X lights producing corrupt FSEQ file.

[Ebuechner]

I had a problem that I was originally blaming FPP for with one sequence that all the lights would go out randomly during playing.
After a bunch of testing I've narrowed it down to  Xlights because the same FSEQ file that will play in Xlights will not play in Xscheduler. I've had a few other people try it with the same result. And no check sequence doesn't come up with any problems.

[Gilrock]

So are you using FPP or xSchedule?  And you mention playing the FSEQ in xLights but xLights doesn't really play FSEQ files.  It plays what's in memory and theoritically that is what gets written when an FSEQ is generated but it doesn't really read from an FSEQ to do playback.

[Ebuechner]

Trying to run the sequence from FPP. I used Xschedule as a test to see if the problem was in FPP or if it was the FSEQ file itself.
Even CaptainMurdock tried this sequence from his server and had the same result of a corrupt FSEQ file.
Something in the sequence is causing problems. Some of the problem is the fan effects on the house with matrix group. But I shouldn't be getting a corrupt FSEQ file. Whatever is going on check sequence doesn't seem to be recognizing it.

[dkulp]

Any idea how to reproduce or what I'm looking for?   I took the sequence, rendered it on my mac, uploaded to a Pi with the models and created an HTTP Virtual Display, ran it, and it looks like it was working properly.  I didn't see anything wrong.  That said, I may have missed something as I wasn't sure what I was looking for.

[Ebuechner]

That should have reproduced the problem. I originally posted this problem on falcon Christmas website thinking it was an FPP problem. Chris and others were able to reproduce the problem doing exactly what you did. I think the only difference being that you're using a Mac. I don't know about anybody else, but I'm running a Ryzen 9 processor using Windows 10.

What you're looking for is all the lights to just shut down for a few seconds randomly, especially when the sequence gets to the portion where all the fan effects are on the whole house with matrix.
On the virtual display it'll look like it just stops for a few moments, but in the actual display all the lights go off.

[Ebuechner]

Here's a copy of the FSEQ file that's bad.

[dkulp]

I've uploaded my fseq file and a screen grab of the HTTP Virtual Display to:

https://dankulp.com/cay/

[Ebuechner]

Your FSEQ file works fine on my FPP. The file made from my computer produces a bunch of errors in the FPPD Log.

