kaareolai Posted September 24, 2018 Share Posted September 24, 2018 At the moment I saw this specification on a DMX product (Showtec Node-8 MKII Artnet-DMX Node or Botex X-NET 8 PRO), I googled on every key word, but could not find anything specific. Someone that might enlighten me on this High Speed DMX protocol? And how much it is implemented. Or is it just a way to tell the node has Gigabyte ethernet capabilities!? Which is far from necessary for just 8 DMX ports. Regards Kåre Olai Lindbach Link to comment Share on other sites More sharing options...
timsabre Posted September 24, 2018 Share Posted September 24, 2018 I had never heard of it... but if you google for "DMX-1000K" you will find some other products with this new type of DMX which runs at 1000kbit/sec instead of 250kbit/sec, it does not appear to be related to the gigabit ethernet. Appears that it also carries 2048 channels per universe, it seems to be on some LED products. I presume one manufacturer must have made a system with it and others have followed, there does not appear to be any standard. Link to comment Share on other sites More sharing options...
timmeh2 Posted September 24, 2018 Share Posted September 24, 2018 Hi I've come across it before, in 2003 or way back in the past, on a certain brand of LED wall. Can't remember which, it had 1024 channels per universe and had a faster baud rate like above. It didn't catch on then, and I doubt this one will catch on now. Like Tim said, it's not an approved standard and I can't see the major desk manufacturers taking it on. Whilst we're not done with DMX just yet, we are moving into distributed network land where you run multiple universes over ethernet and split them out at the ends, e.g. pathport. This is certainly true on tours and shows where you're running a backup desk in tandem with the main one. All the bestTimmeh Link to comment Share on other sites More sharing options...
timsabre Posted September 24, 2018 Share Posted September 24, 2018 I think it is aimed at fixtures with a large channel count. If you have a fixture with lots of pixels which takes more than 512 channels, you have a bit of a problem, you either need multiple DMX-512 inputs or direct ethernet in which significantly increases the hardware cost/complexity. A basic serial DMX system providing more channels makes sense for this requirement. Link to comment Share on other sites More sharing options...
bigclive Posted September 24, 2018 Share Posted September 24, 2018 Hmm, it looks like it may be trying to jam in 4 times the channels with a similar refresh rate. I can understand why with modern fixtures consuming huge quantities of channels, but it really hints at a need for a better network with proper error detection for products like that. Link to comment Share on other sites More sharing options...
timmeh2 Posted September 24, 2018 Share Posted September 24, 2018 Hmm, it looks like it may be trying to jam in 4 times the channels with a similar refresh rate. I can understand why with modern fixtures consuming huge quantities of channels, but it really hints at a need for a better network with proper error detection for products like that. The latest Chauvet stuff has dual input - you can fire DMX at it for pan, tilt etc, and control the pixels via artnet; or both via artnet using separated channel blocks. Works really well. Link to comment Share on other sites More sharing options...
timsabre Posted September 24, 2018 Share Posted September 24, 2018 I have seen some pixel tape drivers with 4 XLRs for 4 universes of DMX-512 in. Not a great solution. You would have thought using ethernet would be a better solution, but it requires quite a ramp up in hardware and firmware design. And neither Artnet nor S-ACN provides error detection. Link to comment Share on other sites More sharing options...
niclights Posted September 24, 2018 Share Posted September 24, 2018 The latest Chauvet stuff has dual input - you can fire DMX at it for pan, tilt etc, and control the pixels via artnet; or both via artnet using separated channel blocks. Works really well.Quite a few do this now. Clay Paky were the probably the first I noticed when they introduced it in a firmware update to the B-EYE's. But as I understand the main thinking behind this is for use with pixel mapping on media servers where it is beneficial to separate out the pixels from the other controls. Even when combined as one mode the footprints still fit within the standard 512 limit and I have not seen any fixtures yet that don't. I think this would be a strange thing to do as it would be practically unsupported. As an aside I am starting to see some 24bit controls. So far this is restricted to media servers as far as I am aware. Crazy resolution! Link to comment Share on other sites More sharing options...
Tomo Posted September 24, 2018 Share Posted September 24, 2018 I have seen some pixel tape drivers with 4 XLRs for 4 universes of DMX-512 in. Not a great solution. You would have thought using ethernet would be a better solution, but it requires quite a ramp up in hardware and firmware design. And neither Artnet nor S-ACN provides error detection.Actually, both sACN and Art-net provide error detection. UDP transport ensures that a "wrong" packet will never arrive, and both application protocols include a sequence counter so a receiver knows that it didn't. Whether a receiver chooses to do anything with that information is another matter of course. What they don't provide is confirmation of receipt. The controller doesn't know if any packets are arriving. Link to comment Share on other sites More sharing options...
timsabre Posted September 24, 2018 Share Posted September 24, 2018 UDP packets can have a checksum, but I thought it was optional. I've certainly had trouble with "wrong" packets getting through. Maybe bad network stack code? Link to comment Share on other sites More sharing options...
matt c Posted September 25, 2018 Share Posted September 25, 2018 You'd hope that if this was to be ratified as a standard, they would use a standard EIA-485 setup. Proper bidirectional communication, and a CRC on the packets. In my custom-built model railway controller, I used a 1Mbit/s EIA-422 link, over cat5 cable, with DMA transfer of 1000 byte packets between two microcontrollers. Works like a dream. No reason why DMX can't go the same way, if it's viable to stick with EIA-485 rather than ethernet. Might have to throttle it back to get an acceptable distance though. UDP packets can have a checksum, but I thought it was optional. I've certainly had trouble with "wrong" packets getting through. Maybe bad network stack code? Correct... UDP can indeed have a checksum, but it's not required. The problem with UDP compared to TCP is that delivery of UDP packets is not guaranteed. They can be unceremoniously dropped along the way or arrive in a different order to what they were sent in. The protocol running on top can mitigate with this by adding sequence numbers and CRCs etc, to ensure a little bit of order in the chaos. Link to comment Share on other sites More sharing options...
kaareolai Posted September 25, 2018 Author Share Posted September 25, 2018 Thanks, so far, for interesting info and associations around the subject! Regards Kåre Olai Lindbach Link to comment Share on other sites More sharing options...
WayneHowell Posted September 27, 2018 Share Posted September 27, 2018 I'm on the DMX512 standards group and there is no such thing as DMX-1000k. So this must be a manufacturer specific thing. Link to comment Share on other sites More sharing options...
timsabre Posted September 27, 2018 Share Posted September 27, 2018 Thanks for the confirmation, Wayne.I've done a bit more digging into this and though it is not an official standard, it is supported by a variety of LED equipment across a number of different manufacturers and also LED driver chip suppliers (e.g. MySemi). So it appears to have become a standard even though not officially one, perhaps like the use of 3-pin XLR for DMX. It is even supported by people like Strand/Philips on their DMXFix DMX analyser although that appears to be a rebadged product. The protocol is basically identical to DMX except that it runs 4 times as fast, and supports up to 2048 slots/channels per frame. It appears that fixtures can autodetect standard DMX or DMX-1000K, presumably by timing the breaktime pulse. Link to comment Share on other sites More sharing options...
WayneHowell Posted September 27, 2018 Share Posted September 27, 2018 That's interesting. Its probably worth noting that a lot of DMX splitters would not work with that. Many RS485 driver chips are only rated for 500Kbps. Link to comment Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.