sprockets Shelton's new Char: Hans It's just donuts by ItsJustMe 3D Printing Free model: USS Midnight Rodger Reynolds' 1950s Street Car Madfox's Pink Floyd Video Tinkering Gnome's Elephant
sprockets
Recent Posts | Unread Content
Jump to content
Hash, Inc. - Animation:Master

williamgaylord

*A:M User*
  • Posts

    907
  • Joined

  • Last visited

Posts posted by williamgaylord

  1. Wonderful work William! And thanks for being so generous as to share it with all of us!

     

    More constructive criticism (understandably far more easily said than done of course....). I think slowing him down, giving him more of a swinging gait would really help the feel of the animation. Also a very low view looking up from the base of one of the pillars (with the pillar base close up in the periphery) as the balrog walks by would also help. Also reducing the ambient light to a very low level, so the Balrog is essentially the only light source would give it more of an ominous feel, too. (Can the spriticles cast light onto other objects? Or do you need to simulate this with another source of light?...this also may be much easier said than done.) This is offered from someone who has greatly admired your work and has drawn much inspiration from it all! And one who has also greatly benefited from your helpful postings on the Forum.

     

    Here's a combination that might work well in the scary vein...put Collin Freeman's boy into the picture with the Balrog! That could make for a scary, suspensful nightmare scene! Would add a sense of scale, too.

     

    On a tangent, it might actually be a fun challenge to make a giant Thom animation don't you think?

     

    Sincerely,

    Bill Gaylord

  2. Outstanding artwork. Are you doing the expressions by "hand" with the modeling tools? If so, wait 'til you learn how to rig the face for animation! Then you'll really begin to see where AM excels. Outstanding work so far. The ears are some of the best formed I've ever seen. (I notice ear details because I build my own dummy head binaural microphones--the ears are a very critical part.)

     

    Bill Gaylord

  3. Really look forward to seeing this story develop. These capsules are "easy to swallow". Amazing amount of personality in such a simple design. Love it!

     

    I can imagine these little guys and gals in all kinds of situations...

     

    ..."Can you hear me, Major Tom?"...(space capsule)...

     

    ...How many capsules can you fit in a phone booth?...

     

    ...Dr. Who?...(time capsule)...

     

    ...the rock band Kiss...encapsulated...

     

    and the list goes on, ad nauseum... :P ...my apologies...

     

    At any rate, I love these little guys!

     

    Bill Gaylord

  4. That was great! Inspiring even! No need to be shy about it. Thanks for sharing it with us. B)

     

    Reminds me that I need to work on logos for two of my little company's registered trade-marks: Art-Techo and Arty-Joke Productions.

     

    Bill Gaylord

  5. This is somewhat of a trial and error process. I'm not happy with the density of branches near the trunk. I'm going to reduce the number of main branches and add more of the short secondary branches. This should give me a much better looking tree form.

     

    I'm also not happy with the uniformity of the arrangement. Doesn't have the aesthetic appeal I want. I'm going to arrange the branches less uniformly to give it a more natural look. More of a bunching of branches and leaves here and there, with some more space between "bunches".

     

    One reason this project is taking so long is the number of "rehearsals" it's taking to optimize the look vs. performance. The number of things to tweak each time is pretty huge. Definitely not an efficient workflow sort of thing at this point. That would take a software utility to build it for you, with some nice features to allow you to tweak it efficiently.

     

    I think I'm really pushing AM's limits in terms of the sheer number of constraints and relationships involved. Some of the branches are loosing their texturing during the camera rendering. And this is only one tree... May have to send another sample to Hash to figure out what's happening and fix it.

     

    Bill Gaylord

  6. This is looking really good! It will be interesting to see this thing moves once it's rigged. Will you maybe add something extra to the webbing to get it to billow when he/she swims? (We need a new smiley with round eyes and a jaw drop! These eyes: :ph34r: ; and this mouth, only more exaggerated: :o .)

     

    Bill Gaylord

  7. Here is a complete tree. Need to tweak the branch positions near the crown, but this should give you an idea of what to expect. Leaf coverage looks good. Leaf clusters are a bit too similar, but I don't think I'll go back and tweak the several hundred clusters again. More realistic lighting with substantial diffuse lighting should bring out the leaves that are farther in, giving it a more natural look. This lighting is not very good. A bit too dark a rendering, too.

     

    FullTree.jpg

     

    I may try rendering an animation. Stay tuned for more exciting developments...

     

    B)

     

    Bill Gaylord

  8. I'd like to introduce the human cast of this production: Meet our tree planters, Missy and her brother Kenny. These are graciously being provided by Frank Hulsey, AKA Iham Wrong. He is refining these character models while I focus on the trees and the set. I really look forward to animating these two!

     

    Kenny-Missy.jpg

     

    Bill Gaylord

  9. Not sure I'd want to swallow one of these. Are these suicide capsules? A lot of character in such a simple design. Look forward to seeing these guys develop.

