xlights and LOR s3 RGB

[frankr]

Hello all,

I am wondering if anyone has tested using RGB channels in S3 and having xlights do the show?  I am very interested in using this approach this year as my experience trying to get LSPs scheduler working for Halloween was scary to say the least...

I am hoping someone has tested this and already found the challenges so that I may learn from their experience before I go converting all my sequences to S3...

Thank You,

Frank

[dowdybrown]

Yes, I ran my Halloween show using S3 sequences. I had 8 RGB fixtures (24 channels) connected to a pair of MR16 DC controllers, plus 9 regular channels on an LE and it ran fine in xLights.

Matt

[JonB256]

Matt, that is great news about S3 and RGB. Did you output using an E1.31 device? or a Lynx Dongle?

[frankr]

Thanks Matt that is indeed great news!

Frank

[taybrynn]

I am curious, in LOR S3 ...

(1) do you need to set up channels as DMX channels?

(2) do you need to map out a PN universe as a series of 512 channel DMX universes in LOR S3?

I can get it to work through a Pixelnet flashed dongle and SS HUB on a 120 node flex string in hybrid mode, using the
xlights RGB tester (EXCELLENT, BTW!) ... but when I play a tested LOR S3 sequence, I can't seem to get anything to
light up ... hmm.

Thx for all you have done with this excellent program.

Scott

[taybrynn]

I also noticed that if I add the LOR S3 channels as regular channels (can add up to 512) ... that I get network designators like "regular, aux A, aux B, etc..."

and if I add the channels as DMX, then I get network designators such as "universe 1, universe 2, etc.".    IN each case I am somewhat limited to 512 channels
per universe, so I was thinking that pixelnet channels would just get set up as a series of 512 channel universes in LOR S3 and then mapped out to a single 4096
range of channels on the xlights 2011b side.

Anyways, I wonder which (type of channel, network designator) xlights is expecting and how to configure them from the LOR S3 sequence side.  I am encouraged that this will work, but I must be setting it up wrong somehow.

I would appreciate an example of how to set up a simple 4096 pixelnet universe from a single Lynx Dongle flashed with pixelnet.

Like I said, I can get the tester to work when setting up things in LOR S3 as DMX channels (in different universes) and in xlights as a single 4096 range ... so testing on channesl 512,513,514 (in string mode) works, and testing on channels 515=red on node1 516=green on node1 517=blue on node1, etc.  ...  but I can't seem to get the LOR S3 sequence to play through xlights for some reason.  Maybe I should read the LOR boards, as I recall someone might have documented how to do this already?

[taybrynn]

I did get it to work if I programmed my SSC to PN channels 1-363 ... and set the LOR S3 controller to to Light-O-Rama controller and "Regular" network and selected 512 channels.

I just don't know how to get pixelnet channels 513 and above working!

[taybrynn]

OK, I got channels 513,514,515 (string mode rgb channels to the flex strip) working ... but hacking the .lms file with an XML editor.

I took the channels #'s for the second set of 512 channels and renamed 1 to 513, 2 to 514 and 3 to 515 ... and then those channels now
work through xlights.

I went back into LOR S3 and it won't show anything via. the program GUI, but 513 is there and can be sequenced to thereafter, but it appears LOR S3 won't allow you to put anything more than 512, so it kind of cripples LOR S3 for use with pixelnet ... except you can hack this way via. the XML.

I don't know, maybe the only solution is to process the .lms file and just generate channel numbers above 512 with a script?

I'll wait for others to chime in, but I seem to have a workaround, but I'm not thrilled with it.

Here is an example from the .lms XML --

<channel name="513" color="12615744" centiseconds="21726" deviceType="LOR" unit="1" circuit="1" savedIndex="363">

change it to

<channel name="513" color="12615744" centiseconds="21726" deviceType="LOR" unit="1" circuit="513" savedIndex="363">

and now it works as channel 512, or the RED componant of string mode for the flex string using PN channels 513->875.

I did the same with other channels and they also worked, all used unit="1" and I just changed the circuit="" value to the PN channel #'s I wanted.

It may be that since LOR S3 is only DMX capable, its also inherently not able to assign channels about 512 in pixelnet with anything but 1-512 ... and
thus this is the workaround?

Would there be any way for xlights to read either the unit= value and basically add 512*<the unit#> to come up with the actual channel# in PN ? 

[charles59]

I use my custom player and file formats to play my show.  I use LOR to sequence.  I have found the way to calculate the actual channel number from an LOR lms script is using the xml attributes as follows (if network is not present, it is the same as 0):

channel  = (((unit-1)*16)+ circuit) + network*512

Now, I don't know if this is what xlights use, it may require the circuit modification (or perhaps if using pixelnet).

