BEYOND THE DMX UNIVERSE

In Uncategorizedby tfwm

Miles Dudgeon of Wybron Inc. breaks down the language you use to control your lighting equipment. As the lighting industry is still following the progression of DMX as a control protocol, other standards have recently been approved that should be closely watched. Dudgeon explains why there is about to be a major shift in the way lighting systems will be controlled in the future.

Let me start by saying that this article is not the definitive source on lighting control. What we hope you end up with is a really good general understanding of how your lighting equipment listens to what your lighting controller is asking it to do.

Most lighting equipment is quite frankly “dumb”, it won’t do anything until told, and it dutifully waits for its next command. Some manufacturers build in playback or demo modes, but for our purposes here, we will be concerned with gear that listens to the controller.

A simple lighting system generally consists of a controller (the light board), dimmers, luminaires (lighting fixtures), and other more advanced gadgets. The controller is the brains of the system; it’s the broadcast center that sends out all the signals. The dimmers listen to this signal and respond by adjusting the electrical output to the luminaires. The system may also have other equipment like moving lights, scrolling color changers or fog machines. The controller sends out a signal called DMX, and the other equipment listens.

So what is DMX? The short answer is that DMX stands for Digital Multiplex. You may have also seen it called “USITT DMX512/1990”, or even “DMX512-A”. Officially it is known as “Entertainment Technology — USITT DMX512–A — Asynchronous Serial Digital Data Transmission Standard for Controlling Lighting Equipment and Accessories”. That’s a little wordy, so we’ll stick to DMX.

DMX started off in the mid 1980s as a project with the United States Institute of Theatre Technology (USITT) to allow lighting equipment from one manufacturer to respond to a controller from a different manufacturer. Up until that point you generally needed to purchase the controller and dimmers from the same company. USITT revised the standard in 1990 and in 1998 the Entertainment Services & Technology Association (ESTA) began a public review process to allow DMX to become an American National Standards Institute (ANSI) standard. ANSI is a group that oversees the creation of industry standards from all aspects of business. ANSI is the official United States representative to the International Organization for Standardization. All these committees and organizations allow us to be reasonably certain that lighting equipment made in one country will work with equipment made on the other side of the planet. This is good for manufacturers and critical for consumers.

DMX uses a cable that looks deceptively like a microphone cable. It has XLR connectors, but instead of three pins, it has five. It is crucial that you do not use microphone cable for DMX. Audio signal has different shielding and capacitance requirements. The first pin on a DMX cable is a ground, pins 2 & 3 actually carry the data and pins 4 & 5 are left free to be used by some manufacturers for talkback or other proprietary communication. The output of DMX (from the controller) is always a female connector and a DMX input is always a male connector.

SPLITTING, DAISY-CHAINING AND TERMINATING
DMX is designed to run up to 32 devices on one line. You might be thinking “But I have 96 dimmers and a bunch of other stuff!” The dimmer rack may contain 96 dimmers, but it only has one DMX input. Essentially it is one device with 96 parameters. DMX can be daisy-chained from one device to the next. Most equipment has a DMX input and a DMX out or pass through connector.

So what about those really big shows with hundreds of moving lights? We can split up the DMX signal using special splitter boxes sometimes called “opto’s”. We need a special splitter so that the signal is regenerated, allowing for 32 more receivers on the line. Many different companies make splitters. They can be as small as 1 in / 2 out, up to very complex routing and distribution systems. The important thing to remember is that you can’t just splice two cables together, like a ‘twofer’. The other benefit we get from a splitter is that frequently they are optically isolated. Optical isolation is an electrical barrier between the input to the output and between the other outputs. If something bad happens on one leg, it won’t take down the signal on the other legs of the splitter, or go back up the line and damage your really expensive controller. DMX is sensitive to grounding issues, so using an opto isolator will help solve those problems. Many people find that just putting an opto in line with their one DMX cable solves all kinds of signal problems. Another common source of problems with the DMX signal is termination. If the signal doesn’t have any place to go, it can reflect back down the line and degrade the DMX. Using an opto splitter with an open output isn’t a problem since it won’t affect the other outputs. However, on chains of DMX, at the end of the run a terminator should be used to absorb the signal. Many people sell these terminators or you can make your own by soldering a 120 ohm resistor across pins 2 & 3. There are also some DMX devices that will auto-terminate if there is nothing plugged into the DMX out or pass-through ports.

