Conflict between Setup and Layout tab channel/universe information

[nutz4lights]

Hey all,

I am STUCK!  I setup all of my outputs in the Setup tab.  I have 141 e131 outputs enabled, each with their own number of channels and e131 universe setting (all set in multicast).  I then went in and created all of my models in the Layout tab.  I am having a problem with two models (out of a much larger number!) which isn't bad... but I want them working obviously...  The one in particular is my oak tree, which utilizes 18 outputs in XLNC, setup as follows:

94 - 390 - 30543 to 30932
95 - 195 - 30933 to 31127
96 - 390 - 31128 to 31517
97 - 195 - 31518 to 31712

so on and so forth.

As you can guess, there are 130 pixels on the first universe and only 65 on the second and that pattern repeats itself.  The reason for this is that I am utilizing 9 outputs from my PixLite16 controller and running 195 pixels on each output from the controller.

The problem I'm having is that, in the model properties for the oak tree, and I have tried this multiple ways (i.e. using an absolute start channel of 30453 and setting 1 string of 1755, using the single string approach starting at Start Channel 1 with From Output set to 94)... for some reason, it thinks that it needs to put the 65 pixel segment on the first output in the repeating pattern... in other words, output 94 has 195 channels output 95 has 390 channels, so on and so forth...

And for the life of me I can't figure out why.  I have tried saving the Setup configuration, the models, closing the program, etc.  Why, even though the Setup is done correctly would the software thing that these were flip-flopped?

Thanks, 1/3 of my tree is not lighting up because it essentially is not sending signals to 1/3 of the strings that it thinks are in the wrong location, please help!!

-Louie

[gerry]

Hi,
I would suspect that your setup is not aligned with your Pixlite definition or physical connection of the outputs on the Pixlite.

1. use the DA_131 utility -Aus forum  to test the channels , by outputting to the selected channels. Next (or u can do this directly)
2. Turn on the channels for 94 ie 30543 - turn on 50 and see which light up  and then extend , then turn those off and do universe 95 till you are happy that exactly the rights one light up.
(this eliminates the model from the equiation)

3. If the above is correct , then only look at the model. By double clicking on the model name in the sequencer, open to the next level and then highlight say 50 of the first nodes in one color and see which light up.   


 

[nutz4lights]

Thanks Gerry,

I have DA-131 and just tried it and the PixLite is definitely setup and working fine (I can also run the PixLite in its test mode and all of the outputs fully light up).  I am playing this through the FPP and it is also setup correctly.  I really think this is in XLNC since the display of the information for that model is definitely showing that it is trying to put 65 pixels on 94, 96, etc and 130 pixels on 95, 97, etc. which is backwards from what it is supposed to be.  I guess I may have to try and delete that model, delete those universes and then recreate them... I just hate to do that because that is alot of universes/outputs to re-enter...

Any other ideas?

-Louie

[gerry]

What happens if u test via the xlights test function - selected channels - is that correct or wrong ?

If that is correct , try deleting the fseq for the seq and resaving it.

Next - copy the model to another model name and create a new seq with only that new model name in it and try .
Otherwise , post screenshots of your setup and model

[nutz4lights]

Ugghhh... this is so frustrating.  I really don't have a good way to test output from XLNC... the only thing connected to the lights is the FPP (the computers upload through wifi to the FPP).

I have attached screen shots of the Model Layout and also the Network Setup pages showing that the Network Output is setup as I wish it to be, but the software goes off and sets the model up in a way that should be impossible (which has to be a bug).

I didn't go as far as deleting and creating a new sequence yet.  I feel like if the network and model are setup properly, this should work, so I must be missing something.

Thanks again,

-Louie

[gerry]

Louie , I am a bit confused ..

If you are using the FP to test your output , then what u have defined in the xl setup tab is not relevant and not used . What will be used are the e131 definitions on the FP .

But why cant you test via xl test ? I assume that you have Pixlite 16 connected to a router/switch and the Fp connected to the same. Are you not able to connect your computer to the router/switch and turn xl Output to lights on and then test directly ?

Looking at your model definition , I think that 94 has 195 defined (not 390) ie the list line of the screen shot ?

[nutz4lights]

Gerry,

Is that really true about the network setup tab in xlights?  I am definitely a complete xlnc noobie, but I thought that was the first thing we had to do?  Why does the model setup seem to rely so heavily on the xl setup tab?  (as you could see from the picture, it is even taking the output number and channel numbers from the xl setup tab).

