DiyLightAnimation
Software => Nutcracker: RGB Effects Builder => Topic started by: smeighan on March 19, 2012,
-
Just Kidding.
well sort of. You may know that I have not yet used any of the three sequencers (Vixen,LSP or LOR). So to test LOR, I launched LOR and the lms file imported with no issues. Only later did I realize that i had launched LSP.
So, I assume LSP can import an lms file. So go ahead and try the lms files i am creating in LSP.
lol, i guess this shows i am an idiot...
-
I have all three if you need help testing
-
I have all three if you need help testing
you could try the lms file i am creating in both lor and lsp and see if they look ok.
thanks
-
Are those posted somewhere or do I need to log onto your site and generate myself? I haven't yet generated anything on your site, but I'm getting closer to trying.
-
Are those posted somewhere or do I need to log onto your site and generate myself? I haven't yet generated anything on your site, but I'm getting closer to trying.
You have to genberate an effect then the links show up for vixen, lor lms and lor lcb
You can just make up a hypothetical tree, like 50 strings with 50 pixels. Once you create a target model, you can then try all of the effects out.
-
Ok good ... thats what I was thinking. Very innovate way to do this.
-
I have made some major changes to the the LOR lms file output. The butterfly animation now loads into LOR with no issues.
Here is what it looks like inside the LOR sequencer
(http://meighan.net/nutcracker/images/FLY_0_0_lor.png)
and here is what it looks like inside Nutcracker
(http://meighan.net/nutcracker/effects/workspaces/2/ZZ_ZZ+FLY_0_0_TEST.gif)
Can someone who has LSP try importing a lms file?
Here is the one i made
http://meighan.net/nutcracker/sequences/ZZ_ZZ+FLY_0_0.lms (http://meighan.net/nutcracker/sequences/ZZ_ZZ+FLY_0_0.lms)
it is 20 mbytes in size. It should be 10 seconds in duration
-
Hello Sean, I was able to import the sequence using LSP Version 2.0.11226.2226 which is not the most recent build of Version 2 of the software, I am pretty sure there have been two newer builds. The Hp Pavillion laptop I tried it on is two years old and has an AMD vision quad core processor so it's not the best laptop out there but good enough to do my sequencing on. I imported the lms file and it went through smoothly with the exception of the effect optimization which took about 15-20 minutes. It looks amazing and runs smoothly, I just wonder if it would load quicker on a better computer, hopefully we will get some more feedback to see if that is the case. I know you have been thanked a million times for doing this, but I just wanted to say it again, so thanks! <res.
-
Sean, i gave this file a go and imported the LMS file into LSP V2, the file loaded without any errors.. The issue is that the controllers and channels imported as individual channels and not as RGB channels as one would expect and as you would have normally set up with RGB lighting so this may be something that needs looking into. Another issue is that the controllers missed the first 3 channels so the channel start is channel 4. One other note and this may be delibrate to save time for this example is that each controllers channel number are exactly the same. The controllers are based on LOR controllers which have controller numbers as where E1.31 and DMX use channel numbers and universes to map out the controllers not a biggie as this can be set up later manually by the user to his channel range, the issue is these are set in LSP as a light controller and not an RGB controller and if changing the controller back to an RGB controller turns all the sequenced channels white thus making the imported effect all white.
Anyway great work Sean its really coming along and growing into one awsome program. Thanks for the effort.
[ATTACH=CONFIG]13628[/ATTACH]
Hi,
missing the first pixel is because i do all animations locgically from the top of the tree (Nutcracker Pixel#1). When people build their tree targets they give me dimensions that dont work out evenly.
Let me give a simple example. If a have a 50 pixel string , folded in half. I have 25 pixels per strand. Assume you said 3" spacing between pixels. If you give me a height for your tree that doesnt add up to having a 75" hypotenuse. I will come to the bottom of the tree without having reached your last pixel. This is happening alot in peoples target models. When that happens i will not reach pixel #1 and pixle #50 while i am creating the animation.
I am going to maybe change how the target model works so that i ignore your height and just use pixels and the hypotenuse.
Anyways, that is why you see some pixels missing.
ill work on gettng it fixed.
I also wish i could do more for LSP.
-
This is what i get
(http://i43.tinypic.com/hukk7t.jpg)
-
Sebastian, How long did it take to load?
-
Not long same as other LOR imports. I have version 2.0.11271.743 on a 5 year old Dell E6500 Laptop with Core 2 Duo processor
-
Not long same as other LOR imports. I have version 2.0.11271.743 on a 5 year old Dell E6500 Laptop with Core 2 Duo processor
When i loaded the 20 mbyte file in LOR it took 8 seconds
When i load the same file into LSP it took 37 seconds.
