Unable to determine the length of this LOR sequence...

[sielbear]

I'm getting an error when I try to convert an exported LMS file from LSP.  The file is a 15 second lms file with approximately 12,000 channels.  The file size is 370 MB.  I've converted the LMS file using setnet.py (per mickpak's tutorial) as well as running the xmlsearchreplace.exe command to set the unit number to 0.  The full error message I'm getting is:

Unable to determine the length of this LOR sequence (looked for length of track 1).

Any ideas?

[frankr]

Hi Alan,

Take a look at the file and search for "track" you should find something like:

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

<tracks>
<track totalCentiseconds="13100" timingGrid="1">

If it is not there then you can probably fake it out by adding:

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

<tracks>
<track totalCentiseconds="1500" timingGrid="1">
                </track>
</tracks>


Good luck,

Frank

[sielbear]

Very interesting...

I loaded the file in Notepad++.  I see the /tracks at the end of the file, but the file starts with:

<sequence author="Alan Sielbeck" createdAt="8/19/2012 8:15:15 PM" musicFilename="C:\Users\Alan\Documents\LightShowPro\Music\Silence\15.wav" saveFileVersion="3">
  <channels>
    <channel centiseconds="113379" circuit="49" color="255" deviceType="LOR" name="T1R" network="0" savedIndex="0" unit="1">

TONS OF CHANNEL INFO..........

        <channel savedIndex="11902" />
      </channels>
      <timings />
    </track>
  </tracks>
</sequence>

So it appears I should have a <tracks> as you suggest towards the top of the file.  It's just missing.

One other oddity - why does the channel centiseconds show such a HIGH value?  I'd have thought on a 15 second file that it would be as you suggest, 1500.  In fact, the effect section looks like this:


  <channels>
    <channel centiseconds="113379" circuit="49" color="255" deviceType="LOR" name="T1R" network="0" savedIndex="0" unit="1">
      <effect endCentisecond="10" intensity="0" startCentisecond="5" type="intensity" />
...
...
...

      <effect endCentisecond="1500" endIntensity="31" startCentisecond="1495" startIntensity="32" type="intensity" />
      <effect endCentisecond="113379" intensity="0" startCentisecond="1500" type="intensity" />
    </channel>

It appears there is a start effect, and the first timing is created at 10 centiseconds, not 5 as you would expect.  Secondly, the file length isn't quite right as an effect is generated at 1500 centiseconds, lasting until 113379 centiseconds? 

This is output from LSP 2.5, exported to LightORama format. 

[frankr]

sounds like the LSP exporting code might be a little finicky now...

I wonder if there is something about your events that LSP did not like. Or if maybe your LSP sequence setup is a bit off? Did you add an audio file to the LSP sequence?

Frank

[sielbear]

I added a 15 second sound file.  It's exactly 15 seconds of silence - wav format.  I found through trial and error that setting an animation length of XX seconds didn't always mean your song was XX seconds.  With the media file, it should be exactly 15 seconds.  I used the media sequence wizard and specified the file as I always do.

I also zoomed in as close as I could and put the start of the transition at the VERY beginning of the file to try and be as accurate as possible. 

I wonder if version 2.5 is dropping the track length identifier...  Based on what I'm doing, it's probably not that difficult to add the duration.  Royal pain for anyone else trying to do this...  I bet a good friend of mine could script something together to add the line after prompting the user "how many seconds is your file?"

I will admit that I hit a limit with setnet.py.  It looks like that file uses approximately 5.4X RAM of the lms file it's parsing??

I was not able to run setnet.py on my 460 MB file (out of memory), but I was able to run it (at 2 GB or RAM usage) on my 370 MB file.  Oddly, the xmlreplace.exe script did run, but holy smokes did it use RAM.  On the 370 MB file, it used around 6 GB of RAM.  Yikes!!!!  That's for 15 seconds of my sequence!!!  I'm clearly going to be building this show in micro-sized pieces this year.  Guess that's the penalty for moving from 4,166 channels to nearly 12,000?

[frankr]

Ouch,  Think we need to set up a dropbox and start sharing some files so I can sort out what is going on.  Parsing all that XML is a lot of overhead but I should be able to optimize it...

[animal]

I'm getting the same error with a small sequence. around 6000 channels E1.31. I've tried a few different things editing the xml but still get the same error. 

Unable to determine the length of this LOR sequence (looked for length of track 1).

   animal

[sielbear]

The line you are looking for is towards the bottom of the file, I think just above the channel name index at the very end. In my file, it was essentially the number of channels in line numbers from the end of the file. If you have 6,000 channels, look around 6,000 lines from the end.

[animal]

I have the line. It's at the end like you said. Do I need to add the hack, because theirs nothing between the tracks comments and does it need to be at the top?

                                  I'm not a big coder but I can fiqure it out. 


               animal

[sielbear]

Just modify the duration - mine was outputting as Centiseconds="0" I believe.  Modify to the duration in 100ths of a second.  I.e. 20 seconds would be 2000, 15 seconds would be 1500...

[frankr]

That is correct. LOR does things in units of centiseconds.

Frank