OK, so I blew up my entire show today

[taybrynn]

So I had to reprogram a few SSCs because my roof had more nodes than I had indicated previous (was causing issues in show).

So I updated my models to match the new actuals.

I went into xlights to open each sequence, and now there is NO GRID -- nada.

So I don't know why, but apparently just changing the start and end channel on a few models has not detached nutcracker
from my sequences.

So I did try doing an new sequence option where you pick the xlights xml ... but that also did nothing.

So I am stumped.  Show is dead for now.

I'm hoping somebody has experienced this before.

Scott

[drlucas]

Yikes!

xlights_rgbeffects.xml file - can you open that in notepad and see if it looks corrupt at all?

Mine looks like this...

Code: You are not allowed to view links. Register or Login

<?xml version="1.0" encoding="UTF-8"?>
<xrgb>
  <models>
    <model name="garage-left" DisplayAs="Single Line" StringType="RGB Nodes" parm1="1" parm2="120" parm3="1" StartChannel="4900" StartSide="B" Dir="L" Antialias="0" MyDisplay="1" offsetXpct="-0.3500" offsetYpct="-0.0861" PreviewScale="0.8600" PreviewRotation="-90"/>
   
 blah blah blah

  </models>
  <effects version="0002">
    <effectGroup name="Favorites"/>
  </effects>
  <palettes>
    <palette name="green" color1="#FF0000" color2="#00FF00" color3="#0000FF" color4="#FFFF00" color5="#FFFFFF" color6="#000000"/>
  </palettes>
</xrgb>


[taybrynn]

No, it doesn't look corrupt at all.

In fact I went into notepad++ and compared the current xlights_rgbeffects.xml from my last backup (earlier today) with the current one ... and the only differences are the StartChannel="" numbers on the RGB roofline models.  So the same named model was StartChannel="1000" and now its StartChannel="1003" because it was say one RGB node (3ch) offset from where it was.

So I was just surprised that when I opened the .xseq there was no grid at all.  I re-installed the prior version, but either way no nutcracker grid with the timings in it.

I am currently playing the existing .xseq files and its play ok, but its not perfect because the models no longer agree completely with the way the SSC(s) are programmed.

Scott

[smeighan]

No, it doesn't look corrupt at all.

In fact I went into notepad++ and compared the current xlights_rgbeffects.xml from my last backup (earlier today) with the current one ... and the only differences are the StartChannel="" numbers on the RGB roofline models.  So the same named model was StartChannel="1000" and now its StartChannel="1003" because it was say one RGB node (3ch) offset from where it was.

So I was just surprised that when I opened the .xseq there was no grid at all.  I re-installed the prior version, but either way no nutcracker grid with the timings in it.

I am currently playing the existing .xseq files and its play ok, but its not perfect because the models no longer agree completely with the way the SSC(s) are programmed.

Scott

have you been doing "File,Backup" daily? Only take 3-4 seconds. This will ensure all sequences are backed up.
The grid doesnt come from models it comes from the sequence.xml files.

pick one and post it here. i can then see what they look like.

[taybrynn]

Yes I back up daily .   I have backups.    I can even run my latest xseq sequences ... But why aj can open the grid has me stumped.    I may restore my prior files into the main directory ... But like I said, doing a diff ... Only the model start channels have changed .

My issue is that I need to update my models to reflect the current channel starts, but doing so appears to make the nutcracker timing/models grid disappear for some reason.   


I just figured I would reopen the sequence, then when AI resave it ,,, I would get the new model generated into the new .xseq file.   

[taybrynn]

Ok, So I figured it out.

It was the ID10T error condition -- user error -- argh!!

 

As it turns out, I had the xlights window less than fully maximized and the grid just wasn't showing up at all.

Once I maximized the xlights window, there it was !!

SO EMBARRASSING!!! 

Scott

[taybrynn]

So while I can see the grid now and my models have been updated.  I have re-saved the .xseq and it appears that things are still messed up ... namely string mode channels and node level channels are conflicting with each other.

So to debug things further, I exported to LOR .lms file and then viewed in LOR the sequencing data.  Sure enough it shows that I have some kind of leftover sequencing on the hybrid/string channels and its also not writing the new model data correctly either.   

I'm thinking of adding a new string mode model for each of these and setting the effects to None,None to see if that'll clear the string mode channels for me.  Now what the new sequence doesn't write to the new channels ranges assigned to the model, I'm not sure.

