robcat2075

As far as I am concerned A:M's depth of field has never been worth the render hit. You need so many passes and even then it never looks right. I agree A:M's DOF is quite time consuming. If you are using a compositing app like Aftereffects, creating DOF via a depth map and appropriate blur filter will be much more flexible and quicker.

Ding, ding, ding! That's exactly right. Or at least it's my distilled version of him. Just wanted to make sure I wasn't fooling myself about whether it looked like anything.

Render the first frame of this PRJ with "use settings from > The Camera". This renders a black frame with the vase's shadow in the alpha buffer. ShadowBufferTest.prj EDIT: This sample PRJ doesn't give the expected result in v19

This is a procedural material I made today. What does it look like?

Black caterpillars aside, that all looks uniformly blurred rather than blurred depending on distance. I have done a test somewhere here showing both DOFs working in their own way, but I don't know what your settings are. Is the caterpillar a one frame thing? on every frame? Same spot in the frame every time? On the same spot in a model every the time? Moving on a spline in a model? If I saw it in motion more would be known, possibly.

I'll guess the "Dan" was Dan Tiedy. Current whereabouts unknown. If you can make something that exports a "standard" that can be used with programs like Vue or Particle Illusion (or terragen) that would be great.

is that multipass or regular DOF blur?

Actually, I see I misunderstood Malo's reference to "Resurface". I was thinking of another program discussed elsewhere, probably not even called "Resurface" that attempted to make splines over polygon model. So basically, I initially misunderstood everything in Malo's first post.

Unfortunately, no. You have to judge it manually. When the prop is in shaded mode you can't see the splines over it at all, so it's not a perfect solution, but possibly useful.

Here... Is this what you are thinking... The head is an OBJ imported as a prop and placed in the chor. Model 1 started as just a spline to make a model that could also be placed in the chor. In the chor I selected Model 1 and then hit the Modeling Mode button to enable editing Model 1 with new splines and drew the new spline that is on the profile of the head. Doesn't work well in shaded mode, but this is what you were thinking except it's in the chor window, right?

Wonderful looking shot, Eric! How could we get that to work in animation? Probably the best live-action film Disney ever did. You oughta rent it.

I've re read your post a bit... I guess what you are wondering is if the "prop" could be a solid, uneditable shape in the modeling window, that you could model around with splines.... hmmm... that sounds plausible... but you can model splines in the chor already (hit the Thom button). Put a prop in the chor and model splines around it, that won't work?

The import of polygon models has been discussed many, many, many times here. Here's one.... http://www.hash.com/forums/index.php?s=&am...st&p=326235 Short answer... if it could be done, it would have been done long ago. You mean automatically? It wouldn't be any easier for a "prop" than a model. If ReSurface is doing an acceptable job, then just use it create mesh suitable for import. Short answer, I doubt it. There are too many other higher priorities that would need to be pushed aside for this one enormously difficult problem. Great looking models! You don't need polygon import!

Can you post a specific test case? Here's my 10 minute test case and it works... ConstraintToActionObject.prj Vase is a model. Action 1 is an action for Vase that uses Cylinder as an action object and moves it back an forth.; In the chor, Action 1 has been dropped onto Vase Sphere begins not being constrained to the action object, then is constrained to it starting at frame 20. From frame 20 onward Sphere moves with Cylinder.

more detail needed. in chor? in action? to what?

That looks great! Are his arms too short? Maybe it's just because he's tall?

in 8-bit color 128 is neutral grey(which means no displacement in a displacement map)so that leaves only 127 levels of displacement up or down. For small, sharp features like these windows it may not matter since the slope is greater over one pixel than the minimum slope that 1 level represents. Here's comparison of my dragon spike test, The first one uses a targa depth map, the second uses an EXR depth map. They are both the same resolution, but the greater precision of EXR allows every pixel to have an exact gray value rather than being rounded off to one of 256 levels. You can see the banding most on the body surface, where the neutral gray wasn't all exactly the same identical value. EXR is able to transition those small changes, TGA can't. There are other problems in both renders. In general displacement maps are interesting but not perfected yet.

For now they will just stretch as the skin stretches. Down the road, I might experiment with scaling their intensity as the knee bends/unbends.

back o' the knee wrinkles

Whoops... I made a mistake on the hand part... constrain the hand controllers to the box not a hand geometry bone (because the hand controller is what tells the hand geometry bone where to be.) This will be the IK hand controller and your arms will need to be set to IK of course. You'll need a "translate to" and an "orient like" constraint for each hand. The box does get constrained to a geometry chest bone. And this will be a "translate to" and an "orient like" constraint. What these bones are called depends on your rig. In some rigs the IK arm target only positions the wrist but doesn't aim the hand bone, so you'd translate constrain the IK arm target and orient constrain the hand bone

For that specific scenario... While picking up the box I'd constrain the hands to the box and then animate the box as if the hands were moving it. (Of course, you may need to unconstrain and reposition the hands a few times before they get to their final gripping position on the box) While carrying the box, leave the hands constrained to the box, and constrain the box to some bone in the chest, like a spine or torso bone. This eliminates having to animate the box moving through space ahead of the body. You can still vary the box's position relative to the body to make it seem not rigidly glued to the chest. In all cases make sure you are constraining with actual geometry bones and not merely the controller that influences those geometry bones. (correction below)

Good looking models. I like the way the penguin design is carried over into the plane. What does a penguin do with cruise missles?

Make a simple PRJ with a very simple model embedded in a chor that does this and post it. I can't imagine.

Saving a file to a drive does not erase other files. Something else is wrong. Rendering to targas is the way 90% of rendering is done.

I didn't really save them so i just ran them thru again. Mine aren't perfect either, you'll probably still have to do some repair. And your stuck with whatever I chose to for the parameters. But here they are. ToyotaLogos.mdl