frosteternal

Well, it has been a busy quarter, and it was spent mostly building a scale model - no, not a digital one, a real one - of a sort of military interplanetary outpost. In order to utilize the footage shot of the model, I built a digital landscape (the group responsible for the model landscape utterly failed at building it) in a nameless Other Software package that specializes in landscapes. However, just a pretty picture won't do, so I brought a robot model I built in yet another Other Software package into A:M, rigged it, re-textured it, and now it looks ten times better than the original ever did. The animating & rigging is going to be a bit of a challenge, since the thing is designed to walk or attack with any appendage. I'll post my test shots here over the next few weeks. =) robotTest.mov

0 makes it all go haywire. Your trick with moving the character after the sim is brilliant and effective. I'll be using it later =)

If you are looking to import actual models - which modern extended versions of both photoshop and AE will accept - the best formats would be OBJ or 3DS, which A:M can export nicely. The resulting objects will be polygonal, and may have some surface artifacts due to differences between polygons and patches, but should work fine.

I do not believe version 12 will bake dynamic constraints - but I'm not sure. If you are having problems with dynamic constraints, you might try rendering final with multipass OFF - rather than ON. It might work better (can't remember ver 12) when you "simulate spring systems" it effectively bakes the motion into the action's channels. it just wasn't called "baking"

right-click on the choreography or action and choose "Simulate spring systems" this will "bake" the motion into the action/chor. The "simulate on the fly" option in the constraint is mainly for preview purposes, not final render.

Well, your sense of mood has really come a long way, Kat - the lighting is very strong here. I also love the little character acting details - like the drumming the fingers on "but its meaning is concealed." One thing stand out though - the eyes - while you have shapes and expressions, they always stare straight ahead. Once you have all other animation done, go back and add some eye darting, swiveling etc.. nothing too extreme - and your characters will really jump off the screen! Your tenacity is admirable, I can't wait to see what you are up to next!

Even when I do actions in the Chor, I like to export to a separate action as well so I have a backup. (Besides, you never know when you want just one pose you did back 3 scenes ago) When doing "The Mountain", this approach helped us pre-sim TONS of cloth and saved HUNDREDS of hours that would have gone to re-simulating (we had many, many changes over 3 years) Even some sims were able t be re-timed without glaring defects. As others have pointed out, using actions has re-timing advantages. Now, back to the original post : I never use them for blinks, that is what poses are for. Auto-blinks are as Robcat said, LAME.

This may sound obvious - but does the problem show up in final render, or just in the real time (shaded) mode. By final render, I mean rendered to a file (not the quick render or bounding box render.) It looks suspiciously like a video card quirk. (Oh, and the double image after you hit "apply" is just the stamp floating above the applied version, waiting to be applied again.)

I like this one, the lighting and camera angles are very strong. I really enjoy the "banshee spin" as it moves into capture mode The only thing I might add is have someone from their group of friends run after the banshee in vain, maybe see them shrink away in the dstance from an over-the-shoulder shot of the banshee? Nice job!

Choreography is where to run sims, and then when you like them, I like to (rightclick on chor action) "export as action" and save the sim in a separate action file. (this has saved me time in a pinch before.) The actual simcloth materials are applied in modeling windows, like any group material.

SimCloth is not hard to use, it just takes a little practice, and what Fuchur said about experimenting a little first is helpful - but really, the best way to get the hang of it is to use it on a character. Cloth sliding down primitive shapes and tables are not the same as a real sim. If you wish to understand simcloth (it is fascinating and powerful) pick a character from the cd (with moderately dense geometry) and add clothes to it. Shirt & pants. (Skirts are fine too.) For your settings, set the stretch values low, and the stiffness high, and adjust from there until you get the look you want. SimCloth becomes more predictable the more you use it, and the setting will slowly start to make more sense as you tinker with them. Oh, and save often, just in case your sim takes too long and you want to force quit the program Also, don't have all parts of the cloth self-collide. Around joint, like inside armpits, elbows, etc, you can group these, and in the group property setting there are "plugins:simcloth" options - turn off "self-collide" where you don't need it. For my "Figure Drawing" project (which I work on in between school and work) I do have full-body cloth sims; all characters have fully simulated clothing. Our previous project, "The Mountain", had simulations on all shots of the main character's cloak. In the years this feature has been available, it has been improved tremendously, and is now very robust. Good luck!