     

    Bill Gaylord

  10. I'd like to edit some percentage pose sliders. However, I can't figure out how to get a view of curves to edit, so I can select the individual keys and adjust the curve behavior at each key. When I set up a percentage pose slider, I don't move the frame position in the time window (I only adjust the attribute of interest at various pose slider percentage points). How do I get a curve view in terms of the pose percentage? Is there such a thing? If not, it would be soooooo handy to have it added!

     

    Thanks,

    Bill Gaylord

  11. I've decided I'd like to animate two children--a little boy and little girl--to plant the trees in this animation. This is pro bono work for a non-profit organization, so there is no money involved. Any decent royalty free models out there I might at least start from? I need to minimize the time I spend assembling them, so I don't want to build them from scratch. Blue jeans and shirt for both of them is fine, to minimize the need for dynamics (a dress would be somewhat problematic, relatively speaking). Some closeups involved, so nicely crafted faces would be appreciated.

     

    Thanks,

    Bill Gaylord

  12. Truly inspiring work, William. Can hardly wait to see it in action! Swinosaurus vs. the Balrog...or maybe feature it in one of the "Marshmallow" shorts we talked of long ago.

     

    Which Wacom tablet do you use. I'm thinking of upgrading to an Intuos2 tablet and adding an "airbush" stylus to my toolkit. I've been using an ArtZII 12"x12" tablet with my workstation and a small Graphire 2 with my laptop (which I love for modeling work in AM).

     

    Thanks for sharing so much with the AM community!

     

    Bill Gaylord

  13. Here's another lighting test, which was also a render speed test as well. Each brownstone apartment has 13249 patches, so I had my doubts whether or not even my new whiz-bang computer could handle it. The long shot helps, but it did just fine! There is hope that I'll finish this project! Later I will alter the brownstone model so that I can easily vary the window treatments for variety.

     

    LightTest05.jpg

     

    The longest shot will take in all nine apartments. I'd like to animate a small boy and a small girl planting trees along this row, with the planted trees growing up behind them as they work their way down the street.

     

    Yes, there is a lot more detail in the models than will show up in this long shot, but there will also be other angles, and plenty of close-ups. The main purpose is for me to learn as much as I can in general and in some cases push the limits of things that can be done with AM.

     

    Bill Gaylord

  14. I've thought of a way around the patch count problem. I'll just do one tree for every two apartments. Then I can render each pair of apartments separately. This results in animated "tiles" I can stitch together into a single very wide shot.

     

    It still would look good from a landscaping standpoint and has the benefit of fewer trees to animate! I could probably use the same tree and just rotate and tweak it a bit. I can add relatively stationary characters, change the widow dressings, and add different props to make each tile look more unique. Wow! I may actually get this project done!

     

    I could even add conventional characters if I keep them inside each tile, or composite them later, being careful to keep them outside of the tree shadows (or simulate the shadows--a good bit more work).

     

    Thanks for the conversations, folks! It really stimulates some good ideas.

     

    Bill Gaylord

  15. Thanks, folks, for the suggestions and encouragement!

     

    The SkyCast lighting to my eye is too flat. Could you add a spot light to give some better contrast, and let the SkyCast rig provide overall general lighting?
    I agree. It's more like an overcast day.  I do have a key light I can turn up to fix this.  (This was a quick test without much tweaking.) I also need to make sure the leave's reflect specular light better, perhaps.

     

    The way the leaves scale up is too obvious to those of us following along... I'd ask my wife to look at it and see if it bothers her the way it bothers me (she not having read the whole thread), but she's out right now. To be sure, actually controlling angle and size would be much more challenging than scaling a bone, but that's still what's bugging me about the animation.

     

    There are just too many leaves on a tree to do much more than scaling, but I certainly could stager the timing so they don't look like they are growing in unison. The angle is more of a challenge. I leave "store roll" on so the leaf clusters stay facing upward without me having to adjust ever one of them when I rotate the whole branch. It's definitely a compromise.

     

    Your telescopic idea sounds like it would work pretty well, but you would still have to scale all the rings, even though you are not stretching the segments. As the tree gets taller the trunk gets fatter. It's starting to sound more complicated to rig and to setup sliders. I wouldn't have noticed the bark if you didn't say anything. All in all it's looking great. Keep it up.
    Actually my rigging is that complicated.  But the final animation controls are quite simple.  Every spline ring has a separate control slide to scale it.  I adjust the scale of each ring and the "ease" of each ring along the guide spline in an action so I have as much control as possible as I animate the growth of each basic "sub-branch".  This boils the controls down to a single slider to control the "growth".  The key is building a hierarchy of relationships.  I then animate the whole branch from these "sub-branches" controls in yet another action window and animate them to sprout and grow in the right sequence.  This leaves me with one slider to work with which allows me to control the timing of the branche's growth, even stretching out some parts of the sequence in time.  Each branch is then used as a separate "character".  (I tried to import the branches into an action to assemble a tree, but this did not work. There were a tremendous number of conflicts.) I assemble the tree in a choreography using the set of branches and the trunk as sort of an erector set for trees.  Still labor intensive, but well within the possible.  I'm hoping we can take the basic assembly technique and automate it with a plug-in or utility.  That would allow a whole tree or other plant to be assembled as a single model and reduce the controls to a small number of controls.  Any volunteers?

