Hasta La Motion!

I have several filmed interviews made with a locked-off camera and need to separate the interviewee (foreground) from the background, so that I can reduce background levels, defocus, etc. I’ve done this laboriously for some shots using a combination of roto, luma keying and even chroma keying, i.e.
http://blog.beyondblueeyes.com/2008/…ip-camera.aspx
but I’m wondering whether I’m missing a more obvious and less laborious way.

The interviewee is nearly always moving to some degree, and I’ve wondered whether difference matting would be an idea (I’ve tried basic difference matting, and it didn’t seem to work that great).

Thanks
Greg

Hi all,

Im gonna start with showing what I’ve achieved succesfully so far and then I’ll explain my problem.

http://www.youtube.com/watch?v=Q4OieoF8n_A

As you can see I’ve made a rotoshape that covers the DJ and I’ve put scrolling text in it.

This I have done the following way. http://users.telenet.be/sickbrain/Afbeelding%202.png

Voorbereiding2_57 is the footage, groep 2 the rotoshape and SFileIn1 a psd with the text that makes a left to right move.

So far so good.

However, the next shot I have 2 DJ’s at the same time. I’ve rotoshaped them both and put the 2 shapes in 2 different groups. I now want to put some yellow on blue text in one DJ and some blue on yellow text in the other DJ. They both are in the same shot and at some point they overlap so at some moment there is a forgroundDJ and a backgroundDJ.
I’ve uploaded a screenshot of how I’ve laid it out for now

http://users.telenet.be/sickbrain/Afbeelding%204.png

Anyone who could provide me with some insight as I’m quite new to Shake. If you need more info just ask and I’ll try to explain what I want to achieve a bit better.

Tnx in advance,

Andries

I just tried to do a luma key which works like the depth slice keyer but I’m too tired too think. Instead I just did some reordering and use the depth slice key to get the result. So here’s the thing I want to know how to recreate (without the depth slice node):

SetAlpha1 = SetAlpha(0, 1);
ColorSpace2 = ColorSpace(SetAlpha1, "rgb", "yuv", 0.3, 0.59,
0.11);
Reorder2 = Reorder(ColorSpace2, "nnnnr");
Copy2 = Copy(Reorder2, SetAlpha1, 1, "rgba", 0);
DepthSlice1 = DepthSlice(Copy2, 0, 0, 1, 0, 0, 1);

// User Interface settings

SetKey(
"nodeView.ColorSpace2.t", "0",
"nodeView.ColorSpace2.x", "1011.45532",
"nodeView.ColorSpace2.y", "372.1047",
"nodeView.Copy2.t", "0",
"nodeView.Copy2.x", "969.5603",
"nodeView.Copy2.y", "195.832474",
"nodeView.DepthSlice1.t", "0",
"nodeView.DepthSlice1.x", "954.340332",
"nodeView.DepthSlice1.y", "106.6597",
"nodeView.Reorder2.t", "0",
"nodeView.Reorder2.x", "1045.136",
"nodeView.Reorder2.y", "299.764374",
"nodeView.SetAlpha1.t", "0",
"nodeView.SetAlpha1.x", "874",
"nodeView.SetAlpha1.y", "481"
);

I don’t need this kind of key at the moment, I just want to know how it works 🙂
Connect a Ramp or some footage to the SetAlpha node

i have everything, if u work in motreal and interesting the shake4 .
add me in msn:dannifirst@hotmail.com
everything is free. i just want to share my library with my friends.

Hey guys,
We’re currently working on a TV show with CG characters and live action plates. I was wondering how we can make the slap-comp process automated.
for example, When the lighting department is done, we just want to see quick slap comps of the characters on top of the plates (A over B) before compers actually start working on the shots.
Is it possible to write a quick script to grab all the files from folder "A" and put it over all the files from folder "B" and write out a new sequence under folder "C"?
thanks in advance…