ACN and RDM

[Bryson]

Has anyone read the article on ACN (Advanced Control Network) and RDM (Remote Device Management) in L&Si this month? (L&Si, February 2004, Technical Focus, P59-61, James Eade) For those who are unfamiliar, ACN is the proposed "replacement" for DMX and RDM is an upgrade to DMX that gives some of ACN features. Actually, it's a touch more complex than that, but you'll have to read the article. It's a good round-up of what is going on with the new standards.

 

Anyway, my point is, that it all sounds very interesting and useful, but I can't help but think that the article sidesteps one of the major issues with ACN. Apparently, the standard only proscribes the way in which the data is transferred in terms of software. The physical hardware connections and cabling are left to the manufacturers to decide upon for themselves. Surely a fundamental part of the reason that DMX is so useful is not just the compatibility of the protocol, but also the compatibility of the hardware. (With some notable exceptions) The 5-pin XLR is such an integral part of DMX's success as a control system, I can't see how allowing the manufacturers to choose their own hardware is going to work.

 

It also makes reference to the use of the existing standards for ethernet as the cabling standard. Surely for our uses, we've conclusively established that cat 5 cable is no good for flexible use. Do we not need a standard that takes into account our enhanced needs for flexibility and ruggedness? The same goes for connections. I can say from close personal experience that standard ethernet connections are not even slightly suitable for the kinds of uses that we put DMX connections through. So what to do? The Neutrik ethercon looks good - but I'm guessing that it's a patented system, and so not totally suitable for introduction as a compulsory choice as part of a standard. (I could be wrong here, feel free to correct me. Who owns the patent on an XLR5, btw?)

http://www.neutrik.com/images/ock/sublevel/thumbs/1_204_170351.png

 

It seems to me that if this is to be a useful standard, the manufacturers will have to agree on a common connector and preferably a flexible, robust cable to be used. And if they agree to that, why not make it a part of the standard?

 

Opinions, corrections and further information please, folks.

[Brian]

You are correct. ACN sits above the IP layer which, of course, sits above the hardware layer. In terms of defining an extensible protocol this is absolutely the right way to do it.

 

By doing it this way you have a choice of how the data is physically moved about. More importantly it allows you to use standard networking hardware. At present you can move IP packets over IEE802.3 - wired Ethernet at several data rates, IEE 802.11 - wirless Ethernet, IEE1394 - Firewire as well as RS-422/485 serial links, and high speed telecoms infrastructure like ATM, SONET, SDH, FDDI. In fact there is even a spec for moving IP datagrams by carrier pigeon (I kid you not).

 

The upside of this is that ACN could evolve to use new hardware as it is invented. The downside is that it can make interoperability a problem. In reality I think that most manufacturers will introduce ACN over wired Ethernet, probably at either 10 or 100 Mbps.

 

Your comments about Ethercon are relevant but you should remember that at its heart is nothing more than an RJ45 8/8. You can use ordinary network leads to plug into an Ethercon socket.

 

The cable issue is a good one. At the moment there are no serious industrial Ethernet cables which will stand up to life on the road. Maybe something like optical fibre is the answer. I have worked on systems for military use where we used 'Tactical' cables which are very rugged indeed.

[Knut]

It also makes reference to the use of the existing standards for ethernet as the cabling standard. Surely for our uses, we've conclusively established that cat 5 cable is no good for flexible use. Do we not need a standard that takes into account our enhanced needs for flexibility and ruggedness? The same goes for connections. I can say from close personal experience that standard ethernet connections are not even slightly suitable for the kinds of uses that we put DMX connections through. So what to do?

 

Hi Bryson,

I am sorry I don´t read the article you mentioned. But I have some sugesstions.

Indeed Cat5/6 won´t be a good stage standard just the terrible nonelastic cables

but in a big theater/disco/congresscenter the new ACN shuld be very useful esp. when lighting the hole venue. The structures of ethernet is cheap to install and could be used by any type of IT/LT for exp.

