Question about MA....

[minglw]

I've seen MA/MA2/MAT2 in action for a long time and I think the screen saver is beautiful and realistic.

One day, I was showing a friend how nice and relastic MA was...
and the friend said "Yes, it's indeed it's very nice and real. Wait, why do I see fish swim off the screen ? If the monitor represents the fish tank, then the fish should be visible (at least swim off the screen to the sides) at all times."

So, I guess my question would be: is this "fish swim off the sides of the screen" behavior normal ? intentional? mis-calculation?

[Pushwall]

Originally posted by minglw:

One day, I was showing a friend how nice and relastic MA was...
and the friend said "Yes, it's indeed it's very nice and real. Wait, why do I see fish swim off the screen ? If the monitor represents the fish tank, then the fish should be visible (at least swim off the screen to the sides) at all times."

So, I guess my question would be: is this "fish swim off the sides of the screen" behavior normal ? intentional? mis-calculation?  

It's normal, intentional, and not a mis-calculation. It's a view of the fish and not a tank. Who wants to look at an ugly tank?

[Jim Sachs]

The tank is slightly bigger than the screen for two reasons:

1. If more than one fish were to crowd against the edge of the tank, there would be nowhere for the fish to go, greatly increasing the chance of a dreaded "pass-through". The fish can go off the edge, turn around, and re-emerge when the space is clear. The fish also can use this area to pass from the front of the tank to the back and vice-versa.

2. If I allowed the edges of the tanks to show, I would have to put in reflections of the fish as they approach it. This would greatly increase the complexity of the program, and reduce the frame rate.

[minglw]

Originally posted by Jim Sachs:

The tank is slightly bigger than the screen for two reasons:

1. If more than one fish were to crowd against the edge of the tank, there would be nowhere for the fish to go, greatly increasing the chance of a dreaded "pass-through". The fish can go off the edge, turn around, and re-emerge when the space is clear. The fish also can use this area to pass from the front of the tank to the back and vice-versa.

2. If I allowed the edges of the tanks to show, I would have to put in reflections of the fish as they approach it. This would greatly increase the complexity of the program, and reduce the frame rate.  

Jim, thanks for the explanation!

For 1, is it true that when proper edge-detection/collision is implemented, this "pass-thru" problem would be solved ?

I can understand all the extra work (and complexity) of implementing reflections.

I guess it's my problem of seeing the view as a "fish tank". Not sure where I get this idea, but that's how I see it since day 1.
It's correct if I see it as a fixed-view of an underwater scene with some fish swimming around. ;-)

[Jim Sachs]

Yes, if I can come up with a better collision-avoidance scheme, the fish would be able to get closer together. This will be necessary for schooling behavior, chasing, nibbling of corals, etc.

When I first designed the Aquarium, there were no LCD displays, so I sized the tank to be roughly correct for a large CRT monitor. The areas where the fish can go off the sides of the screen are only about as wide as the bezel of the monitor, so when the scene is viewed with 3D shutter glasses, the effect was extremely convincing - the glass of the monitor was the front glass of the tank, and the water extended in about 14 inches. Unfortunately, stereoscopic glasses are nearly a dead issue with everyone going to LCD screens (which can't come close to the 100-120 fields-per-second necessary for a smooth stereo effect).