Channels work in some sequences, not in others?

[Don]

Before you continue reading please note: This has ALREADY BEEN SOLVED. I'm posting this message to help people learn how what steps were taken, and how it was resolved. (Some numbers/networks changed, because, well, I want to. Does not affect the story.)

Recently a customer of mine wrote me back to say that on some of the songs the mini-trees were working just fine. However on other songs nothing would happen. This customer is using S5, and I did the programming of the sequences. As I'm writing this message please know that I'm writing this out in much the same way my thinking went while troubleshooting. I want you to see the steps taken to get to the resolution.

The first thing I did was write "Joe" (not his real name) and told him to check that the Previews were the same on all the sequences. There were new elements added this year, though the mini-trees were there last year. Thus my first thought was that somehow the wrong preview had been applied to the 'broken' sequences (Broken = mini-trees not working.) The customer spent the morning looking through the sequences and found that all of them are using the same Preview.

The next step was to launch a Zoom session and take a look at it myself. I mean, there is no way that this can be happening, right? It's the same preview on all sequences! What could be going on? So I open a working and broken sequence. I look at the Preview for each one and go through the String Summery and make sure they are on the correct channel. Yep! All of the mini-trees are on Unit 4 in the working and the broken sequences. Yep! All of them are on the REG network. Yep! I can see the trees (thanks to Joe's phone connected to Zoom as well sitting out in the display area) are not working when I can see events programmed for them. What is going on?

Since it was daylight I couldn't tell exactly which colors were lighting up on the trees, but the way Joe described what was happening started to unravel the mystery. "They are flashing Red/White then Green/Blue." (LED lights, btw.) Interesting, I thought to myself. Since the trees were set up in the Channel Grid in RGBW order, things started to come together. *1

By this point I had pulled in an extra set of eyes to look at the issue. I highlighted a section of sequence where the trees had NO programming, but were still lighting up, and then played that on a loop. Over and over it played, and then we started to look through the rest of the sequence and noticed that a portion of the sequence matched the effect we saw.

But wait ... how are the trees working in other sequences when NOTHING is programmed for them, and then NOT working with something IS programmed for them?

The problem? Are you ready? Any guesses? If you said the Unit ID was wrong on the controller, (1602Wg3) you are right! But wait, they worked some of the time, but not all of the time. How is that even possible? How would a wrong Unit ID lead to this?

In the heat of the moment, when all eyes were watching what I was doing, and the customer waiting on me to solve the problem I had forgotten one crucial element. When I do my programming I create a 'Working' channel grid (track for you S3/S4 folks) so I can put the channels in the order I want. This year the customer had removed some LED arches and replaced them with RGB arches. This meant that the previous 1602's that were running those arches were no longer in use. Thus I had removed them from my Working grid, however I never removed them from the preview; they were still in the programming - and they were sending commands to the mini-trees. The controller Unit ID for the mini-trees had been wrong for YEARS, but no one ever noticed because the trees worked and the arches worked. Joe mentioned today that sometimes he thought the effects on the arches were a bit odd, but with so many other things to worry about, he didn't get around to mentioning it. 

Only much later (hours after the call) did I realize that only the new songs were affected by this. As the LED Arches were not in the Working portion of the grid for the new songs they never got programmed. When it came time to run the sequences the mini-trees didn't run because nothing was programmed for them, even though I thought they were programmed. The old songs retained the programming for the 'trees' even though it was really the Arches.

Lessons Learned

1. Always troubleshoot from the "Show All Items" view.
2. Don't be a packrat! If channel or group of channels isn't going to be used, remove them from the Preview. If you want to keep the programming around for archival purposes, leave the data in the Archived channels. Not live and sending data, as I had done.
3. Don't be afraid to pull in extra help. I've been doing this since 2004/2005. Surely I don't need anyone looking over my shoulder 18 years later, right? Wrong. I'm still learning.

I hope you find this helpful. Please remember that the solution may be something so simple that you don't think to look for it in the beginning.
 

*1 - This meant the arches would have been filling from the base to the top of the arch, then turning off in the reverse order. Since the mini-trees are in RGBW order in my Grid, it was the first clue that gave me an idea where the programming was coming from, or even what the effect might 'look like' in the Sequencer.

Ps. Special thank you to the person who jumped in to assist.

[k6ccc]

Hell of a story Don!

Glad you finally figured it out.  So often it's those little details that get you.

 

[tlogan]

Great job trouble shooting!

Being an old school Cobol programmer, I could tell some stories about tracking down intermittent problems!

I have found that a couple of times when I had lights blinking when they shouldn't be, that it usually involved copied blocks of sequencing. Maybe it's just my eyes, but I finally figured out that where long fades end (of particularly blue and purple) become invisible in the sequencer. I had to enlarge as much as possible to see where the fade ended. The fix was to highlight a block and hit delete. Not to mention pasting a block of copied cells one row off of where you intended. Everything looks good, except that one light on the next prop below the one you just pasted to. And even though all my sequences are purchased, I found a few times where one entire row in the middle of something needed to be nudged right one or two cells.