There were many things to learn from this projection exercise. First off, Ben's camera/projector system for calculating object and camera positions in physical and digital space worked excellent. Secondly, Nikki and I were reminded of the importance of checking and double checking projector settings. When we began working on the 2d portion, we neglected to check our keystone values on the Optoma. We would have noticed earlier if we were projecting in 3d when our camera's didn't align, but since we were drawing live in photoshop we unknowingly adjusted for the keystoned projection. When I got to projecting 3D, we noticed the problem so it was necessary to move the duck and sphere to match the projection, so we decided to first wrap up 2D before moving on so that we wouldn't have to move objects back and forth.
The first thing I did before beginning building objects in maya was figuring out the their physical dimensions. It was an easy process. The duck was built from a previous 3D model I have made so it was just a matter of scaling the model to match. The sphere we wrapped with a string to find its circumference and did the easy math. Since the objects are organic, there was no easy way to measure their position in physical compared to its surroundings, (back wall). If it were a cube we could just measure from a corner and match it exactly in maya. I ended up attempting to find a center point on both the sphere and the duck and making an estimate of there positions. With Bens new system of calculating camera positions/settings, I knew that our 3D camera was correct so it was just a matter of projecting the 3D image and nudging the physical objects into place. Once everything aligned I could place tape on the table to help realign the objects if we had to move them.
After everything was aligned it was just a matter of creating the animations in maya to explore lighting and geometry deformation. I stared with lighting. I created a mallard texture to project onto the duck and a basketball for the sphere. I created a light pointing at the objects from the right rear. I animated it to move clockwise around the objects so that we could see how the shadows looked in the physical space. The shadows translated well and without deformations that relied on perspective, it was viewable from a wide range of angles.
For the next part of the exercise I extruded a series of squares from both the duck and sphere. Initially I animated random squares extending to random distances. It seemed like a good idea, but when watching the animation play as projection it became distracting and too difficult to analyze. So I reworked the maya file and made everything much more uniform. The squares were all extruded 'every other' and I made sure to extrude them together. I also slowed the animation down so we could see how the extrusions worked at many distance. I began with pulling them inwards and then slowly extruding them outwards. As expected, the shorter the distance, whether pulling in or pushing out, the wider viewing angle to view to illusion properly. As the squares traveled further outward towards the viewer, the illusion of depth was lost unless you stood very close to the projection source. The "intrusion" of squares were much more effective from a range of viewing angles when pulled in to greater depths.
We also learned a very valuable lesson about projection surfaces. The sphere was made of foam, and something about it's pouruss and reflective material dulled all of the projections. Also, purely white materials tend to over saturate the projections as well so you have to either a.) adjust the animations color balance or b.) paint the physical objects a slightly darker shade of gray. White semi-glossy posterboard was used for the floor and back walls. The surfaces caused a lot of light bounce which washed out the lower parts of the projections even when projecting black.
The first thing I did before beginning building objects in maya was figuring out the their physical dimensions. It was an easy process. The duck was built from a previous 3D model I have made so it was just a matter of scaling the model to match. The sphere we wrapped with a string to find its circumference and did the easy math. Since the objects are organic, there was no easy way to measure their position in physical compared to its surroundings, (back wall). If it were a cube we could just measure from a corner and match it exactly in maya. I ended up attempting to find a center point on both the sphere and the duck and making an estimate of there positions. With Bens new system of calculating camera positions/settings, I knew that our 3D camera was correct so it was just a matter of projecting the 3D image and nudging the physical objects into place. Once everything aligned I could place tape on the table to help realign the objects if we had to move them.
After everything was aligned it was just a matter of creating the animations in maya to explore lighting and geometry deformation. I stared with lighting. I created a mallard texture to project onto the duck and a basketball for the sphere. I created a light pointing at the objects from the right rear. I animated it to move clockwise around the objects so that we could see how the shadows looked in the physical space. The shadows translated well and without deformations that relied on perspective, it was viewable from a wide range of angles.
For the next part of the exercise I extruded a series of squares from both the duck and sphere. Initially I animated random squares extending to random distances. It seemed like a good idea, but when watching the animation play as projection it became distracting and too difficult to analyze. So I reworked the maya file and made everything much more uniform. The squares were all extruded 'every other' and I made sure to extrude them together. I also slowed the animation down so we could see how the extrusions worked at many distance. I began with pulling them inwards and then slowly extruding them outwards. As expected, the shorter the distance, whether pulling in or pushing out, the wider viewing angle to view to illusion properly. As the squares traveled further outward towards the viewer, the illusion of depth was lost unless you stood very close to the projection source. The "intrusion" of squares were much more effective from a range of viewing angles when pulled in to greater depths.
We also learned a very valuable lesson about projection surfaces. The sphere was made of foam, and something about it's pouruss and reflective material dulled all of the projections. Also, purely white materials tend to over saturate the projections as well so you have to either a.) adjust the animations color balance or b.) paint the physical objects a slightly darker shade of gray. White semi-glossy posterboard was used for the floor and back walls. The surfaces caused a lot of light bounce which washed out the lower parts of the projections even when projecting black.
Taken from extreme left of projection source.
Same as above but with more extreme extrusion. Notice how quickly the illusion is lost.
Close up and to the left....
Table height and extreme right.
Straight on and short deformation
Straight on with lengthier deformation. Most of the illusion stays intact from this view.
No comments:
Post a Comment