robcat2075

Although we rarely do it anymore, in A:M you can set up a chain of bones, pose them as if they were IK and they will interpolate as if they have been keyed in FK. So I wonder if this Pixar thing is a similar result.

I think they are trying to get something other than the straight-line inbetween of two poses, but you'd have to digest a lot of jargon to know. "Evaluation time"?

the language in the patent is pretty dense, I'm not sure what it is they are patenting that is new... http://www.freepatentsonline.com/8358311.html

The realtime view has an adjustable subdivision... PageUp: smoother, slower PageDown: flatter, faster

Markw is likely right. If you have created a lathe outline with some CP on the center (or worse... crossing over the center), nudge it with the cursor key until it is just barely to the side. It doesn't have to be much. Welcome to A:M!

If you like the topology of a polygon model, I'd recommend importing it into a model (instead of as a Prop) and deleting the splines you don't want and leaving just the ones you do want. I have used that subtraction method on polygon models with Quads and that works well. A little bit of spline breaking and reconnecting is needed to make good A:M patches in some places but I think this is faster for this case than remaking the mesh from scratch wit Snap-to-Surface.

Ouch. He must have landed in something really sticky.

In the Project Workspace, on the Images folder> Import> Image Sequence I just checked that in v10.5 Welcome to A:M! you are in the right place to ask questions.

OK, I read this a bit closer now. Force Keyframe, by default, will only force keyframes on things that already have a keyframe somewhere in the timeline. Holding down the SHIFT key while using Force Keyframe brings up an option window that can key EVERYTHING. This is risky to do. Have you seen my "keyframing basics" videos. I think one of them , at least, covers that. In the Screencams link in my signature.

First, make sure you don't have the A button OFF.

I Haven't tried a Darksim but if you are in a pinch you might temporarily edit a material to make its bump pattern appear as a grayscale color, then bake that and swap the map that results from that with the regular bump map.

Yeah CPU speed is probably still the biggest indicator. We're comparing apples to prunes here. Mmmm... prunes... I need lunch.

That looks threatening! I suppose if you could try a soft edge on the clouds But I like solid clouds you have.

I just tried it with a decal and i got visible stuff on the baked bump map. Something else is wrong.

you did have some bump maps applied before you baked, right? Or a bump material?

Your i7 CPU has the "SSE4" extensions that mine does not and that probably is a bigger advantage than having 64-bit.

I just tried it and I did indeed get a shadow image in the alpha buffer. Possibly your After Effects was defaulting to "ignore" the alpha channel. A simple test is to load the image into Photoshop and see if the alpha channel has anything in it.

I think the relevant values are from .1 to 1.0 Try those on a simple case and see what you get. 1.0 wont' necessarily give the same size bitmap with different models.

On to three!

You need a booth babe to sit at your table and sell DVDs.

I have an Intel Q6600 which is a generation or so earlier than your i7 CPU It has four cores at 2.4 GHz. NetRender takes about 5:16 per frame to do the benchmark scene.

TGA will also work. The shadow info will be in the alpha buffer. That will reveal the portion of the otherwise black RGB image to make a shadow.

"Add-lock" is obsolete. The Current "Add" mode does what Add-lock used to... allow you to create more than one CP without having to repeatedly press the Add mode button. The Add-lock button can still be found inthe Customize window but i find it non-functional.

I would suggest making the clouds opaque so the arrangement of the elements is more clear.

The hyper threading is the culprit. The two virtual cores get you some extra overall performance but not as much as two real cores would and they have the drawback of occasionally getting in the way of the real cores, slowing everything down. In most consumer apps that would never be noticed, but A:M saws away continuously for many minutes at a time which is different than most apps. You might try a test with hyperthreading disabled ( and two nodes running) and see if the result is much different. On my four-core, no hyperthreading machine the render nodes are almost as fast as A:M. They do have some extra overhead in that they need to completely reload all assets for every frame (although that is not a significant issue inthe benchmark scene)