Failing Universe but only when running as show - works in HU and Pix Editor

[Welsford]

I'm lost with what appears to be a serious software bug.
LOR Pro 4.3.24
E1.31 output of universe 90 to IP address on an AlphaPix Controller
universe 90 is assigned to the RS485 output
RS485 Output is connected to CTB16 (I have a mix of commercial metal boxes and residential boxes but all have same issue, all are on most recent firmware)

Using the hardware utility I can operate the dimmers of U90 as expected
Using the pixel editor I can play a sequence that operates the dimmers as expected
When I save the sequence in the pixel editor and the intensity file and then run the sequence as part of a show, this particular universe is ignored.
So it works with everything except when running using the scheduled show.
In this sequence dozens of other universes run when running as a part of the show.
The comm port listener shows the universe 90 opened to the controller
I feel like I'm missing a check box somewhere but I can't find anything wrong anywhere
I'm also using U91 in the same manner to a different set of AC controllers with the exact same issue.

Thank you in advance!!! 
Merry Christmas!!!

[k6ccc]

Very unlikely that it's a software bug.  Lots of people here are using DMX outputs from e1.31 controllers just fine.  More likely something that you're doing.

By chance do you have the channels in Universe 90 and 91 as available channels in the Sequence Editor portion of the sequence.  In other words, are the channels in the intensity file from Pixel Editor and the channel grid in Sequence Editor?

 

[Welsford]

So you are on to something here.  Definitely a software bug.

The controllers have ID numbers 1-3 on universe 90 and 4-5 on universe 91.

The pixel editor has all of the appropriate DMX channels for each prop based upon the start unit numbers.

In the sequence editor I had two blank controllers, so unit ID 1 - 16 ch and unit ID 2 -16 ch.  Both were listed as running on network Regular.

I removed all channels in the sequence editor except unit ID1 ch1 and now the channels play correctly while running as a show.

It shouldn't have mattered that those were there since there was no overlap given that the LOR units connected to the DMX output should have no signal crossover to the LOR regular network (which is unused at this point).  In other words, from a software perspective it didn't matter that there were LOR units on the end of that DMX chain.  They could have been any DMX device and there is no feedback to the software that that is a problem.  I should be able to run unit IDs 1-2 on the LOR network and it shouldn't matter.

So I've proven a software bug here it would seem.  I suspect it has nothing to do with the units though and more to do with opening and saving the sequence in the sequence editor AFTER saving in the pixel editor.  I seem to recall running into that in the past at some point as well.

Any thoughts on the above?  Hopefully it remains working

[k6ccc]

It's too strange as described.  I still think something is getting missed here - not a software bug.  I'm going to ask for confirmation on some very specifics to make sure your configuration is fully understood.  I'm going to describe what I think you are doing.  Please correct me if I'm in error of if I'm omitting something.

You have an AlphaPix controller which presumably is connected via E1.31.  On that AlphaPix controller, there are two outputs that configured for RS-485 output.  One of those has Universe 90 and the other is Universe 91.  On the output that has Universe 90, there are three LOR controllers (obviously operating responding to DMX commands, not LOR commands) which have Unit IDs 01, 02, & 03 - which would be DMX channels 01 - 16, 17 - 32 & 33 - 48 respectively.  On the output that has Universe 91, there are two LOR controllers which have Unit IDs 04, & 05 - which would be DMX channels 49 - 64, & 65 -80 respectively.  The Universe 90 & 91 channels are being sequenced in Pixel Editor as DMX channels.

In Sequence Editor, originally there were two 16 channel controllers configured on Regular Network Unit ID 01 and Regular network Unit ID 02.  They were not used, so presumably the channels were completely off at all times in the sequence.  Confirm that in both Pixel Editor and Sequence Editor, there are NO OTHER channels assigned (by that I mean they exist, not whether they are used or not).

What you are saying is that if the channels for LOR regular network Unit ID 01 (channels 1 - 16) and LOR regular network Unit ID 02 (channels 1 - 16) exist in the Sequence Editor channel configuration, then the lights on Universe 90 & 91 don't work.  I don't buy it - there is something that has not been told. 

A large number of LOR users are running DMX channels off of E1.31 channels - both as connected pixels and via a RS-485 output to other DMX devices.  Personally I'm using 17 universes for my Christmas 2017 show.  All of those are pixels connected to four SanDevices E1.31 controllers.  To the LOR software, a DMX channel via E1.31 then to some other device via a DMX output from an E1.31 controller is no different than a pixel connected directly to an E1.31 controller.

Did I capture your configuration correctly, or is there more than this that has not been mentioned.

 

[Welsford]

you captured it correctly.  The only nuance there is that 90 and 91 are on different alphapix controllers.

I honestly don't think it has anything to do with having the LOR controllers in the sequence.  I think the action of editing the sequence after the work in the pixel editor is what made the difference. 

I've got a help desk ticket open for a software bug that I need to submit the files for but there is something fishy there.  It's not clear when the show software reloads intensity files and sequences.  It's not on each run of the sequence.  That is easily reproducible.  In most cases the entire LOR software must be killed in memory in order for the new files to take effect.

[k6ccc]

I can answer your last part.  You are right that unless the show player runs low on memory, one sequence files are loaded, they stay there at least until the end of the show.  If you watch the Status Monitor, you can watch it reload sequences when a new show starts.  For example during my year round landscape lighting, I have an evening show and an overnight show.  Several of the sequences are in both shows.  When the overnight show starts, all the sequences are reloaded.  With that said, intensity files are different.  Intensity files are read by the show player almost in real time each time the sequence is played.  The are NOT loaded in their entirety at the beginning of a sequence and then played out the data ports as needed.  Nope, don’t work that way.  The show player only has to determine which data port to send each block of data and send it out.  LOR does not give out the details, but that’s the general answer one of the developers has said here.  That means that if an Intensity file is modified between platings in a show, the revisions will be seen next time the sequence plays. I personally have confirmed that part.  My entire show is intensity files, and I have watched a sequence play, walked inside and made some change, and exported the revised intensity file.  Next time that sequence played, I could see my revision.  BTW, that works just fine as long as you don’t try to export a new version WHILE that sequence is playing - don’t ask how I know that?

[Welsford]

The very last line is you mentioned is what I was running into.  This is a result of two issues:

1. Windows is stupid and locks the file.  The intensity file is opened when the show is loaded and the only way to release the file lock seems to be to unload all of the control panel stuff.

2. The pixel editor doesn't understand that the file is locked for writing and doesn't properly read the response code back when you go to save the intensity file.  Hence, it says saved and in fact it has not.  And that's the confusion.  Save it 20 times while it's locked and you'll go mad trying to find why the intensity file didn't update.

Thanks for helping my brain work on this.

[k6ccc]

What version of Windows are you running?  I guarantee that on both my Windows XP and Windows 7 show computers, you absolutely can write a new intensity file at any time during the show.  The only problem that causes is if you overwrite the intensity file WHILE that file is being played.  That results in your song crapping out.  I have MANY times exported a new intensity file during the show, and walked out a few minutes later to watch my changes.  Note that this was with SuperStar, not Pixel Editor.