Basic crowd sim possibility

Started by Kadri, August 06, 2015, 09:08:02 PM

Previous topic - Next topic

Kadri

#60

The sticks were for uniting the gears obviously (that only so so works).
But just after the render finished i found out that i could have made animated ropes (kinda) too.


[attachimg=1]


Dune

Oh my Gob, I wish I could do something. You guys are incredible! But at the moment I am struck down by severe lumbago, and just manage some typin once in while. Keep at it!! By the way, for moving trees in different directions with the T&S only one pop would suffice, I'd say, with several masked T&S's in different directions. Even windgusts traveling acrsoos land can be simulated that way, I'm sure (lots of time to lie down thinking  :( )

fleetwood

Cool stuff. Still laughing from the little cars gif (and somehow wishing for tiny clowns to climb out of each one).

Kadri

Quote from: Dune on August 15, 2015, 06:42:38 AM
...But at the moment I am struck down by severe lumbago, and just manage some typin once in while...

Get well soon Ulco.
Since nearly 1 week i am using my left hand on the computer. The pain in my right hand is better now.
Now my left hand is beginning to hurt :D

We are getting old  ;D

Dune


Hannes

Thank god none of my hands or anything else hurts. Get well soon, my old friends! ;)

So I did another test. I created five populations: the body and each wheel as a seperate population. The body's compute terrain has a gradient patch size of 20 and the wheels have a different compute terrain with a patch size of 0.1.
The wheels' rotation is animated (at more or less the correct speed). Unfortunately sometimes the wheels seem to rotate backwards or stand still. But this phenomenon accurs in real life too due to shutter speed or whatever it is. Anyway. it looks a bit weird.
I wanted each wheel to react differently to the terrain.


Kadri

Quote from: Hannes on August 16, 2015, 05:25:18 AM
...
I wanted each wheel to react differently to the terrain.

Would one population of 4 wheels look too much different? Curious.

Hannes

Thanks Kadri. Yes, I think so. If the terrain is very bumpy the left front wheel might be higher than the right one for example.

Just found out that I connected the wrong compute terrain to the planet. Doing some tests with a seperate compute terrain for the body with a patch size of 100 and "Smooth surface" checked, which is NOT connected to the planet. So the body reacts to an identical but smoother terrain which is not visible. Only the normal terrain, the wheels react to is visible. So I'm hoping I'll get the illusion of a spring loaded (is this the right word?) car.

Kadri

Quote from: Hannes on August 16, 2015, 05:54:37 AM
...If the terrain is very bumpy the left front wheel might be higher than the right one for example.
...

But even if separate when the values are the same on all the 4 wheel populations they would react in the same way on the same terrain.
If you ask me you are making it harder then needed.

Quote from: Hannes on August 16, 2015, 05:54:37 AM
...
...Doing some tests with a seperate compute terrain for the body with a patch size of 100 and "Smooth surface" checked, which is NOT connected to the planet. So the body reacts to an identical but smoother terrain which is not visible. Only the normal terrain, the wheels react to is visible. So I'm hoping I'll get the illusion of a spring loaded (is this the right word?) car.

I made a different test with a hidden surface similar (I won't upload it.It is not usable) to yours (from a surface approach).
I think it isn't related for using a hidden-or bypassed?- surface but sometimes some objects in the population are disappearing.
By using less displacement or smaller displacement spike limit they came back mostly.
I tried it with smaller and bigger compute normal settings and smooth normal and much higher Planet displacement tolerance etc.but they weren't much useful then the above settings. It looks related when over higher displacement ground.
Have you encountered anything similar Hannes?

Hannes

Quote from: Kadri on August 16, 2015, 06:19:48 AM

But even if separate when the values are the same on all the 4 wheel populations they would react in the same way on the same terrain.
If you ask me you are making it harder then needed.

I made a different test with a hidden surface similar (I won't upload it.It is not usable) to yours (from a surface approach).
I think it isn't related for using a hidden-or bypassed?- surface but sometimes some objects in the population are disappearing.
By using less displacement or smaller displacement spike limit they came back mostly.
I tried it with smaller and bigger compute normal settings and smooth normal and much higher Planet displacement tolerance etc.but they weren't much useful then the above settings. It looks related when over higher displacement ground.
Have you encountered anything similar Hannes?

No, it really makes a difference. You'll see it in the animation I'm rendering at the moment. Each wheel reacts to the position it is in the terrain, which might be quite different for each wheel.

There is no hidden surface in my setup. I just added an additional compute terrain with "Smooth terrain" checked for the cars' bodies, which is not connected to the planet. See attached image.
Luckily the cars' bodies just react smoother than the wheels which was my intention, without having to create another (invisible) terrain.

Kadri


I looked at the scene ones again. I think they don't disappear.
They just sunk more in the ground.Much more then they should or it looks from look of the ground displacement.
Hmm...not sure what to do but just in case you see anything similar be prepared :)

Kadri


Exactly the same setup Hannes thus i said -bypassed-.
There is no second Planet but still a different one...kinda :)

Still not sure about the different wheels (if not different values are used of course). Curious of your test.

Hannes

Thanks for the warning! At the moment nothing disappears. The wheels individually react more or less (!) correctly to the visible terrain, and the body moves way smoother. Still rendering.

Here is another screenshot of my setup. The last one didn't show everything.

Kadri


Slightly different but same method.

[attachimg=1]