ProSoundWeb Community
Sound Reinforcement - Forums for Live Sound Professionals - Your Displayed Name Must Be Your Real Full Name To Post In The Live Sound Forums => Lighting Forum => Topic started by: Bob Faulkner on June 17, 2021, 09:31:36 AM
-
I'm not a lighting person, but I understand DMX addressing and setting up DMX fixtures.
I have some DMX wireless units (from about 7 years ago). I got these from Springtree. Never used them for a show, till recently, a band is asking for lights... For testing, I'm setting up some scenes with 8 x PAR LED (RGB) stage wash lights using a Chauvet Stage Designer 50. All PAR lights have their own addressing. For the configuration, wireless DMX was setup between the light trees (to avoid running a cable between the trees). There is a DMX cable between the controller and the first tree. Observed latency between the trees was extremely high. I'm guessing close to 100ms of latency - you couldn't miss it! I remove the wireless option between the trees and move it to the controller. Observed latency between the controller and first tree appeared well below 100ms (manageable). However, the latency between the trees (now with a DMX cable) was still there, but much less than the original 100ms. It's almost as if the controller was delaying the sending if the commands to the second tree.
I removed all wireless from the setup and went back to all DMX cabling; there is virtually no latency.
Is DMX wireless worth the expense/effort? Is the controller the problem?
It would be nice to have a wireless solution. At this point, DMX wireless does not appear to be an answer.
Should I abandon wireless DMX?
-
I use the Donner DMX wireless stuff and they work great. I'm sure there is a degree of latency, but for bands I'm calling it good. If I were to be doing lights for a theater company, it would be all hard wired.
Most drummers are off by 100ms! lol
-
I remove the wireless option between the trees and move it to the controller. Observed latency between the controller and first tree appeared well below 100ms (manageable). However, the latency between the trees (now with a DMX cable) was still there, but much less than the original 100ms. It's almost as if the controller was delaying the sending if the commands to the second tree.
That's not possible unless the delay is programed in, but you also proved that the wireless link isn't the problem.. you have some other issue. All hardwired fixtures should respond at the same time... assuming they are programmed to do so, if they don't I'd suspect the fixtures first... maybe a setting or a fault with one of them.
One thing to check is that all fixtures are set to slave mode, there cannot be any masters as they will hijack control of downstream fixtures.
-
I use the Donner DMX wireless stuff and they work great. I'm sure there is a degree of latency, but for bands I'm calling it good. If I were to be doing lights for a theater company, it would be all hard wired.
Most drummers are off by 100ms! lol
For sure! I thought about just going with it, but the delay was too distracting.
That's not possible unless the delay is programed in, but you also proved that the wireless link isn't the problem.. you have some other issue. All hardwired fixtures should respond at the same time... assuming they are programmed to do so, if they don't I'd suspect the fixtures first... maybe a setting or a fault with one of them.
One thing to check is that all fixtures are set to slave mode, there cannot be any masters as they will hijack control of downstream fixtures.
I agree, it should not be possible! It was baffling to watching. No delays were programmed. The trees function correctly (timing) when using all cabling. Good point about the master / slave configuration; though, from the manual, that setup shows it only works when using the fixtures without a controller.
I'm assuming DMX controllers send their data in a linear manner. Makes me wonder if the wireless units are unable to process the data fast enough. Sort of like using a 10-BaseT network components on a 1G network attached to 1G devices...?
-
Is there any chance that any of the fixtures have some type of delay/fade built-in to them? Some of our old Blizzard HotBoxes had a setting that made it seem like there was latency... It wasn't immediately obvious, but once we corrected the setting on the fixtures, things were more responsive.
And more power/good-on-ya to Mark for getting the Donner stuff to behave... mine always seemed to drop connectivity/fail about 3-4 minutes before showtime. I eventually moved over the the SHoWBaby stuff and it was a massive stress reducer.
-
DMX is a relatively simple and very forgiving protocol. Simplified way that DMX works is that it sends out serial data, sending out DMX addresses in order with their values in a loop. When a device sees it's DMX value, it listens and responds accordingly. If for whatever reason, the fixture doesn't hear it's address in the string, it just waits until the next loop comes around and responds then. So, in the event that there is data loss, the light still will eventually respond.