THE LANGUAGE AND THE UNIVERSE
The DMX signal itself has up to 512 channels being broadcast. It’s a lot like cable television in that the channels are all being sent, but your receiver is only listening to the one you tell it to. We call this group of channels a universe. Many modern controllers can broadcast multiple universes of DMX. Simple lighting equipment might only need one channel to operate, but a dimmer rack may need 48 channels, as it has 48 dimmers.

Devices with blocks of parameters generally use a start address. At the device, you define what channel it should start listening to. So a 48 dimmer rack can start at address 49 and the last dimmer will listen to DMX channel 96.

It’s important to remember that DMX travels one way, and doesn’t really care what is plugged into the signal. For instance, it doesn’t know that you may have two dimmer racks both listening to start address 49. The racks generally don’t talk to each other, and have no idea that the dimmer rack that is plugged in one link over, is at the same address. DMX equipment just passes the same signal it used on down the line. This is great if you need to have multiple devices doing the same thing. Moving lights also use a start address with multiple parameters. A moving light will start listening to DMX commands at a specific channel, but depending on its complexity may use 32 or more channels. These extra channels tell it how much to open the dimmer, how far to spin the pan or tilt motors, what gobo to use, etc. DMX doesn’t actually tell the moving light “go to gobo 5”. Instead, the channel has a value of 0 to 255. For example, the light will know that a value of 157 on the fifth channel from the start address means to put gobo 5 in the beam. Dimmers know that a channel value of 0 is completely off and a value of 255 is full power.

A full universe of DMX consists of up to 512 channels with each channel containing a value of 0 – 255, being refreshed by the controller between 20 and 44 times a second. Like television, changing the picture 30 times or more a second, we can effectively make a smooth transition from one value to another. In the DMX specification, controller manufacturers are allowed to send only the amount of DMX channels they need for the devices attached. So if you are only using 48 desk channels, you only need to send the first 48 DMX channels. However you can’t send channel 1 then channel 10, you must send all channels up to the maximum channel you are using. This has caused problems in some equipment; so modern controllers generally send the full universe, even if they aren’t using all the channels.When you are looking at systems that have multiple universes, it’s important to keep in mind that your controller may think of a device having an address of 513, but actually the device is listening to channel 1 on the second universe. Controllers do some other things designed to make your system more user friendly, but may confuse matters a bit. Many controllers, especially those designed primarily for dimmer control, present the channel value as a percentage. It’s just plain easier for a human to know what 50% is compared to 128.

Controllers also use a construct called “soft patch”. Soft patch is used so you can group dimmers together so similar purpose lights appear next to each other on the screen. Say you have 12 dimmers running your front light; you may have those luminaires plugged into several different dimmers (ie: 1, 4, 12, 35, etc). Using soft patch, you change how they are displayed on screen so your front light is 1-12. When the DMX is actually sent out, the controller sends the right value to a different channel. Soft patch is really not a function of DMX, though it’s important to keep in mind that the channel number on the screen might not actually be the DMX start address of your device.

DMX is a ‘one transmitter, multiple receivers’ system, but what if you wanted to have multiple transmitters? There is special equipment that allows you to merge multiple DMX streams. These merger devices can work on either Latest Takes Precedence (LTP) or Highest Takes Precedence (HTP) settings. Some devices will even allow you to selectively replace specific DMX values. It can be tricky designing a system that allows for multiple controllers to run the same equipment. Many large systems will completely split out the moving lights from the conventional luminaires and dimmers. It becomes two discreet systems in the same venue. Some venues might have wall-mounted button stations to control the architectural lighting with a link to the performance lighting system. These might be either DMX based or have a DMX link to an architectural control protocol. The possibilities are vast and take up more room than I have space for in this article. Just keep in mind; your light board may not have the final say in which lights are turned on.

