For about 357 days out of the year, I’d rather be in Austin, where I live, than in Dallas. Nothing against our neighbor to the north; after all, Dallas is the cradle of the automated lighting industry, the birthplace of the DLP chip, and the exporter of both Stevie Ray Vaughn and Lance Armstrong, both of whom took up residence in Austin. It’s not that I don’t like Dallas — I do. It’s just that I prefer Jerry Jeff Walker to Jerry Jones, Kinky Friedman to Ross Perot, and Willie Nelson to Vanilla Ice.
But for at least eight days out of the year, I want to be in Big D. That’s because ESTA holds two of its four annual meetings there, and much of the future of entertainment production is forged by the big hammers of the industry during those eight days.
The meeting last January was no exception. I made the four-hour drive in about two hours and arrived around dinner time. By the time I made my way to the sixth floor of the host hotel, it was late in the evening for the rest of the world, but it was high noon at the RDM Plugfest. The Plugfest was a sort of digital jam session where several manufacturers unpack their RDM gear and play it with other manufacturers RDM gear to see what kind of music emerges. It was a chance to experiment and see what worked and what didn’t, to try to break the software in the lab before it breaks in the field, and to ring it out in a real-ish world setting.
When I walked into the room I found Bob Goddard of Goddard Design, Doug Fleenor and Milton Davis of Doug Fleenor Design, Scott Blair of High End Systems, Peter Willis of Howard Eaton Lighting and a few other hacker jammers. Gack and geeks were strewn everywhere around the 20-foot conference table. The bits and bytes were flying.
Goddard had four of his new DMXter4 RDM testers connected together, one acting as a controller sending DMX data and RDM commands, two as intermediate devices intercepting the data, and one acting as a receiver getting the data and commands. Bob was taking RDM timing measurements, going through the RDM discovery process, and doing general RDM housekeeping. He changed the display on the first one from numeric to graphic while it was outputting data and played around with the DMX, adjusting the dimmer levels one slot at a time. With the intermediate units he captured the data as it streamed through the network. On the other end, he grabbed the data and displayed multiple slots at a time. He was dribbling data like Meadowlark Lemon dribbles the basketball.
At the other end of the table, Willis brought some Howard Eaton Lighting RDM Labpacks and he connected them to an Enttec RDM USB Pro interface. On his laptop he was running Enttec RDM Controller software and it was spitting out DMX and RDM data through the interface.
The RDM Labpacks are ideal for testing RDM data networks because each one has six sets of five LEDs, one each of red, green, blue, white, and amber. The RGBW LEDs represent a single pixel and the amber LED is the RDM status indicator. Peter was able to send DMX data to the device, light up individual LEDs and when he sent RDM data the amber LED lit up while it was receiving it.
He demonstrated a basic feature of RDM, the discovery process. That’s where the RDM controller polls the network by broadcasting to all connected devices a command asking them to respond if they are RDM-enabled. If there are any such devices then they will all respond at once, resulting in data collisions. So the controller then sends an RDM command to mute half of them according to the User ID, which is a sort of electronic serial number, unique in all of time and space that the manufacturer programs into it. Once half of the UIDs are muted, the controller multicasts a message to the unmuted half asking for a response. Again, if there are data collisions then the controller mutes half of the remaining UIDs and tries again. It repeats this algorithm until it isolates a single device, and then it records its UID. It then works its way back up this binary search tree, recording the UID of each device it encounters.
At the end of the discovery process, the controller will have recorded the UIDs of every RDM device on the network, and it does it quickly with no operator intervention. Since the UID is 48 bits long, there are 281,474,976,710,656 possible UIDs. Thank goodness computers are very fast; otherwise it would take hours for the discovery process to finish. In this case, there were only six “devices” so it only took a matter of a few seconds.
After he discovered the LEDs, Peter set the DMX addresses remotely from the laptop. In order to see which device he was dealing with, the controller would ask a device with a particular UID to flash (or IDENTITY, in RDM-speak), enabling him to set the DMX address according to the needs of the operator rather than just doing it randomly.
In a short while, Peter brought up the Robe Dreambox software on his laptop and connected the interface that comes with it. He demonstrated how it also discovered the RDM Labpacks and allowed the alteration of the DMX addresses that had already been set using the Enttec application.
About that time, Milton nearly jumped through the ceiling. He and Doug had one of their RDM Splitters connected to a power supply that was feeding an LED fixture. They were trying to figure out why the fixture was not behaving correctly. Milton had discovered that the power supply was passing along the RDM data to the fixture instead of filtering it out as it was supposed to. The problem was easily identified and resolved, thanks to their efforts. This is why the digital jam session, better known as the RDM Plugfest exists — to iron out the wrinkles in this emerging technology.
There’s a very popular bumper sticker philosophy that you’ll find on T-shirts and on the bumpers of cars all over Austin. It says, “Keep Austin Weird.” I think Dallas should adopt one of their own: “Keep Dallas Wired.”
Talk back to the author. His UID is rcadena@plsn.com.