A couple of things that will cause delays. 1. The speed of re-broadcasting data. Since data is sent in a serial fashion at a predetermined rate, the more addresses used will mean a longer time between repeating the loop. Some devices will only broadcast addresses up to the highest address in use. Eg, if you use 10 addresses, it sends the value for those 10, then repeats. Other devices will send all 512 channels on each loop. This means that the response time would be slightly lower.
In the event that there is data loss, it can take time for the loop to get around again and re-broadcast the signal. Using wireless hops that are un-responsive typically means there is data loss or corruption in the signal. Not enough to completely break things, but enough that it's only hearing the signal some of the time, thus giving you the perceived delay you are experiencing.
-
DMX is a relatively simple and very forgiving protocol. Simplified way that DMX works is that it sends out serial data, sending out DMX addresses in order with their values in a loop. When a device sees it's DMX value, it listens and responds accordingly. If for whatever reason, the fixture doesn't hear it's address in the string, it just waits until the next loop comes around and responds then. So, in the event that there is data loss, the light still will eventually respond.
A couple of things that will cause delays. 1. The speed of re-broadcasting data. Since data is sent in a serial fashion at a predetermined rate, the more addresses used will mean a longer time between repeating the loop. Some devices will only broadcast addresses up to the highest address in use. Eg, if you use 10 addresses, it sends the value for those 10, then repeats. Other devices will send all 512 channels on each loop. This means that the response time would be slightly lower.
In the event that there is data loss, it can take time for the loop to get around again and re-broadcast the signal. Using wireless hops that are un-responsive typically means there is data loss or corruption in the signal. Not enough to completely break things, but enough that it's only hearing the signal some of the time, thus giving you the perceived delay you are experiencing.
Brian, nice explanation!
-
Something else I just thought of that I encountered with my own fixtures, If they have selectable diming curves and they are not all set to the same one that can make some of them look like they are slow to respond.
-
I removed all wireless from the setup and went back to all DMX cabling; there is virtually no latency.
Is DMX wireless worth the expense/effort? Is the controller the problem?
It would be nice to have a wireless solution. At this point, DMX wireless does not appear to be an answer.
Should I abandon wireless DMX?
I'll continue to maintain my opinion that wireless DMX should only be used when absolutely necessary - not as a matter of convenience. I'm all for minimizing time and effort when setting up a rig, but wired reliability isn't something I'll give up unless I have no other choice.
That being said, when I must use it, I implement a truck and branch philosophy (console to splitter and then branch out the wired runs with the wireless transmitter on its own splitter output). I never go from console to fixtures to transmitter to more fixtures - way too many opportunities for compromised signal to hit the transmitter and glitch everything downstream. If you decide to keep wireless I'd suggest getting a basic splitter and using a branch for wireless only. I can't guarantee that will fix your problem completely but it certainly won't hurt. Good luck!
-
Wireless DMX has gotten a LOT more reliable. Even the cheapie eBay units have been extremely reliable for me. The key is having realistic expectations. No, you aren't going to run 500 feet through walls and expect a reliable signal! But, for a simple LED tree, they are awesome.
Yes, I agree with you in putting the transmitter first in line. Segmenting DMX runs makes life a heck of a lot easier when it comes down to troubleshooting problems. DMX splitters are cheap enough these days to easily implement them into your designs.
The neat thing too about wireless DMX is that it's a splitter in itself. You can have one transmitter and as many receivers as you want. For my portable rigs, I have a receiver on each tree or truss segment and one main transmitter at the console. Way less wires and each segment is completely isolated. If there's a problem, very easy to track down.
-
Is there any chance that any of the fixtures have some type of delay/fade built-in to them? Some of our old Blizzard HotBoxes had a setting that made it seem like there was latency... It wasn't immediately obvious, but once we corrected the setting on the fixtures, things were more responsive.