[drlucas]

Good to know where the grid comes from....makes sense, was just thinking in this instance what changed last (a model) and the fact all sequences were impacted....Now what Scott found makes even more sense....and good to know the saving of the xmls is rock solid code!!

I have noticed in a few of my sequences if I am importing a sequence from say HLS and then overlay an xml file that I create manually by importing some timing marks and don't set the start time to 0 and add an effect to the beginning of a sequence I get some crazy effects going on by default. I will one day make a post that makes more sense than how I just described it...but I use dmx universe 9 in HLS for all my singing trees and LE controllers and then use pixelnet for the hubs and SSCs in universe 1-12. Could be a hardware compatibility issue, could be programming....could be user error. Anyways, left over sequencing is a good way to describe this...outside of the fact there is nothing that I'm aware of in my other universes when I import my HLS data file.

[taybrynn]

I'm finding that doing an export or LOR is my only way to really "see what the final sequence looks like" at a detailed (per channel) level.   It exports the entire 32k setup since that is what I have defined in the setup tab (only use 19k channels of those). 

But it's confirming that I'm getting something other than expected ... It's actually not generating effects into the exact new ranges the way you would expect.     I'm finding that there are a lot of "leftovers" that are not getting erased.   I'm learning a lot through trial and error.

[smeighan]

I'm finding that doing an export or LOR is my only way to really "see what the final sequence looks like" at a detailed (per channel) level.   It exports the entire 32k setup since that is what I have defined in the setup tab (only use 19k channels of those). 

But it's confirming that I'm getting something other than expected ... It's actually not generating effects into the exact new ranges the way you would expect.     I'm finding that there are a lot of "leftovers" that are not getting erased.   I'm learning a lot through trial and error.

Suppose you used 1-1000 as lor, 2000-3000 as your megatree. You save. You get an Xseq for all of that and it saves an xml for the song.

You now decide to move your megatree to 1500-2500 channels.

You will have leftover stuff in channels 2001-2500. This is because we cannot blank out channels that are outside our channel range, you might have loaded lor stuff there.

How to get a blank slate?

First "file,backup"

Delete the Xseq and the xml for your song!

Now create a new file using the audio file. Add one display item and one row into the grid.
Save it as xxx, so u have xxx.xseq and xxx.xml

Now rename xxx.xseq to the original Xseq name

Copy,from your backup directory, the latest xml file into your show directory.

Finally load in your lor file into this sequence.

Remember u don't need the Xseq files, they can be recreated as I just described.

Maybe we will add a button to do this next year

Thanks, sean

[taybrynn]

Many thanks Sean, I will try this out today!! 

[taybrynn]

Sean, I believe your process has worked!

The channels on my LOR export now look clean.

Since I also defined string models for the roofs, it also nicely shows each string channel as an RGB channel now ... with no programming in it.  WOW!

You are a true miracle worker !!!

Thanks!!!!!

Scott

p.s. I just played the new .xseq and everything looks perfect, so I can tell this process has worked!!!

[smeighan]

Sean, I believe your process has worked!

The channels on my LOR export now look clean.

Since I also defined string models for the roofs, it also nicely shows each string channel as an RGB channel now ... with no programming in it.  WOW!

You are a true miracle worker !!!

Thanks!!!!!

Scott

p.s. I just played the new .xseq and everything looks perfect, so I can tell this process has worked!!!

next year we will make a button that will do this.

1) Blanks every cell in the xseq.
2) Now when you press save, it rewrites the nutcracker stuff into the cells.

you would be done except you lost your LOR import. so
3) Import LOR again using the Open LOR button, now resave

we are just missing a button to blank the xseq file. we will add it next year. people need to know , if needed, to reimport lor,lsp,vixen,hls data after doing this.

my short term process will work for this year.

glad it works for you

sean

[arw01]

Ok, So I figured it out.

It was the ID10T error condition -- user error -- argh!!

 

As it turns out, I had the xlights window less than fully maximized and the grid just wasn't showing up at all.

Once I maximized the xlights window, there it was !!

SO EMBARRASSING!!! 

Scott

You are not alone on that one, I was making a sequence in a new version and the first time I used it, spent a good 20-30 minutes uninstalling, reinstalling, opening a new sequence, blah blah.  Drug the window bigger one time, and wala there was the grid to work in, felt like a total noob!

Alan