2023-12-04 08:55:03.418 (17920) [Sequence] /opt/fpp/src/fseq/FSEQFile.cpp:1015: Serious problem reading sequence data
2023-12-04 08:55:03.418 (17920) [Sequence] /opt/fpp/src/fseq/FSEQFile.cpp:1016:     Block: 8 / 52
2023-12-04 08:55:03.418 (17920) [Sequence] /opt/fpp/src/fseq/FSEQFile.cpp:1017:     Offset: 8
2023-12-04 08:55:03.418 (17920) [Sequence] /opt/fpp/src/fseq/FSEQFile.cpp:1018:     Offset+1: 8
2023-12-04 08:55:03.418 (17920) [Sequence] /opt/fpp/src/fseq/FSEQFile.cpp:1020:     Size: 3256782
2023-12-04 08:55:03.419 (17920) [Sequence] /opt/fpp/src/fseq/FSEQFile.cpp:1021:     Max: 3249954
2023-12-04 08:55:03.419 (17920) [Sequence] /opt/fpp/src/fseq/FSEQFile.cpp:1023:         Block 0:    Offset: 1487    Size: 0
2023-12-04 08:55:03.419 (17920) [Sequence] /opt/fpp/src/fseq/FSEQFile.cpp:1023:         Block 1:    Offset: 1487    Size: 0
2023-12-04 08:55:03.419 (17920) [Sequence] /opt/fpp/src/fseq/FSEQFile.cpp:1023:         Block 2:    Offset: 1487    Size: 0
2023-12-04 08:55:03.419 (17920) [Sequence] /opt/fpp/src/fseq/FSEQFile.cpp:1023:         Block 3:    Offset: 1487    Size: 0
2023-12-04 08:55:03.419 (17920) [Sequence] /opt/fpp/src/fseq/FSEQFile.cpp:1023:         Block 4:    Offset: 1487    Size: 0
2023-12-04 08:55:03.419 (17920) [Sequence] /opt/fpp/src/fseq/FSEQFile.cpp:1023:         Block 5:    Offset: 1487    Size: 0
2023-12-04 08:55:03.419 (17920) [Sequence] /opt/fpp/src/fseq/FSEQFile.cpp:1023:         Block 6:    Offset: 1487    Size: 0
2023-12-04 08:55:03.419 (17920) [Sequence] /opt/fpp/src/fseq/FSEQFile.cpp:1023:         Block 7:    Offset: 1487    Size: 0
2023-12-04 08:55:03.419 (17920) [Sequence] /opt/fpp/src/fseq/FSEQFile.cpp:1023:         Block 8:    Offset: 1487    Size: 0
2023-12-04 08:55:03.419 (17920) [Sequence] /opt/fpp/src/fseq/FSEQFile.cpp:1023:         Block 9:    Offset: 1487    Size: 0
2023-12-04 08:55:03.419 (17920) [Sequence] /opt/fpp/src/fseq/FSEQFile.cpp:1023:         Block 10:    Offset: 1487    Size: 0
2023-12-04 08:55:03.419 (17920) [Sequence] /opt/fpp/src/fseq/FSEQFile.cpp:1023:         Block 11:    Offset: 1487    Size: 0
2023-12-04 08:55:03.419 (17920) [Sequence] /opt/fpp/src/fseq/FSEQFile.cpp:1023:         Block 12:    Offset: 1487    Size: 0
2023-12-04 08:55:03.419 (17920) [Sequence] /opt/fpp/src/fseq/FSEQFile.cpp:1023:         Block 13:    Offset: 1487    Size: 0
2023-12-04 08:55:03.419 (17920) [Sequence] /opt/fpp/src/fseq/FSEQFile.cpp:1023:         Block 14:    Offset: 1487    Size: 0
2023-12-04 08:55:03.419 (17920) [Sequence] /opt/fpp/src/fseq/FSEQFile.cpp:1023:         Block 15:    Offset: 1487    Size: 0
2023-12-04 08:55:03.420 (17920) [Sequence] /opt/fpp/src/fseq/FSEQFile.cpp:1023:         Block 16:    Offset: 1487    Size: 0
2023-12-04 08:55:03.420 (17920) [Sequence] /opt/fpp/src/fseq/FSEQFile.cpp:1023:         Block 17:    Offset: 1487    Size: 0
2023-12-04 08:55:03.420 (17920) [Sequence] /opt/fpp/src/fseq/FSEQFile.cpp:1023:         Block 18:    Offset: 1487    Size: 0

[dkulp]

I think I've figured this out.   The uncompressed size of your sequence has exceeded 4GB in size (exceeded a uint32_t) which is causing a calculation issue.   

[Gilrock]

Wow just how many pixels do you have in your show?

[Ebuechner]

So this brings up a couple of questions.
I have 73 sequences and this is the only one acting up on the same layout, why don't the others act up?
Why did the FSEQ file that you produced work when the one that I have doesn't?
What can be done to prevent this?

[Ebuechner]

And Gil, it wasn't my fault. I was left unsupervised.

[dkulp]

So this brings up a couple of questions.
I have 73 sequences and this is the only one acting up on the same layout, why don't the others act up?
Why did the FSEQ file that you produced work when the one that I have doesn't?
What can be done to prevent this?

With your layout, any sequence under about 4m, 15s wouldn't hit it (25fps).   For longer sequences, the overflow may be enough of an overflow to allow enough compression blocks.  Bunch of possible variables.

 

[JSOrange97]

I'm also chasing an issue with an FSEQ file - rendered and plays fine on 2021.38.  Re-rendered in 2023.16 and it flashes constantly EVERY pixel and every DMX channel. Re-rendered in 2021.38 and its fine again.

[Ebuechner]

I was wondering if going into my 2024 show if there's anything I can do to avoid this problem. Right now this is only affecting one sequence, but there are things I want to add to my show. I tried rendering that sequence at 50 ms but it looked like crap at that frame rate. I looked to see if I could edit that song down to less than 4 minutes, but it really doesn't have a good point that I can edit. As I add things to my show, I can only imagine that this problem will get worse. And I know there's people out there that have much larger displays than mine.