View Full Version : how do you put a 2D character in 3D space?

01 January 2010, 12:26 PM
Hi guys! I have a question. I'm a college student who's planning to do a 2D-3D mix up something like Prince of Egypt or Titan A.E. I've recently watched, Prince of Egypt the making, and well, they said they used something called exposure tool. I imagined it like some kind of poly plane, where your moving sketch is, and you can move the plane around a 3D environment, giving the feel of your drawing moving in 3D space. Hope I explained clearly. So, my question is basically what is that program or something similar to it. :) thanks a bunch for answers :D

02 February 2010, 10:36 PM
I did 2d characters in a 3d environment for my thesis film back in school.
I basically rendered out the 2d animated character (+alpha) into an image sequence, and put them on planes in my 3d scene.

The only trick came to the final composite, If I wanted to render the 2d characters separate, I had to render out 2 passes for them: the color pass and an alpha pass. This is because their embeded alpha works for the shader so you do not see the card, but the alpha of the render shows the card, not the character's alpha)

I used a toon shader to render my 3d stuff and it turned out pretty cool.

hope that helps

02 February 2010, 01:49 AM
A "2D" character necessarily must occupy a plane in the 3D space.

Realistically, that plane is going to have to be perpendicular to the camera's POV at all times.

Therefore, you need to arrange things in such a way that you will be able to render two sets of 3D-generated outputs:
Everything which occurs behind the 2D-image plane. Everything which occurs in front of it.

In each of these two cases, as well as your 2D material, you will need to have at your disposal some method of knowing "what is image" and "what is not-image, i.e. transparent." This is known as the alpha channel (where A=1.0 indicates "opaque" and A=0.0 indicates "transparent")

When you have that, the rest is truly a piece-o-cake for compositing. You simply "layer" the three strips of material, back to front, to produce your final image.

02 February 2010, 12:31 AM
Thanks for the reply guys! :D, but it's still well, not kinda clear :blush:

@frizDog- what do you mean by image sequence? I was thinking of having several planes placed like some dominoes each containing one movement of the char.

I'm sorry if I didn't understand:sad:

02 February 2010, 07:05 AM
Grab a piece of paper and go to a nearby cluttered table. Hold the piece of paper vertically over the table, there among the clutter. Now, look at what you've got.

The paper represents the plane where the 2D imagery is going to be. You'll see that there's some clutter in front of the piece of paper, and other clutter :) behind it.

Now, instead of a piece of paper, let's imagine that we've actually got a piece of clear plastic film ... the kind of stuff you used to use with overhead projectors, and a frame of your 2D imagery has been drawn onto it. The plastic is opaque (Alpha = 1.0) anywhere the imagery has been drawn, and transparent (Alpha = 0.0) everywhere else.

This "piece of plastic film" is actually magical because your imagery, as a movie, is "projected onto it." You've done the usual 2D stop-motion animation in the usual way ... it's all right there on that magical piece of plastic in your hand, just like in the Harry Potter movies.

Now, for the intuitive leap ... you're going to combine that with a foreground and with a separate background.

When you set out to do, in the computer, more-or-less what you see on that cluttered table, you're going to do it using three "transparent pieces of plastic":
One contains the 2D imagery, with alpha as described. Another contains only an image of everything that will be behind it, also using alpha as described. A third contains an image of everything that is in front.

With me so far? Hope so. :shrug:

We have three planar "layers" of information, and each of those layers has an additional channel of information, called "alpha," which basically tells us where useful information is and where it isn't. The "alpha channel," therefore, will act as a mask.

We use the 3D software to separately produce the second and third "pieces of plastic," and then we use compositing to "sandwich them all together."

Since the final image is, of course, flat, and because there is a clear indication of what's "in front" and what's "behind," the resulting image appears properly three-dimensional.

Incidentally, many years ago Walt Disney actually figured out a way to do compositing through the use of sheets of plastic photographed in a camera rig that was many feet tall. Although the implementation is different, the principle is the same.

02 February 2010, 09:24 AM
@sundialsvc4 - thanks for taking your time to explain it further. and yes it is getting to me now :). hhhmm ok here's what i've thought of so far, so 3 layers, each individually containing a background, foreground and the 2d drawing.

And now what got into my head as I thought of this is say, like using a linetest say of Flipbook or Flash to animate the hand drawn char. then using Maya to make the whole background and foreground, and then using after effects or premiere to composite them together.

Am I getting there? I hope so, wouldn't want to annoy anyone if I was wrong:sad:

CGTalk Moderation
02 February 2010, 09:24 AM
This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.