Author Topic: Pixelnet/DMX start address problem <SOLVED>  (Read 3294 times)

Offline wildwillie

  • Full Member
  • ***
  • Posts: 45
Pixelnet/DMX start address problem <SOLVED>
« on: October 03, 2012, »
First off, thanks to RJ and everyone else who has contributed their time and effort to designing all of this. You're much smarter than I am, and give a tip of the ol' hat to all for your hard work.

For the last 4 years, I've used a couple Firegod (128 ch) controllers. I didn't want to re-sequence everything, so I added the ETD in Vixen starting at Ch 257. A DMX universe for the Freestyle (512 ch), and that left the pixels starting at ch 769 (128+128+512). The ETD plug-in is set to start at 257 to 4352. Hub is set to DMX and Pixelnet 1

Here's the problem: No matter what the Vixen plug-in start channel is set to, it appears the ETD wants to start at Channel 1. I couldn't figure out why the pixels were acting strangely, and finally deduced that the SSC's were programed 256 channels too high. If the SSC was supposed to start at 1000, it had to be programmed to start at ch 744. I changed the start/stop addresses in the Vixen plug-in, but it made no difference. The ETD still started at channel 1. I can't tell if the FG data is getting hosed. It's still in the attic.

I know, you'll say "Just program the SSC's 256 lower!". That's not the right way to do it, and sets-up channel conflicts in Vixen, which is a recipe for disaster.

Can anyone verify this condition? Any fixes?
« Last Edit: October 04, 2012, by wildwillie »

Offline rrowan

  • Administrator
  • Sr. Member
  • *****
  • Posts: 5899
  • 08096
Re: Pixelnet/DMX start address problem
« Reply #1 on: October 03, 2012, »
Seems like you are trying to get the Etd to split between two pixelnet universes. You would be better off to start at pixelnet universe #2

Universe #1 = 1 to 4096
Universe #2 = 4097 to 8192

See if the wiki chart helps clear up something You are not allowed to view links. Register or Login

Rick R.
Light Animation Hobby - Having fun and Learning at the same time. (21st member of DLA)
You are not allowed to view links. Register or Login
Warning SOME assembly required

Offline tbone321

  • Sr. Member
  • ****
  • Posts: 4054
Re: Pixelnet/DMX start address problem
« Reply #2 on: October 03, 2012, »
Maybe I'm not understanding what you are saying.  What starting address are you setting your smart string controllers to?  You have to remember that the start channel in the controller does not always match the channel in the sequencing software.  Regardless of what you set the start channel of a dongle in Vixen, the start channel of the devices connected to the dongle will always be 1.  If you did take this into account and have addressed the SSC's correctly, then there is a bug in the EtherDongle pluggin.  If that is the case, then you may need to do as Rick suggested and address the added units into the next pixelnet universe. 
If at first you don't succeed,
your not cut out for sky diving

Offline tbone321

  • Sr. Member
  • ****
  • Posts: 4054
Re: Pixelnet/DMX start address problem
« Reply #3 on: October 03, 2012, »
You are not allowed to view links. Register or Login
Seems like you are trying to get the Etd to split between two pixelnet universes. You would be better off to start at pixelnet universe #2

Universe #1 = 1 to 4096
Universe #2 = 4097 to 8192

See if the wiki chart helps clear up something You are not allowed to view links. Register or Login

Rick R.

If this is the case, then there is a serious bug in either Vixen or the EtherDongle pluggin for it.  While it is true that a Pixelnet universe is 496 bytes long, there is no reason that it has to start at address 1 or 4097 in the sequencing software.  If that was an assumption during the creation of the pluggin, then that was a serious error that should be corrected in the future.  It is up to the pluggin to translate the Vixen addresses into the correct Pixelnet addresses for the dongle.  The issues could in fact be with Vixen itself and the fact that it may have been modeled around a 512 channel universe which could make it difficult or impossible to create a pluggin that could make the required translations between such different universe sizes.
If at first you don't succeed,
your not cut out for sky diving

Offline dpitts

  • Restrictive
  • Sr. Member
  • *
  • Posts: 466
Re: Pixelnet/DMX start address problem
« Reply #4 on: October 03, 2012, »
You are not allowed to view links. Register or Login
First off, thanks to RJ and everyone else who has contributed their time and effort to designing all of this. You're much smarter than I am, and give a tip of the ol' hat to all for your hard work.

