Bad apple!! LF version

[masterpj]

Bad apple LF version.


LF means Laser Friendly.

It's not really perfect yet but it have removed and photoshopped some parts frame-by frame. like the tree and I removed the shadows on most of the characters.

I also took -bart-'s advice by instead of the constant swapping between white and black characters I made it so the characters are most of the time white.
He told me it was easier for the tracer then.
I also did my own tracing, as most of the ones i've seen (no offence) had very jumpy lines).

it's on the photonlexicon ftp
in the following directory:
/Laser Shows/ILDA Frame Shows/Masterpj/Trace Content/Bad apple!! LF version
Containg the avi and the ilda.


Bad apple is getting a bit too overused though
now have fun people

[masterpj]

Bumping this as bad apple full color version is coming up in a few hours if this experiment succeeds!

[soforene]

Damn You !!

As you know, I am only 8 years away from producing my own 2 colour version !!

[masterpj]

I found a way.
Currently first attempt failed but i am very close to success.
(it caused a program to crash but i know why)
So adobe illustrator is going to assist me making the process work )

The first go wont be perfect once it's done (color wise)
but an example frame of an automatic colored frame is this:
*the colors were grabbed automatically*


@soforene, i doubt it will be as accurate as a hand colored version though

[masterpj]

There! with loads of tinkering the process is finally starting.
I need to have a .dxf sequence to use the image colors of the MMD color version.
(if i feed ildsos with an ilda it will crash when you want to use an image sequence for color, despite having an equal amount of frames).
The only way to do this is through an automated process in Adobe illustrator.

Currently I made an macro in adobe illustrator to do the tracing, it required me to hack a bit into illustrator to make use of custom tracing presets (to trace the video editted bad apple shadow art which has the same amount of frames as the colored image sequence).

The macro is slow so it has to work overnight.
After that I end up with .ai and nameless files (because of a bug it doesn't properly output .dxf in the macro)
So i just remove all files with the .ai extension and then use a batch extension renamer to rename all () extensions *() because there isn't an extension to the spit out sequence files by illustrator* * to dxf extensions


Then after this i throw both .dxf sequence together with the image sequence and let ildsos do the job, hopefully it wont crash because of this.

Then after that I have to open this in LD2000 and zoom in so much that it gets rid of the borders created by the dxf through illustrator (and do this in batch through making this progress happen in a block of frames)
I note down the upsize value's I used and then after the block of frames is sized up, i size them down again the outer border is gone.
Then after that I export it to full color ilda and optimize it (since the blanking needs to redone)

If this works then i'm done

Pfew... a lot of stuff yes

*edit*

DXF export and illustrator tracing has been done overnight and looks really good!
Took a long time for ildsos to read the files and he is currently optimizing it... then after that i hope he wont crash at using the colored images.


Shit! it crashed again.. something is wrong... perhaps the aspect ratio of the image and .dxf ???
Contacted drlava, it is really ready to be done but only the crash is preventing me from finishing this.

[teslablade]

hey masterpj, i was trying to use the bad apple .ild and it works great, but i can't find the framerate i have to use (in LaserBoy) to be able to match it up to the laser friendly video. i tried for half an hour, and the speed was always skewed and it didn't go out as i planned. by the way, I'm not using a laser, but I'm planning to build or buy a laser scanner. I'm using a hacked EVF CRT as a x-y oscilloscope. can you tell me the framerate you used when you traced Bad Apple?
thanks
-teslablade

[Grix]

hey masterpj, i was trying to use the bad apple .ild and it works great, but i can't find the framerate i have to use (in LaserBoy) to be able to match it up to the laser friendly video. i tried for half an hour, and the speed was always skewed and it didn't go out as i planned. by the way, I'm not using a laser, but I'm planning to build or buy a laser scanner. I'm using a hacked EVF CRT as a x-y oscilloscope. can you tell me the framerate you used when you traced Bad Apple?
thanks
-teslablade

It is 30 frames per second

[james]

Something that is important to know about LaserBoy is that it only does its best to approximate a frame rate when exporting to an optimized wave.

You set the parameters you want for max lit and blank distance and dwell. Those things usually add points to the frames.

When LaserBoy renders a frame set to wave it does so in whole frames. You will never get a partial frame scan. If you set your frames per second to, let's say, 10, then each frame should be displayed for 1/10 of a second.

LB looks at the number of samples in each optimized frame and figures out how many times that frame should be displayed to last for 1/10 of a second.

Some frames will need to be repeated. Each frame of a frame set will be displayed at least one time, no matter what.

LB does not add any blank padding between frames.

So frame sets with frames of fewer points tend to render most accurately.

If you have a frame set that was rendered from some other, non-laser application like an svg animation, there's no telling how many points you'll get in each individual frame.

Keep in mind that typical wave exports from LB are at 48K samples per second. So 1/10 of a second is 4800 total samples.

That includes all the optimization points and the transition to the next frame.

So, if you have a frame that has more than 4800 total points, it will be shown once and there is no way it will be 1/10 of a second.

That's why sometimes it looks like your animation is slowing down or speeding up..... because it is!

But your scanners are always happy!

The best thing to do is to make sure that your original frame set is as stripped out of as many unnecessary points as possible before you optimize it and render it to wave.

You can also fiddle a bit with the optimization parameters for better timing results.

@teslablade, if you are using a CRT (and not scanners) you can do whatever you want with optimization parameters. The only thing that you need optimization for is to make more visible points.

On a CRT, a shorter the distance between points tends to make a denser, more visible line.

So you can think about blanking spans. You don't want to add any points to them. Set your max blank distance to 65535.

http://www.photonlexicon.com/forums/...861#post309861

James.