Problem possible bug: setup universe number doesn't match _networks.xml

[scuba]

I was having issues with channel numbers and testing the light setup.  I checked the xlights_network.xml and it didn't have all of the 50 universes I setup in setup. IT's missing universes 32-50.  Idea's?

SCREEN Shot is just part of the setup screen

[scuba]

update:
I add a new universe and was able to find it in the _network.xml file at the end of the file.  Since I'm in the middle of bugging the channels, I decided to delete universe 32-49, then add them back in...only added universe 49.  all the others were not added.

Help me figure this thing out!  Halloween is tonight.

[Phrog30]

Your screenshot shows otherwise. If you are having issues just delete and start over, it takes 30 seconds.

Sent from my Pixel using Tapatalk

[scuba]

I check last years xlights_network.xml file. I can see universes up to 45 and LOR information for 3 controllers...all of which is lost in the current file. 

The screen shot is a partial due to the length of the screen.

[Gilrock]

If you reopen xLights and the universes are there then they are in the file.  Maybe you are looking at the wrong directory when you inspect the file.

[scuba]

Hey Gil
I deleted them all and reentered them..looks good.  I was in the correct directory...failed to mention I'm on a MAC.

thanks

[Gilrock]

You do have to hit the "Save Setup" button for those network setup changes to be saved.

[scuba]

I've spent some time trying to account for a 48 channel shift between the setup/models and the test feature.

A bit of a long story: I've connected all of my controllers to my network.  SanDevices (10).  same setup as last year with some changes to get a few new features connected.

If you look at the setup screen shot, you will see that my first network type is 48 channels of LOR, followed by 49 universes of E1.31.  The first E1.31 universe channel number is 49.  Note most of the num of chns are 510, with several configured to allow a model to bridge across two controllers: standard method I've used for years.

I've attached my xlights_networks.xml and xlights_rgbeffects.xml files.

In TEST I turn on any model and I get a 48 CHANNEL shift in the actual RGB LEDs (the first 16 pixels are dark) and 16 pixels in the next model (by channel number) are on.

So I'm looking for the 48 channels. 

Your thoughts

[Gilrock]

Hmm...48 LOR channels and a 48 channel offset.  That's not obvious at all where the offset is coming from...

I'd say a bug in the Test code.

[scuba]

It may not be TEST.  I first noticed the issue with a sequence running with the lights ON.  There were several issues with the  rgbeffects file.  It seems to get corrupted if you you make several changes to a model.  I went from start channel to indiv start chans and back again; only to find the individual channels were still in the RGBeffects file.  The model was a circle, with four rings.  yes I did save the layout!

 I'll try to use a sequence tomorrow that runs thru the channels to verify if the issue is TEST.  I'll also run a FPP test sequence.

Thanks for the help to "find the lost 48"....sounds  like a movie plot or campaign slogan 

[Gilrock]

Are you saving to a dropbox or something in the cloud?  That's when I start hearing about rgbeffects files getting corrupted.  Or one guy finally admitted he was saving to a USB drive...also not the best idea.

[scuba]

I'm not saving to dropbox or the cloud as a working file, but I do transfer some files from my mac to the PC...but I check the files after transfer...all good.

[scuba]

just ran a sequence with the above files....no joy the offset is still in play..tried several different models..same results.  Apparently its not the TEST software.

[Gilrock]

Ok well your problem description has changed once or twice since you started the thread.  Originally the problem was stated as universes 32-50 missing.  Then its switched to XML file corruption and a 48 channel offset issue.  I can only provide some troubleshooting tips or ask questions since this needs to be tested on your setup.

Lets just stick to the 48 channel offset since I can't fix the source of any file corruption issues.  Just pick one model after the LOR channels and turn on its first light (drop On effect on the node).  If its 48 channels off is that in your house preview, the real lights, or both?

[scuba]

Ok Gil good approach to solving the problem.

A simple sequence with only one model on, shows the 48 channel offset. I've tried the first  model after the LOR channels. same thing 48 channel offset; I've tried other models many channel away from the LOR channels...same thing.  I've reviewed the SanDevices setup...looks good.  Could it be due not having the LOR usb dongle inserted?

I've not tried FPP with these simple sequences.

Any other suggestions?  Kill the LOR channels? or change them to 32?etc.