For the last 4 years, I've used a couple Firegod (128 ch) controllers. I didn't want to re-sequence everything, so I added the ETD in Vixen starting at Ch 257. A DMX universe for the Freestyle (512 ch), and that left the pixels starting at ch 769 (128+128+512). The ETD plug-in is set to start at 257 to 4352. Hub is set to DMX and Pixelnet 1

Here's the problem: No matter what the Vixen plug-in start channel is set to, it appears the ETD wants to start at Channel 1. I couldn't figure out why the pixels were acting strangely, and finally deduced that the SSC's were programed 256 channels too high. If the SSC was supposed to start at 1000, it had to be programmed to start at ch 744. I changed the start/stop addresses in the Vixen plug-in, but it made no difference. The ETD still started at channel 1. I can't tell if the FG data is getting hosed. It's still in the attic.

I know, you'll say "Just program the SSC's 256 lower!". That's not the right way to do it, and sets-up channel conflicts in Vixen, which is a recipe for disaster.

Can anyone verify this condition? Any fixes?

I hope I am following your setup correctly. To summarize...

Two 128 channel controllers then a 512 DMX Freestyle controllers and last some pixels after that. If this is correct the first observation I see is that you have 768 channels of DMX so unless the DMX controllers will share some of the same effects you will need two active hubs to output all the DMX channels. The Etherdongle should be configured to start at 1 not 257.

I think to cover all your DMX channels you will string two active hubs together with one on DMX universe 1 and the other on DMX universe 2. The Etherdongle should be set to 1 to 4096 if you are only using one Pixelnet universe.

In Vixen start addresses:

First 128 Controller: 1
Second 128 Controller: 129
512 Controller: 513
Pixels: 1025       

if you are trying to save channels

512 Controller: 1
First 128 Controller: 513
Second 128 Controller: 641
Pixels: 769     


There are many other configurations but this would be mine.
« Last Edit: October 04, 2012, by dpitts »

Offline tbone321

  • Sr. Member
  • ****
  • Posts: 4054
Re: Pixelnet/DMX start address problem
« Reply #5 on: October 04, 2012, »
I don't see where he will need more than 1 DMX universe.  The Firegod controlers have nothing to do with the Lynks DMX universe so they don't count at all.  The Freestyle is a 128 channel controller and if he is dedicating an entire 512 channel DMX universe to it then so be it.  Either way, it is still just one DMX universe.  As for Pixelnet devices, it appears with this configuration and if Vixen is working correctly, his first SSC should be set with a start channel of 513.  That along with setting the beginnig channel of the EDT to 257 should have the first smart string node channel at 769 in Vixen.
If at first you don't succeed,
your not cut out for sky diving

Offline dpitts

  • Restrictive
  • Sr. Member
  • *
  • Posts: 466
Re: Pixelnet/DMX start address problem
« Reply #6 on: October 04, 2012, »
Like I said I may not have been following him correctly. I do not have Firegods. So they are not 128 channel DMX controllers?

Offline tbone321

  • Sr. Member
  • ****
  • Posts: 4054
Re: Pixelnet/DMX start address problem
« Reply #7 on: October 04, 2012, »
Even if they are, what difference does it make?  Unless they are also being driven by the EDT, they would be assigned their own group of channels.  The fact that he is starting the EDT at 257 shows that they are not.  There is no rule that I'm aware of that forces you to set a DMX dongle to a full Universe of 512.  512 is the maximum count that you can set a DMX dongle for and the minimum should be 1.  If the Firegod controllers are directly controlled by the PC, then the first one should be set with a starting channel of 1 and an ending channel of 128. The second one should have a starting channel of 129 with an ending channel of 256.  This is the only way thast Vixen would know which controller to send the commands to.   This should also allow him to set the starting channel of the EDT to 257 and set the ending channel to whatever he wants provided it doesn't exceed 4096 channels from the start so that Vixen knows to send all channels above 256 to the EDT.  As said before, if he wants to waste an entire 512 channels for a 128 channel controller he can but he doesn't have to. 
« Last Edit: October 04, 2012, by tbone321 »
If at first you don't succeed,
your not cut out for sky diving

Offline rrowan

  • Administrator
  • Sr. Member
  • *****
  • Posts: 5899
  • 08096
Re: Pixelnet/DMX start address problem
« Reply #8 on: October 04, 2012, »
You are not allowed to view links. Register or Login
--- snip ---- This should also allow him to set the starting channel of the EDT to 257 and set the ending channel to whatever he wants provided it doesn't exceed 4096 channels from the start so that Vixen knows to send all channels above 256 to the EDT.  As said before, if he wants to waste an entire 512 channels for a 128 channel controller he can but he doesn't have to.

I don't agree with that.

Sure he can start with 257 but a pixelnet universe will stop at 4096. If he needs more channels than it will go to universe #2. I believe that the Etd will send anything over 4096 to the second pair of wires and a second hub will be needed to get those channels.

Now the fly in the ointment would be if. There was a bug in the early Etd firmware that it would start with channel 1. RJ fixed that but if he is using DMX firmware, I am not sure if that version has the fix in it.