OTHER WAYS OF TRANSMITTING DMX
Besides our trusty 5 pin XLR cable, there are other ways of transmitting DMX. Several manufacturers have developed systems that translate DMX into a more friendly protocol for TCP/IP networks. If you have a newer installed system, chances are good that you have some of these devices in it. All the major control and dimming manufacturers have developed these systems; Net 2 & Net 3 from Electronic Theatre Controls and Shownet from Strand Lighting, for instance. Other systems like this include Pathport from Pathway Connectivity, Artistic License’s ArtNet, Sand Systems’ Sandnet and DMX-Link from Goddard Designs.

Basically what all these systems are doing is taking a serial signal and chopping it up to be used on the same network type that your computer uses to access the Internet. Most commonly these systems use Ethernet cables and networking gear. The really neat thing about these systems is that you can have one relatively small cable transmitting many universes of DMX instead of one larger cable per universe. Theoretically one system can transmit 255 universes (130,560 addresses), practically less than 64 universes is the norm. The problem is that your lights still expect to have a 5 pin XLR serial DMX signal, so you will need special hardware to translate the networked signal back to standard DMX. Some consoles output directly to a network, but you may also need input hardware for the network too. Most manufacturers call these ‘nodes’.

Now that we are in the realm of having much more bandwidth and configurable nodes all over our venue, things can get really tricky. These systems allow you to remotely change which universe is coming out of the ports on the nodes. You can even combine universes and have selected channels from one universe and the remainder of the channels from different universes, but on the output it looks like one stream to the fixtures. The complexity of these systems can get overwhelming.You can also send DMX wirelessly. Some people have had luck using the DMX over Ethernet systems and using a wireless router. This isn’t actually a recommended practice, at least with the wireless router you might pick up at your local mega-office-superstore. Most consumer grade WiFi gear isn’t designed for the constant type of traffic lighting control uses. However, plenty of manufacturers have developed wireless systems designed for exactly what we need. Essentially the hardware is similar to the DMX over Ethernet systems; you have an input/transmitter node and receiver/output node, with no wires in between. Most manufacturers use a spread-spectrum radio system; there is even one that uses infrared, but only for short distances.

The wireless DMX systems are purpose-built for our needs, most include some error checking and correction, they use radios that are designed for use in areas with lots of other radio devices, like wireless microphones. They are great for places like churches where you probably have power (at least a little), but for some odd reason they forgot to run a DMX cable through the wall when it was built.

THE NEW STANDARDS
In the past few years there have been significant efforts to expand what we can do with lighting control. In the past there have been manufacturer specific talkback schemes, but now communication can be two-way. Last year, ESTA ratified a new standard called Remote Device Management, or RDM because we like acronyms. At its most basic, RDM will allow you to set the DMX address of a fixture from the controller, actually see if the device is there (called discovery) and identify the device. RDM uses the same cable as DMX, actually on the same pins as DMX. So you don’t need to replace everything in your rig to use it. However it is important to remember that not all DMX splitters will allow RDM traffic to go backwards through them. Besides the basic RDM commands, manufacturers are allowed to add specific device information and commands.

For example, Wybron has added RDM features to their most popular color changers. To use the Coloram as an example, it can tell the controller what colors are loaded on the string, if the fan has been clogged up or blocked, if the gel string has broken, if the device isn’t receiving enough power, even if the light behind the color changer is on, etc. When you start up a new rig with RDM equipment, the first step is to perform a discovery. Every RDM device has a unique identifier or UID. The first part of the UID identifies who made the device, and then it is basically a model number and serial number. Each piece of equipment has its very own unique name that no other piece of gear in the world has. The discovery process asks everything in the rig to report back its UID’s. After receiving UID’s, the controller then goes through and systematically asks each device what it can do, how many DMX channels it needs, what its start DMX address is and so on.