I have an HP desktop computer
Windows Experience Index of rating of 5.9 (right click computer, properties to see your number)
Processor: Intel i7 4 core, 8 threads at 2.67hz
Windows 7, 64 bit
Memory: 12gigs
Fast eddy made the point that when loading into LSP it has to first convert the lms file into msq. Then it loads the msq file.
Here is a video of me using LOR and LSP to load this sequence. Note: I have NEVER used vixen,lsp or LOR yet so there could be an operator error here someplace.
http://meighan.net/nutcracker/tutorials/LOR_LSP.swf (http://meighan.net/nutcracker/tutorials/LOR_LSP.swf)
-
I tired loading it into LSP 2.0.11338.1054 it took about 37 seconds (had to select and audio file) and the result was the same as posted by Sebastian as far as appearence. Running on AMD Phenom II X4 840 with 8 gigs memory with Windows 7 64bit
-
Same here... looked like Sebastian's import...
-
Not long same as other LOR imports. I have version 2.0.11271.743 on a 5 year old Dell E6500 Laptop with Core 2 Duo processor
When i loaded the 20 mbyte file in LOR it took 8 seconds
egads - 1 model, 10 seconds = 20 MB
now I deal with multi-gig video files all the time, but still, these are ultimately going to be huge data files when you are talking about a whole show.
-P
-
Not long same as other LOR imports. I have version 2.0.11271.743 on a 5 year old Dell E6500 Laptop with Core 2 Duo processor
When i loaded the 20 mbyte file in LOR it took 8 seconds
egads - 1 model, 10 seconds = 20 MB
now I deal with multi-gig video files all the time, but still, these are ultimately going to be huge data files when you are talking about a whole show.
-P
Yep, not alot i can do. The massive rgb devices will produce massive files.
The 10 second file was from a 32x64 tree. This tree will have 2048 pixels and need 6144 channels. I was running a 50ms frame.
I will need 122,880 bytes of info per second. In 10 seconds 1,228,800 bytes of RGB. Now you have the overhead of having these bytes be in an ASCII xml file.
The 10 seconds file ended up as 20mbytes. I am expecting LSP xml file to be at least as large.
So if i create all the effects for a megatree for a 3 minute video, 600mbytes of xml.
I am hoping RJ's conductor might offer a way to save some of this.
Also dont try 25ms frame timing unless you really need it. That would double the size of the file.
If there is intelligence in the xml that would allow looping parts, i could take advantage of that in a future export.
I already have logic so that i dont output data if the rgb value is zero
Here is a sample
<channel name="R1-P3-G" color="65280" centiseconds="10000" deviceType="LOR" unit="1" circuit="8" savedIndex="4">
<effect type="intensity" startCentisecond="50" endCentisecond="55" intensity="15" />
<effect type="intensity" startCentisecond="55" endCentisecond="60" intensity="50" />
<effect type="intensity" startCentisecond="60" endCentisecond="65" intensity="88" />
<effect type="intensity" startCentisecond="65" endCentisecond="90" intensity="100" />
Notice all the overhead to finally be able to start outputting the GREEN value for this string.
Most of the nutcracker animations do one cycle and then repreat. sure is a waste of speace to keep sending out the same animation sequence, say 6 times.
i do believe space management is going to be something i will try to work on.
Maybe LSP 3 (summer 2012) and LOR S3 have some xml features that allow the files to be smaller.
thanks
-
I think we are already seeing that very large RGB sequences (hopefully turbo charged with Nutcracker effects in 2012) ... are likely to require some kind of hardware based playback, like what the etherdongle conductor daughter card would allow for This means that regardless of whether I'm using LOR, LSP ... I might be converting my sequences into .xseq and putting it onto a memory card for playback. It least based on the reports from RJ that he's playing very large sequences with zero lag ... which is AWESOME !
After experiencing more lag than I've ever seen before last year, I'm convinced that you must minimally use DMX / or / PIXELNET & DMX as the protocol ... as LOR protocol (on half my show) was the main bottleneck for me, even using only approx. 1 pixelnet universe of channels. As soon as I switched to DMX control over those same LOR controllers, using the same base sequences, the lag went away ... and I saw the sequences come to life in a whole new way (for the better). If course this all happened after Christmas, so nobody other than the fire department guys got to see it.
I also saw some lag on some distant pixelnet based SS nodes ... but I think the pixelnet signal itself was getting degraded from (crossing) bundles of power cords (because when I would turn on my neighbors roof ... lots of C9s ... lots of power ... I'd see degraded control/response in a distant pixelnet based display item in his yard) ... but once I turned his house lights down or off ... I'd see that same display item become responsive again, without the lag.. I think once I can get the pixelnet signal insulated from the power cords (told my neighbor to do this, didn't happen) ... the signal won't degrade and I'd get rid of that issue. Another option might be to someday run a second etherdongle over there and use 802.xx wireless to bypass all the wiring from my house, across my yard and over his roof.