Rick R.
Light Animation Hobby - Having fun and Learning at the same time. (21st member of DLA)
You are not allowed to view links. Register or Login
Warning SOME assembly required

Offline jnealand

  • Patron Member
  • Sr. Member
  • ****
  • Posts: 2494
Re: Pixelnet/DMX start address problem
« Reply #9 on: October 04, 2012, »
The original poster needs to understand that the physical channels and the sequence channels are not necessarily the same.  For instance, I have a couple hundred DMX channels and take that off the active hub as DMX universe 1.  But in vixen I skip all the channels from my last DMX channel and start my pixelnet programming at 513 leaving room for potential DMX additions.  There are blank (unprogrammed) channels in vixen from the end of my active DMX up to 513 where I start the pixelnet programming.  But with Vixen I just rearrange my channels so all blank/unassigned/unused channels are at the end of the channel list and I do not see them when I am programming.  Pixelnet will always start at channel 1 and go thru channel 4096, but that does not mean they have to be programmed or assigned as such.  The original poster also needs to understand that if using DMX univ 1 then all channels from 1 to 512 are also sent to any pixelnet device so if he has sscs with addresses in the 1 - 512 range they will receive the exact same signal that the dmx devices will see.  And he needs to understand that this anomaly/feature can be used if desired for duplication of programming to two or more devices.
Jim Nealand
Kennesaw, GA

Offline wildwillie

  • Full Member
  • ***
  • Posts: 45
Re: Pixelnet/DMX start address problem
« Reply #10 on: October 04, 2012, »
<Written earlier this morning>   First, thanks for all the replies. Now that I've had my morning cup of coffee, let me try to explain what I'm seeing, and what _I_ think is going on....

<While composing a long reply>   While having that 2nd cup of coffee, it hit me. Right up side of the head.

The problem lies in user error.    <fp.

  From the dongles point of view, it IS address 1. It doesn't know/care what the vixen channel is, just that the first channel of data sent to it by Vixen is it's first output channel. The dongle and SSC's are going to start at Pixelnet ch1, not whatever it is in Vixen. I was programming the SSC's with the Vixen channel, not the pixelnet channel.

I can't believe I didn't see this before.  Sorry.  Hope you weren't all up late looking for a bug in the plug-in.....

I think I need to go solder something now.......

Bill

Offline tbone321

  • Sr. Member
  • ****
  • Posts: 4054
Re: Pixelnet/DMX start address problem <SOLVED>
« Reply #11 on: October 04, 2012, »
THa's what I thought it was ad why I said to remember that the EtherDongle always starts at one for the controllers, regardless of what Vixen calls it.
If at first you don't succeed,
your not cut out for sky diving

Offline tbone321

  • Sr. Member
  • ****
  • Posts: 4054
Re: Pixelnet/DMX start address problem
« Reply #12 on: October 04, 2012, »
You are not allowed to view links. Register or Login
You are not allowed to view links. Register or Login
--- snip ---- This should also allow him to set the starting channel of the EDT to 257 and set the ending channel to whatever he wants provided it doesn't exceed 4096 channels from the start so that Vixen knows to send all channels above 256 to the EDT.  As said before, if he wants to waste an entire 512 channels for a 128 channel controller he can but he doesn't have to.

I don't agree with that.


Then you would be wrong.

You are not allowed to view links. Register or Login
Sure he can start with 257 but a pixelnet universe will stop at 4096. If he needs more channels than it will go to universe #2. I believe that the ETD will send anything over 4096 to the second pair of wires and a second hub will be needed to get those channels.

You are confusing Pixelnet with Vixen.  Vixen doesn't know or care what protocols your devices are using at this level.  Vixen channel numbers are just that, Vixen numbers.  They are not DMX, Pixelnet, LOR, or even Renard channel numbers.  It is up to the Pluggins to pick out the correct data and send it to the device with whatever format the device needs to function.  When you set a starting and ending address for a pluggin, it acts like a variable length controller and grabs the channels you told it to.  It then renumbers that channel group starting with 1 and sends it to the controller such as the ETD.    If you set the pluggin to pull channels starting at 257, the first channel that the ETD sees is not 257, it's 1.  The ETD should never see a channel higher than 4096 on any of its universes if the pluggin is working correctly. 

You are not allowed to view links. Register or Login
Now the fly in the ointment would be if. There was a bug in the early ETD firmware that it would start with channel 1. RJ fixed that but if he is using DMX firmware, I am not sure if that version has the fix in it.

Rick R.

Talking to Dpitts last night I came to the conclusion that wildfire may have come to the same confusion and programmed his devices with Vixen channels rather than Pixelnet ones and he confirmed this with his last post and I'm glad that he is good to go now.  This also shows that there is nothing wrong with the pluggin and all is good.
If at first you don't succeed,
your not cut out for sky diving