone in the area to flood your wireless network with 'bogus' packets, thus preventing your valid DMX data from getting through, and so crashing your show. They don't need to crack your WEP key to do this.

 

*) WEP is relatively trivial to break, and there are 'off the shelf' tools to allow those so inclinded to do so.

[bruce]

*) Once you've gone wireless, its relatively easy for anyone in the area to flood your wireless network with 'bogus' packets, thus preventing your valid DMX data from getting through, and so crashing your show.

 

Hence my suggestion to use directional antennae. If they're pointing directly at each other, it should be possible to reduce the receive gain at each end by a huge amount, so interference from off-axis sources should be minimal.

 

Remember, a couple of years ago wi-fi was virtually unknown. But today, it's pervasive. I've got at least 4 wifi devices in my house, and several thousand at work. I carry one, sometimes 2, around with me (a PDA and a phone). No doubt there will be someone in your audience with similar devices.

 

I've come across several instances where a rogue wireless device was "swamping" a network. It was not intentional - each time it was due to a misconfiguration, or a virus, or sometimes just someone being stupid.....

 

Bruce.

[martinw]

A very good example of wireless DMX currently in practical use is the Space Cannon installations on the temporary bridges over Oxford Street in London. These are in groups of 4, spread out over a distance of a couple of miles, and are controlled from a single console (a Hog 2 I think) using wireless DMX with directional antennae.

 

Martin

[Pete McCrea]

I can see the bonus of WiFi on things like that. But for the average show where you've got a 50m jump from FOH to the splitter/rack I'd rather go for wired. A lot less potential trouble shooting issues with a wire: No DMX at the splitter. Check desk out....good. Must be the wire, swap it out (or patch into a spare line on the Hum scum multi to check it :-) ). Now try it with a wireless connection..... A whole new range of pontential problems. And the possibility of only getting parts of the data to the required places.

 

I think I won't be an early adopter of WiFi for every day gigs.

[bruce]

I can see the bonus of WiFi on things like that. But for the average show where you've got a 50m jump from FOH to the splitter/rack I'd rather go for wired. A lot less potential trouble shooting issues with a wire: No DMX at the splitter. Check desk out....good. Must be the wire, swap it out (or patch into a spare line on the Hum scum multi to check it :-) ). Now try it with a wireless connection..... A whole new range of pontential problems. And the possibility of only getting parts of the data to the required places.

 

I think I won't be an early adopter of WiFi for every day gigs.

<{POST_SNAPBACK}>

 

I'd agree - you should only consider wireless if the benefits clearly outweigh the disadvantages. It introduces a lot of complexity, and an unfamiliar technology, and considerable expense, just to get rid of one piece of wire!

 

Having said that, I'm sure that in the past people said exactly the same things about radio mics!

 

To compare this with the data networking arena, I design and manage a network with around 12000 nodes, spread across 60 buildings on 2 sites. The majority of the interconnects are via fibre running in privately-owned ducts, with dual routes into most buildings. From a reliability point-of-view, it's rock-solid.

 

I also have a few outlying buildings that are connected via wireless links. Experience has shown that these are also extremely reliable. But I would be reluctant to use them for mission-critical stuff. If they should fail, it will take time and effort to restore them. The analogy we use is "I wouldn't want to run the payroll over them".

 

If I DID want to run a critical service over a wireless link, I probably wouldn't use wifi-type services. I'd look at either a microwave point-to-point link, which uses licensed frequencies, or an optical link. Both much more expensive, but less susceptible to interference. I would probably also provision a wi-fi-type link as backup. Belt & braces.

 

Bruce.

[gareth]

For what it's worth, I think that, by and large, I'd have to agree with the comments that have been made so far. I had my first experience of wireless DMX the other week, and I was reasonably impressed. It certainly did what it was intended to do (saved us from having to run in about 600m of DMX!), and got the DMX signal from the desk to where it was supposed to go. But it's still not 100% perfect - had to do one or two "big red resets" on one of the encoders, and I noticed a few data "dropouts" which lasted a second or two each. We didn't experience any problems with interference from other wireless networks, but that's not really surprising as I very much doubt there were any within range! So ... reasonably impressed, but on balance I think I'd still rather just run in a bit of cable if it's just a 'normal' FOH-to-stage run.