You shuld control the normal stage rig with many of movinglights and convetionals.

than you have another rig (FOH) with many movinglights too. and you have to control

venue lights. normally you must have five hands to control 3 desks. If now all signals could be converted to one ethernet type you only have to split back to DMX at:

• stagerigs
  
• stagefloor
  
• FOHrig
  
• Dimmer City
  
• Venuelighting dimmers
  

The other thing is to have a fu**ing big lighting console to have about 2045 DMX channels availible to control all of above listed. Sometimes my fantasy works hard :huh:

 

The only thing to do is press "go" on your dedicated palm connected to the "desk" via

wireless.

 

Sorry spelling, and dreaming but this could be done using DMX and ACN.

[Wilf dLampy]

One of the cable manufacturers at PLASA last year was selling Cat5 data cable that felt like a standard mic lead. I've got a bit here in fact (a 25 cm sample that I made into a patch lead). The writing on it says:

eurocable extraflex cat-5 s-ftp data cable.

It's only slightly stiffer than a normal mic cable, and is black. I've been using this for a quick network test lead, and its been in and out of pockets, drawers etc since September and it's still working OK.

 

It crimps into standard RJ45s (like the one's inside the ethercon).

 

I like the proposals in principal. Use of standard ethernet cabling will also greatly aid the installation market, all new buildings are generally flood-wired with cat-5e/6 cable, meaning you can have the equivalent of DMX outlets all over your venue, patchable back in a central location.

[Ellis]

I have read (and understand) the draft specification that was availble from the ESTA web site late last year. My copy of L&S I came this morning, I will look at the article tonight.

 

As far as LX is concerned ACN is more akin to ArtNet/SandNet etc, but it goes much further. The Multipurpose Control Protocol is exactly that. It can be used to carry many DMX universes over a single wire. It can also be used for MIDI, SMPTE Timecode, or even video (given a sufficiently fast network).

 

Maybe a future High End mover might use ACN for all attributes including the image to be projected.

 

As for the physical networking, ACN is too abstract to specify that. As Brian said, it could run over a serial line or even RS485 (just as DMX does). It does not even specify TCP/IP other than as an implementation of ACN.

 

I suspect that when the abstract standard is ratified, further implemntation standards will follow, probably going as fas as specifying a suitable connector. You never know, Neutric might even licence the concept of putting an RJ45 inside an XLR housing freeley.

 

Where ACN comes into its own is in the field of discovery. If knut's fantasy rig have ACN movers, dimmers, etc. with no DMX anywhere, he could plug an ACN lighting desk in and it would find every dimmer and fixture and know exactly what each attribute channel was. The dimmers might even be able to send back a current reading or a message that a lamp had blown, and the lamp life of the movers could be viewed on screen.

 

As a software engineer by day, I can see many uses for ACN outside the industry as well.

 

As for XLR 5 connectors, I am pretty sure that they have been around for more than 20 years and as such are out of patent. Hence the almost complete absence of Cannon from the connector market.

[Brian]

The structures of ethernet is cheap to install

is a claim often made. However, it is no cheaper than DMX. DMX will work quite happily over CAT5 cable if you want. True, some Ethernet hubs are cheap but they are cheap for a reason ie wall-wart power supplies, nasty connectors, built to a price.

 

At installation time the actual cost of cable/connectors is tiny compared to labour costs. This is why IT installs now flood-wire buildings. I've seen figures of £10 per Ethernet outlet when installed at the start against £100 when retro-fitted.

 

The approach suggested by Knut is probably they way people will go initially, run Ethernet around the building and break out to DMX at fixed locations but...

 

you can have the equivalent of DMX outlets all over your venue

which will be of no practical use unless you flood wire the building with 63A mains outlets. Unless you have ready access to everything you need at every location there is nothing to be saved by having your 'DMX' signal in the foyer for that product launch if you still have to run in a mains feed.

 

eurocable extraflex cat-5 s-ftp data cable

this cable still has one problem, it has a foil screen (as well as a braided one) which is not really suitable for a flexible cable. After a while the foil kinks under the jacket.