The way my show is setup, I have the FPP master in the garage going out to a switch in the yard with a really long CAT5 cable and then there's an FPP slave on the other side of the yard and another switch.  There isn't a computer in the garage, I upload through wifi to the FPP master and then initiate the playback.

Can you clarify your last comment?  I'm a bit confused.  The model definition is definitely where i'm having a problem.  It does show that 94 has 195 not 390 but it is doing that automatically and it isn't correct.  That is my problem.  I can't seem to change that, it is automatically filled in.  That line is exactly opposite of the network setup tab for that output... can you suggest a way to change it?

-Louie

[nutz4lights]

Ok, I just clicked that "Individual Start Chains' box and it is letting me edit it now... I will report back.

-Louie

[nutz4lights]

Ok, that still didn't do it.  I am attaching the FPP channel configuration screen showing what I have for that as well.

Like I said above, if I switch the PixLite into "Test" mode, it lights all of the lights on those outputs.

-Louie

[gerry]

Gerry,

Is that really true about the network setup tab in xlights?  I am definitely a complete xlnc noobie, but I thought that was the first thing we had to do?  Why does the model setup seem to rely so heavily on the xl setup tab?  (as you could see from the picture, it is even taking the output number and channel numbers from the xl setup tab).

The way my show is setup, I have the FPP master in the garage going out to a switch in the yard with a really long CAT5 cable and then there's an FPP slave on the other side of the yard and another switch.  There isn't a computer in the garage, I upload through wifi to the FPP master and then initiate the playback.

Can you clarify your last comment?  I'm a bit confused.  The model definition is definitely where i'm having a problem.  It does show that 94 has 195 not 390 but it is doing that automatically and it isn't correct.  That is my problem.  I can't seem to change that, it is automatically filled in.  That line is exactly opposite of the network setup tab for that output... can you suggest a way to change it?

-Louie

The models define which channels each model is going to use. It does not rely on the setup tab at all.


If you using xlights to run your show (ie as your player) or want to test your show directly to the controller and actual lights , then xlights will use the setup definition to map the channels to the Universes and use the multicast/unicast /ip definitions to send data to the controllers.

If you are not using xlights to run/test your show ie you are using the FP , then the 'setup', must be defined on the FP  (ie on the E131 screen) and then FP (as the player) will use the FP E131 setup definition to map the channels to the Universes and use the multicast/unicast /ip definitions to send data to the controllers.
In this case , all u need in xlights setup screen is "any" definition to cover all models - it can be one block of say 20000 NULL definition.

What do I do ?  Define everything correctly on the xlights setup , connect my laptop to the show router via Cat5 , use the Output to lights to run and test from xlights and then duplicate exactly the same setup from my xlights to the FP definition finally.
 
Where your PI master is in the garage , can u not disconnect the PI from the Ethernet cable and plug that into you computer ? Or plug the computer into the switch in the yard ? Or are you using a desktop ?

Note - there is a Bridge option ... but am not sure if I should describe that , as it is a another layer of potential confusion.

Hopefully u have managed to change your model defn

[gerry]

Unfortunately , I don't know that way of definition , so will have to have someone else respond.

Have u managed to change the model , such that 94 has 390 defined ? Please try from output as 1 (am just guessing here).

[nutz4lights]

Ok, SOOOOO... even though the lights were all working just fine with the PixLite in Test Mode... it turns out that the issue with this whole thing WAS the PixLite configuration...  I honestly have not dealt much with controllers that do multiple universes out of a single output (I have numerous J1SYS p12__ controllers).  It turns out, even though in my head I thought I would do 130 pixels on universe 94, 96, 98, etc and then 60 pixels on universe 95, 97, 99, etc... the PixLite config automatically just fills out the universe for you... which means it was putting a full 170 pixels on universe 94, 96, etc... and because I had told it that there were 195 pixels on that output... only 25 pixels were going on universe 95, 97, etc... so basically, the 40 pixels 131 through 170 that I thought were going on the odd universes were not lighting up because they were actually residing in a different universe out of the controller...  I adjusted the channel output from the FPP configuration screen to match the PixLite and whoila... all is well.

Thanks for helping me way talk and think my way through this, Gerry.  It was a learning exercise and hopefully someone will come across this in a search at some point down the line...

Take care,

-Louie

[gerry]

glad you got it working Louie

[nutz4lights]

Me too!  Everybody seemed to love it... No music this year, maybe next...  Thanks again!