[taybrynn]

That sounds very good.

Don't get me wrong, I think xlights is absolutely wonderful.  It could easily be I'm doing something wrong here.

I noticed in xlights, the single network mode uses this formula:

xlights channel# = ((LOR_Unit_# - 1) * 16) + LOR_circuit_number

....

In the Multi-Network mode is seems to allow multiple networks to be mapped ... but again, I think the problem seems to be that Aux A= second xlights network, but 1-512 don't translate to 513-1024 for pixelnet.

The problem I had for mult-network mode was that I could assign the same come port more than one, so this might have worked for multiple DMX dongles, but when using a pixelnet flashed dongle, I'm thinking it has to be all 4096 channels assigned right there? ... and then translates into correct PN ch# by xlights??

It seems that for pixelnet, you could use networks for each pixelnet universe and multiply by (network-1) * 4096 ... so for regular network, add zero, for Aux 1 network, (pixelnet Univ#2), add 4096.

And then perhaps use unit# as the dmx multiplier (for PN sub universes (1-> ).

So pixelnet ch# = ((network#-1)*4096) + ((unit#-1)*512) + circuit# ... and you could roughly translate as follows:

LOR Network Name:  network#
-----------------------  ------------
Regular                    1
Aux A                      2
Aux B                      3
Aux C                      4
Aux D                      5
Aux E                      6
Aux F                      7
Aux F                      8

or for DMX types from LOR S3
Universe 1                1
Universe 2                2
...
Universe 8                8

I'm just thinking that maybe this hasn't been used with pixelnet over 512 channels yet??

[frankr]

Hi Taybrynn,

I haven't had a chance to experiment but if I read in the xlights app correctly the same equation is used for multi-networks as is for single.

So for the Aux A network you could get 512units *16channels each = 8192 (i.e. more than one pixel net universe)

I plan to play with this Tomorrow to validate my assumption

so you can set things up in S3 as if they were all LOR networks and xLights will take care of the translation.  This would probably be tedious to setup but I am thinking I should be able to hack a script together to convert my channels properly.

Right now my bigger challenge looks to be getting the sequences out of LSP cleanly. I am seeing some very weird behavior there.

Frank

[taybrynn]

Thanks interesting ... never thought of adding 512 16ch controllers ... but I think thats a tedious task, in terms of adding so many 16ch controllers
just so xlights can compute the higher channel numbers for you. 

So for a 120 node flex string, which is 360 channels ... I would have to add 23 16 controllers and then combine them with groups ... thats ugly.

I think an easier solution would be to have a 3rd network type which uses either just unit# only

PN ch#=((unit#-1)*512 + circuit#

or handles multiple pixelnet universes like this using the formula mentioned above.

[taybrynn]

 OK, I just tried adding 33 16ch controllers on the Regular network into a test LOR sequence.

The first 32 make up the 512 DMX channels I am wanting to skip or rather reserve for the DMX out on the SS HUB. ( would be used as needed)

So I just added one additional 16ch controller (#33 or Unit 21) to see if I could control the entire flex string and the first few nodes ok ... and I could !!

So then I converted some of the remaining channels in Unit 21 to RGB channels to see if I could control the first 4 nodes using RGB ... I could !!

Config:  I have a 120 node flex strip in hybrid mode programmed on the SSC for pn ch#513-835.

It does seem to work using single network node for me.  I would likely also work (as you said) in multi network mode if I wanted LOR on Regular network and
all the pixelnet stuff on Aux A network.  Another option might be to just go with all DMX on the LOR boards and that would use up some of the DMX512 channels, but
I was adding them anyways.  The only downside to that is I'd have to convert existing sequences to DMX and I"m not sure its worth it, if I could just use multi-network
mode instead in xlights.

Using the channel configuration window in LOR ... it wasn't too bad to add all these.  I think the LOR groups would save the day on the organization side.  I really love those
and the ability to convert to RGB channels is also useful.  It's very easy to create a string mode group for the flex string and then also turn the rest into 120 rgb channels and then group those into a flex strip node mode group.  So in a way, no need for hybrid mode if using those groups, unless its just faster performing or something.

THis must have been what xlights was intending.

[therealbigjim]

Did you get the super star add on. I would be intrested in seeing if it will program smart strings like it does the LOR CCR strings....

[taybrynn]

I don't have superstar, but one concern is that the highest version only supports 3xxx channels and I have more than that.  The auto sequencing looks interesting, for sure.  But I'm still a little skeptical.