 

The potential benefits of ACN are more to do with the ability to move data in both directions at high speeds in large amounts than anything else.

 

RDM is a different beast completely and I think is likely to have a greater impact in day-to-day use.

[Knut]

I have read this question before but I don´t remember the answer:

What are the possibilities to run DMX 512 on ethernet over swiches and hubs ?

 

I don´t think DMX will be transfered over networkdevices.

 

Greets

Knut

[Brian]

If you mean use Ethernet hardware to process DMX then the answer is no. However, have a look at ArtNet from Artistic Licence. Artistic Licence

[Ellis]

You could also look at Sand Net(Uses IPX not TCP/IP).

[martinw]

The current work-in-progress ACN spec is just the transport layer upwards. As others have said, as this runs on TCP, it can in theory run over many different physical layers.

There will almost certainly be a 'Recommended practice for ACN over Ethernet Networks' specification or guide produced once the protocol has been developed further, much in the same way that the DMX spec is about to get a separate (but totally unnecessary IMHO) spec for cabling. ACN is still some way from complete, and once it is, it will still have to compete with other standards such as Artnet. Whilst Strand and ETC are actively working on it, this does not guarantee its acceptance throughout the industry.

 

Martin

[TeeJay]

True, but then ACN is meant to be for a different market. It was originally conceived way back when as a universal for show control, as opposed to Art-, Strand-, ETC-Net which are primarily for lighting control. That said, just browsing through the list of adopters of ArtNet means that it can be used as the protocol in a show control system.

 

Of course, there is still some die-hard support for MSC and work still progressing (last I checked) on MWPP, MIDI over Ethernet if you will.

 

I have a feeling this is a debate that is going to resurface regularly for a good while yet . . .

[Ellis]

As far as I am aware,ACN will become an ANSI standard. As such, Art-, Strand-, ETC-Net will probably fade away slowly.

 

Most lx desks are software upgradeable so adding ACN will not be difficult.

 

Many of the intelligent ethernet based devices have upgradeable firmware, some of them are being marketed as ACN upgradeable.

 

It will get big in the US and we will just have to follow suit (good thing too).

[martinw]

As far as I am aware,ACN will become an ANSI standard. As such, Art-, Strand-, ETC-Net will probably fade away slowly.

Possibly, but there are more than a few ANSI standards which have never caught on. The standards process takes so long compared with the pace of technology, that they can become obsolete before they are finished, often becoming superseded by a manufacturer or group of manufacturers opening up a proprietary standard, much like has happened with Artnet.

 

Many of the intelligent ethernet based devices have upgradeable firmware, some of them are being marketed as ACN upgradeable.

Yes, there is a lot of marketing BS regarding 'ACN ready' at present. It will be interesting to see where this actually gets implemented.

 

It will get big in the US and we will just have to follow suit (good thing too).

Time will tell I guess, the spec is only a first draft, and it still has to prove to the industry that it can do the job required.

 

Martin

[Madmac]

For a manufacturer to spend money on either ACN or RDM development there must be a clear return. The development cost for ACN is going to be much greater than that for RDM so manufacturers will implement the latter to test the market. Both systems will increase the cost of fixtures in both development and parts cost for manufacture. Will the limited advantages be worth the extra cost to the user?

 

The ACN software stack is going to need resources that most fixtures in use today do not have spare so it very unlikely that any will be upgraded to support this standard (when it is a standard). It is also unlikely that Ethernet to the fixture is going to appear for a long time, so some form of TCP/IP over RS422 /RS485 would be needed to keep cost down, possibly compatible with RDM DMX hardware.

RDM needs a lot less to be implemented, so assuming the hardware is capable a number of existing fixtures may be upgraded.

 

Having fixtures implementing these standards is only half the problem, console manufacturers have to put in the development to integrate fully these standards to get the most from them and this will take a while. Remember that the majority of consoles on the market have output only buffers on the DMX lines and that rules out any bidirectional communication unless hardware upgrades are made available.