And more power/good-on-ya to Mark for getting the Donner stuff to behave... mine always seemed to drop connectivity/fail about 3-4 minutes before showtime. I eventually moved over the the SHoWBaby stuff and it was a massive stress reducer.
Good thought. I don't see anything. For these fixtures, looks like everything with the fixture is "disabled" once it is connected to a DMX controller. A lot can be set on the fixture if they are not attached to a controller.
DMX is a relatively simple and very forgiving protocol. Simplified way that DMX works is that it sends out serial data, sending out DMX addresses in order with their values in a loop. When a device sees it's DMX value, it listens and responds accordingly. If for whatever reason, the fixture doesn't hear it's address in the string, it just waits until the next loop comes around and responds then. So, in the event that there is data loss, the light still will eventually respond.
A couple of things that will cause delays. 1. The speed of re-broadcasting data. Since data is sent in a serial fashion at a predetermined rate, the more addresses used will mean a longer time between repeating the loop. Some devices will only broadcast addresses up to the highest address in use. Eg, if you use 10 addresses, it sends the value for those 10, then repeats. Other devices will send all 512 channels on each loop. This means that the response time would be slightly lower.
In the event that there is data loss, it can take time for the loop to get around again and re-broadcast the signal. Using wireless hops that are un-responsive typically means there is data loss or corruption in the signal. Not enough to completely break things, but enough that it's only hearing the signal some of the time, thus giving you the perceived delay you are experiencing.
Good info. Thanks for the re-broadcast info. My 2-tree setup is only using 22 addresses. The fixtures work great without wireless. I'm thinking the wireless units are crap.
Something else I just thought of that I encountered with my own fixtures, If they have selectable diming curves and they are not all set to the same one that can make some of them look like they are slow to respond.
I think my fixtures have that control, but is only available if they are not on a DMX controller.
I'll continue to maintain my opinion that wireless DMX should only be used when absolutely necessary - not as a matter of convenience. I'm all for minimizing time and effort when setting up a rig, but wired reliability isn't something I'll give up unless I have no other choice.
That being said, when I must use it, I implement a truck and branch philosophy (console to splitter and then branch out the wired runs with the wireless transmitter on its own splitter output). I never go from console to fixtures to transmitter to more fixtures - way too many opportunities for compromised signal to hit the transmitter and glitch everything downstream. If you decide to keep wireless I'd suggest getting a basic splitter and using a branch for wireless only. I can't guarantee that will fix your problem completely but it certainly won't hurt. Good luck!
Thanks! Yep, my use of wireless is for convenience and speed of setup.
Wireless DMX has gotten a LOT more reliable. Even the cheapie eBay units have been extremely reliable for me. The key is having realistic expectations. No, you aren't going to run 500 feet through walls and expect a reliable signal! But, for a simple LED tree, they are awesome.
Yes, I agree with you in putting the transmitter first in line. Segmenting DMX runs makes life a heck of a lot easier when it comes down to troubleshooting problems. DMX splitters are cheap enough these days to easily implement them into your designs.
The neat thing too about wireless DMX is that it's a splitter in itself. You can have one transmitter and as many receivers as you want. For my portable rigs, I have a receiver on each tree or truss segment and one main transmitter at the console. Way less wires and each segment is completely isolated. If there's a problem, very easy to track down.
hmmm... did not think of using one transmitter with multiple receivers. I have 4 wireless units; I'll try setting up one transmitter with a receiver on each tree (like you have done).
-
did not think of using one transmitter with multiple receivers. I have 4 wireless units; I'll try setting up one transmitter with a receiver on each tree (like you have done).
Oh... I missed the fact you had a wireless X-mtr daisy chained in system in your original post.. yeah that won't work well at all.
Put a reciever on each tree, and the transmitter can plug directly into the console but I find it helps to elevate it overhead in crowded venues.. it's the line of sight is best for wireless communications thing.
-
Oh... I missed the fact you had a wireless X-mtr daisy chained in system in your original post.. yeah that won't work well at all.
Put a reciever on each tree, and the transmitter can plug directly into the console but I find it helps to elevate it overhead in crowded venues.. it's the line of sight is best for wireless communications thing.
Good points. Thank you Paul.