The subgroups are key. As tempting as it is to just apply cloth and let the sim run, all cloth sim technology is still very touchy. (I read an interesting list of cloth sim tips from a Pixar cloth-sim lead once - made me feel better...even the pros have intersections and failed simulations.) Sub-groups tell the computer what is truly important, and how that importance is to be dealt with. Essential for good/stable sims. Oh, and the rule of thumb on cloth mesh density - build the mesh to the minimum resolution to display the maximum detail desired. If you need to simulate wrinkles 3 cm wide, then you will need patches no wider than 3 cm. Ask yourself, "does this tell the story, sell the effect, etc?" If the answer is no, then explore other options. "Less is more," is not true - that is a logical fallacy - but less can be more effective.

Cookie-cutter map is just a decal type. Ttransparent parts of the image are rendered transparent. Looking back, now that I see what you were modeling, it wasn't the best solution. Glad you seem to have worked this out!

Slap a cookie-cut map on the surface if you want a hole that you didn't model. Depending on your graphics card, it may display as cut out in real-time too. If you simply want patches "turned off" so you can model another part better, use groups, and "hide" I do that all the time. If you want a permanent solution, delete the points that make the patches valid, or simply and an extra CP here and there with the "y" key (insert) Other solution, if you just need part "out of the way", model the parts you need to see separately, and move them into place & attach later. If you post an example of a screen shot showing what you have/want we could help you more effectively. (Last note, Render-lock is not always an effective preview tool, as it uses the old renderer. Just do a bounded preview with what render settings you intend to use for production.)

Very nice. Perhaps just a slow truck backwards with the camera to heighten the empty hopeless & isolated feeling as she is dying(?)

Fantastic! Facebook and the Forums are pretty much my only social networking sites.

Apply it as a stamp, THEN change the repeat. And use TGA, not JPG. 1. Adjust your view so that the whole floor is visible, nothing else. 2. Right Click on the Decal in the Project Workspace, and choose "Start Positioning" 3. Scale it up to cover your whole floor model, then Rightclick, choose "Apply" 4. RightClick, "Stop Positioning" Now go into the decal instance and adjust the repeating images, etc. Otherwise, applying the file to a group can make the patch images can go every which way and get wonky. That is the technical term.

Nice one, your splines are all flow-y and stuff now =) Now maybe Nancy & I won't harass you about spline flow. Well, sometimes.

Cool beans. I particularly like the bear destroying the butterflies.

You're very welcome, Kat.

Nice page turns, maybe just a bit more "plop" of those snails? Who is she talking to? Herself? I suppose witches, being solitary would be self-talkers... OR....can this scene, as it is mostly explanatory, be condensed? Suggested dialogue - "I need no oracle to tell ME what to do" "The ancient book of shadows has the spell I need" (fairy squeaks in protest) "Yes, this will do quite nicely..." "If I follow the spell precisely - I'll finish with rotting entrails, pinch of newt, a handful of snails!" "Foulest of humours, boiling and hissing form vaporous tendrils, a ghostly apparition!" Short, succinct, to the point. And it seems more a chant than a monologue. This is all coming along so well!

Increase either the damping or the stiffness. Increase bend damping first.

Also, you may have "Brightness" set to "Brighten Facing Hairs." With Hair, you often must adjust the brightness for your purposes. And like Nancy hinted, maybe start with a darker green. Cool tree face model. Looks like you are getting the hang of spline continuity!

If you hold down CTL while dragging it onto the parent bone, it may make a copy like a normal bone would...

Every patch has a "front" and a "back." Normals are vectors that emanate from the front of the patch, pointing from the center outwards. They are used by the software for shading, simulations, etc. If you turn on normals (SHIFT+1) you will see them represented by pointy lines coming from the patch center. (They are usually visible only ion modeling modes.) In order for smooth, predictable shading, all normals in a surface should point the same direction. If you right click in the modeling window, you will see a "Refind normals." It can help to run this tools after you have finished some major modeling overhaul. (Twisting points around, stitching, copy/flip, etc.) If you right click on a group, you will see a "Flip Normals" option. These tools help you make your models less ambiguous for the computer. (Just because you know which side of the patch is the outside doesn't mean the computer does.) Normals are necessary in computer graphics because surfaces are built of infinitely thin shells. A patch has no thickness, you are either on the front of it, or the back. (For definitions of terms, this page is invaluable, (if a bit dated) http://www.pixelburg.com/am_glossary/#00index )