With this information you can do things like change the personality of the device, change its address, and ask it to identify itself. The identify feature is important because you may not have set all the DMX addresses before you hung the device, and now you need to find out where in the air it is. In the case of Wybron’s color changers, they run back and forth for a few moments during identify. If you are lucky enough to own one of Doug Fleenor’s DMX/RDM coffee makers, it will buzz after receiving the identify command. (Yes, he really made an RDM coffee maker). Different devices will identify in different ways, its up to the manufacturer, just as long as they do something out of the ordinary enough for you to find the unit.

When working on a rig with RDM equipment, you need not do your addressing on the ground. As long as the device matches what’s drawn on the plot, you can hang it. Then after the rig is up in the air, and other departments are working on the deck, you can go through and identify each piece of gear and set its address. When something is addressed wrong, it doesn’t mean pulling out the ladder, or lowering the position, you just type it into the laptop sitting next to the console. The same is true if you need the device to operate in a different personality, you can change the gel string move speed to “slow” so it is quieter. You don’t even need to set the addresses on the dimmer racks, it’s all done from the controller.

ARCHITECTURE FOR CONTROL NETWORKS (ACN)
In October of 2006, another new standard was ratified: Architecture for Control Networks, commonly known as ACN. It’s not, as many people believe, Advanced Control Networks. ACN is actually a whole host of standards under one common umbrella. It defines how entertainment equipment can communicate using the UDP networking protocol. UDP stands for User Datagram Protocol. UDP is faster than TCP/IP and suited for time sensitive communication.

This time it’s not just the lighting folks; audio and other types of equipment manufacturers are interested in what can be controlled with ACN. It’s important to remember that while you may be used to seeing TCP/IP and UDP run over Ethernet, ACN may use other styles of networking. Because Ethernet equipment is so common and easy to get hold of, it will be a de-facto standard. You will start to see more equipment with an RJ45 jack instead of a DMX connector. There are new consoles that only have a network output; you can get a node to go to DMX.

True ACN devices won’t have DMX start addresses. They have a Media Access Control address, MAC. If you can network your computer, it has a MAC address. Your cell phone has a MAC address. Basically anything that can be attached to a network has one. These act as the unique identifier to the network, and tell the controller what’s out there. The theoretical limit on the number of devices attached to a network is seemingly infinite. However controllers will have some fixed amount of equipment they can operate. With ACN, the problems associated with multiple controllers get much easier to tackle, and the rig can be easily segmented while still communicating with every piece of gear. ACN control commands don’t look like DMX at all. In DMX you say, “Channel 2 is at value 127.” Then it’s up to whatever is supposed to be listening to channel 2 to decide what 127 means it should do. In ACN, the command is more like, “Set your pan motor to 50% of its potential”. Or more specifically, “Go to pan value 250°”.

Because ACN is bi-directional you’ll know what gobos are loaded into a moving light. The light will transmit back to the console a picture of what gobos are loaded. Clicking on a picture is a whole lot faster than looking it up in the manual and typing in a value. The concept of cross-fading can be thrown out too. You can tell a moving light to go to a certain gobo by clicking a picture and it takes 10 seconds. You can then start rotating the gobo by 20° per second. You can send a single command that encapsulates all sorts of information that you would normally be sending as a stream of constant changes. It’s possible that ACN commands could even include image, video, or audio files. RDM and ACN are brand new, there is gear out there that will use it, but as an industry, everyone is still trying to figure it all out. As users, you have a lighting system that works today. What you will probably start to see is ACN nodes that will output DMX, a lot like the DMX over Ethernet nodes, but a lot smarter.

You’ll see dimmer racks that only speak ACN. You may even see a moving light without a DMX connector at all. But for the time being, most lighting control will go back to good old DMX and then into the device. All of these systems will make day-to-day operation of your rig more complex, but pay back in amazing versatility and abilities to foresee and fix problems from the ground, in the middle of service. The most important part of all of this is to illuminate your message.