     

    ]As an aside, I am currently working on a building that I need for closeups and I made the walls one extrusion thick. This resulted in an insane number of unneeded internal patches.

     

    My brownstone apartment is just a front, not a box, so it was easier to avoid internal patches. You do have to be careful about the number and placement of control points in the figure you plan to extrude. I would create a box by hand with open ends that already avoids inside patches and then extend that. The grid for my building has a patch for each window so I can make those patches invisible. Probably could have done the same thing more efficiently with decals and alpha channels. My brownstone has a fairly high patch count, mainly because of all the fine details like the moldings and (especially) the fence.

     

    I may have missed something, but you said you hit a problem with patch count and reduced the patch count of the tree by approx 1000. However, did you not say you were going to have 9 trees? If so you are going to hit the same problem on tree number two and then be stuck.

     

    Yep. May have to recruit a render farm.

     

    Again, thanks for the comments and encouragement!

     

    Bill Gaylord

  16. It might make the animation look a little better, but it would be a good bit more complicated, or a lot less accurate. (I'm doing this for some tree experts.) Aside from the fact that the actual growth would be seasonal, this model simulates the way a tree actually grows. A tree grows at the tips of branches and at buds. The shape of the previous growth stays the same except for thickening as new layers (what makes the rings of the cross-section) are added each season. The locations of newly sprouted branches and leaves stay fixed rather than migrating along as the branch lengthens. So my model (being a magical tree that "catches up" to the current growth season, doesn't sprout any leaves until the branches reach the end point of "last season's" growth. Then it sprouts the new leaves and branch tips. If the leaves started growing at the half-way point, they would have to travel as the branch lengthens, which is more like scaling a branch than the way a branch actually grows.

     

    My model isn't entirely accurate in the way the branch growth happens. I don't grow it from the end, as you can see by following the bark texture. A real branch would only grow at the tips, with the bark texture behind the growth staying in the same position and only becoming thicker and more coarse. To make the model manageable, I stretch tube forms along a spline path to simulate growth from the ends, but a close inspection of bark texture gives the "trick" away....

     

    Actually, describing it to you has made be think of a refinement to the model that might actually work very well. Instead of keeping the spline ring bones evenly spaced as the tube stretches along the spline, I could keep the rings near the growing end "collapsed" together and only expand one segment at a time--like "telescoping". Then the stretching would only be visible near the growing end and the texture along the rest of the branch would stay put like it should! Cool!....and @#$%, I'll need to modify the branch growth actions again! Might be worth it though.

     

    Anyway, thanks for the suggestion. I could probably stagger the sprouting of leaves a bit so they don't look quite so much like they are growing in unison.

     

    How do you like the SkyCast lighting?

     

    Thanks!

    Bill Gaylord

  17. Here's the movie! To save on rendering time I cut the duration in half compared to what I would have prefered and set the frame rate at 15 frames per second. I noticed that the longitudinal splines of the branches did not render quite right--they should be smoothly curved. (This is an occasional glitch I've noticed now and then--usually goes away when I restart AM. Just didn't have the time to start over.) Overall, I am very pleased with the results!

     

    Lighting Test

     

    Much thanks to the folks at ArtBox for the wonderful SkyCast! Bravo!!!

     

    Any comments, suggestions, etc., are very welcome. Bear in mind that this is a "quick" test, with much tweaking.

     

    Bill Gaylord

  18. Each leaf has 4 patches. That's just enough to give it a simple, but natural shape, including a slight cupping. The stems are now just splines rendered as lines.

     

    This project definitely involves a large amount of overkill, but it has a lot of potential for spin-offs.

     

    By the time I finish this, character animation will seem like a breeze!

     

    A movie of the branch growing in the SkyCast light is only about an hour away. Stay tuned!

     

    Bill Gaylord

  19. Here is a new sample of things to come. This is a rendering with one of the new branches and a brownstone apartment in the background. The lighting was done using ArtBox Animation's SkyCast. I must say I am very pleased with the results. The stems of the leaves are now just splines rendered as lines. Saved a load of patches to render!

     

    BstnSkyCast01.jpg

     

    Let me know what you think of it. I may render an animation of the branch growing in this light. Stay tuned!

     

    Bill Gaylord

×
×
  • Create New...