bullet time effect of exploding ice rock how ?

[tom schuelke]

Hi sorry i post this here,

but there are so many talented guys here, il do it .

we do an animation of an icerock exploding.

What we plan to do is a bullet time effect also. Scene consists of two parts, a reactor explosion and also particle flow fragments, smaler , flying arround.

my idea for an bullet time effect was to maybe use the motion mixer. as i´ve seen, i need to write out a XAF file of the reactor animated parts. which did work allready.

but no way to write out the particleflow thing.

and at last how do i bring them back to the motion mixer...

any ideas ? we are rather new tho these parts of max..


heeeelllllp pleas,


Tom Shocked

[Da_elf]

ive been able to use a method that stops time. however unfortunately it doesnt slow down then speed up. its just 100% speed - 0% speed - 100% speed. but it can stop anything (even something like aura) while still moving the camera. (sort of) due to my method i cant get motion blur done in max.

[tom schuelke]

Hi Da elf, this is right now the method, we will use if our bullettime effect will not work properly,

but its maybe not what our client wants.


Tom

[studioDIM]

For bullet-time you mean revolving around the exploding rock while it stands still?

In that case, it should prove pretty simple: Create a spline arc (or circle), and a standard camera (say the "master" camera).
Assign to the camera a path controller, pick the spline as path.
Then, shift drag the camera, and create as many copies (shifted by the first amount) as you need.

Those will form a nice are/circle around the object.

Then is just a matter of writing a simple script (or doing it by hand) to render the bullet-time frame from multiple views in sequence.
In a post software you can then blend the frames to create the moblur feel, or animate the cameras moving from the frame before to the frame after the bullet-time one, so to have 3d moblur in the renderer.

It worked well as a previz for a past job, lemme know if it cuts it for you...

re,
Lele

[studioDIM]

Herm, quick fix:

for i=1 to cameras.count by 2 do (
cameras[i].pos.controller.percent = ((100.00 /cameras.count)*(i-1))
)

This to make sure the cameras are spaced evenly along the spline.
If the cameras are free cameras instead of targeted ones, take the "by 2" away.

It seems something's not working properly in the path controller (max 7.5) and they all get packed in the first 25% of path travel, no matter the number of copies...

Lele

[tom schuelke]

Hi Studio Dim.


Thanks alot. This would be my secound wish. what i rather want to do isnt a bullet time effect that stops time totaly, but a slow motion part.

I thought of this feature in the motion mixer with bipedes, and looking for the new features in max 8 it was told, it would do this in the whole programm.

but if i dont find a way i definitly will try this


Tom

[studioDIM]

Hey Tom
If it's just a slowmo, can't you just, as you said, slow down the fliying debris' keys with the motion mixer?

I think that's the quickest way, and it'd allow you to spin one camera at ease...

By now i'm curious to see the result

Lele

[Da_elf]

there was an old script that could slow down things. even particles. but i think that was back in the max4 days

[joconnell]

If you wanted to you could bake out the particle flow motion to keyframes and apply an ease controller to the entire lot or even just use drag spacewarps to gradually slow things down - I used that for a shot of a mirro exploding and it seemed to work fine. The script you're talking about is called maxtrix by the way - bobo wrote it a while back, not sure if there's a max 7 /8 version.

[Morbid Angel]

Tom Im not sure if this will be of any help, but you may read about it and deside
http://www.scriptspot.com/download.asp?ID=890&inf=1

[studioDIM]

doh, i missed the particles out...

Maxtrix does start up with no errors, but it's either me not knowing how to use it, or it doesn't work on PFlow (as it has no motion curves).

http://www.scriptspot.com/bobo/mxs4/maxtrix/

I'd try scripting, so to bake the particles pos and rot out.
But it might prove quite pricky...

Or, you could try render out subframes directly, bearing in mind one of the 4800 ticks per frame would be one of your 25 fps...
A heck of a job by hand but, again, pretty easy with maxscript.
That, of course, if your particles' integration step is set small enough.

Lele

[tom schuelke]

Ahhhh....


Did i ever tell you all , that i love you.. ..


This is realy a great place..

So now i downloaded the scriptfile for max4.. i am not sure it will do the job together with reaktor and pf. reaktor saves the keyed frames on each debris part of the ice rock. but

the particle flow particles are procedural... but ill give it a try now.

well if this doesnt work ? (we are newbies to reaktor and pf rather).. how do i do this, baking out the particle flow motion to keyframes... this would do the job i think, cause this way i think i also could use the motionmixer (what i also did never do allready) and i also could aplay a ease curve in the track view....

thanks again. ill soon post some first test shots..


Tom

[tom schuelke]

Ok here are three possible screenshots. a testanima ill post later.








these are only first ideas of the look, deadline is small.

the middle one shows how the big parts fall appart, the debris of pf is still missing. and the animation doesnt allready rock as i want it.

tom

[tom schuelke]

and a small (the very first) testanimation to get a feeling for the look.
also cam movement is bad, bullettime is still missing and also the pf particles.

the secound film shows another test.

http://www.tom-schuelke.de/forum/ice/06-02-12_ice1_.m1v

http://www.tom-schuelke.de/forum/ice...ice_ra_pf_.avi

[studioDIM]

Woah, nice job there.

All in all i'd suggest you to turn the timeline to frames:ticks, turn "snap to keyframes" off (so to actually see the ticks), and simulate the particles at higher "integration step" (reactor has curves and would work anyways).
Then render per frame in normal conditions, and say, to the integration step in the slowmo conditions.
You shouldn't be needing any scripting with it, either.

Again, nice nice job

Lele