Stuart Rogers

I suspect this is a 'normals' problem. Each patch has a front side and a back side. Hairs and other particle systems emit from the front side. Most of your patches are facing outwards, producing hairs. Some patches are facing inwards, which produces ingrowing hairs - the little dots on the bare patches suggests this - they're the roots of the hairs. If you look in the Tools->Options panel you should find an option to "show normals" - turn this on. It might be useful for this to turn hairs/particles off. In the show normals mode you get a single pointy hair in the middle of each patch, pointing out from the patch's front. Select any patches that are facing inwards and flip their normals (in the context menu of the selected group). Once you're happy all your patches face outwards, turn off normals and turn hair back on. Of course I could be wrong about all of this...

Oh it doesn't have to be all that big. A few years back I came across a web site describing an experiment that put a 25MHz 486SX machine inside a domestic fridge-freezer and managed to cool it enough to crank the clock speed up to 247MHz.... (link)

One thing I noticed the other day, which is obvious now you've shown them side-by-side, is how the original 3DSM version doesn't have the mapping quite right - it pinches at the poles a bit. If you look closely you'll see some of the white spots on Jupiter's northern latitudes are stretched north-south - in reality such spots, if they vary from the circular, stretch east-west due to the prevailing winds. I found the figure of 1195km quoted on NASA's web site, and 1132 km elsewhere. Most quoted figures have errors of +/-20km. Stick with the value you have. I would check my old univertsity text books (I have a BSc in Astronomy & Physics) but they're waaaaay out of date.

My PC used to sit on the floor and it used to get a bit dusty (but not too bad given that the floor was carpeted). My PowerMac has always lived on my desk, and that's almost as dust free inside as it was when I bought it two years ago. So, if you can afford the deskspace...

But I bet it picks out the skin texture really well when it's animated! Nice work

Wow, that's quite a tour! I wouldn't be surprised if this really was the case. Many years ago I wrote a program to simulate n-body problems with objects on an interplanetary (and, ultimately, intergalactic) scale, and I found that I couldn't work with 'normal' measurement units as it would always end up with rounding errors, overflows, and underflows - I had to use, for example, 10000km as my basic distance, and mass in terms of milliEarths, and time in hours to keep the numbers sane. This, of course, meant I had to redefine the gravitational constant too. If precision is the problem here, scaling the whole thing down might help, but I suspect it might not. If not you might have to resort to compositing a render from a scaled-down 'orbit lines only' choreography with a full-scale 'planets only' choreography.

First things first... If you can create a small example project file for which this a repeatable problem, consider going to A:M Reports to raise a bug report. Does it do it for all final file types? For example, if you've been rendering to a movie file, does the same thing happen when rendering to individual TGA files? If so you'll at least have a workaround. (And if you've been rendering to TGAs, does it fail for movie files?) If you're trying for Final Quality, then no. (Out of curiosity, do you get problems rendering to file for other-than-final?) I've never had an occasion in which weird glitches like this were solved by a reinstallation, so I would hold back on that for the moment. Have you tried rebooting your PC? Every now and again even hardened A:Mers start weeping about strange problems that simply go away after a complete system restart. Hyperthreading has nothing to do with graphics - your Athlon will do just fine.

If the replacement model has the same rig as the original, and is of similar proportions, then it's a piece of cake - each instance of the model in the choreography has a pointer to the model definition, and changing the pointer to point to the replacement model is no more effort than choosing the replacement from a pop-up menu. If the replacement model has the same rig but different proportions, some tweaking of keyframes might be necessary, but it'll typically be fine-tuning rather than wholesale changes. It appears to me that the keyframe channels created in the choreography refer only to the names of the corresponding items in the model definition and not anyunderlying mechanism. So, if both the original model and its replacement hava a pose named "extend tentacle", the pose will still work even if the models differ vastly in how the pose works. And if the replacement model doesn't have a pose with that name, the channel created using the old model will simply be ignored for the replacement.

Excellent - well done!

Fabulous - I love the hip movement.

It looks good to me - but creating 20,000 spectators will be quite a challenge, though!

Don't make them quite so black. If you haven't done so already (it's hard to tell from these images) I suggest you bevel the edges between surfaces - instead of a sharp angle between the side of the knob and its end face, put in a curved edge. Very few edges in real life are infinitely sharp. Such a bevelled edge would improve the chances of catching specular highlights. Also, consider adding knurling to the sides of the knobs (a serrated grip, if you like). Again, this would help the knob catch specular highlights. The knurling could be done as modelled splines, or a bump map, or as a displacement map.

Are you working in wireframe mode? If so, try doing a Quick Render - hit the Q key and then click in your model's window. Do you see the decal there? (Please find your Caps Lock key and use it - when you write in all capitals it makes me think that YOU'RE SHOUTING!)

Woo-hoo! An interesting approach. Here's how I would have done it... I would make the umbrella an entirely separate object. For those scenes in which the umbrella stays in Ebon's back I would create an action with the umbrella as an Action Object. For any other scene I wouldn't use the Action, but add a separate umbrella to the Choreography and constrain it there. That way it would be no problem transferring the umbrella from one character to the other. Of course, if you'd used my (unreleased, ever-evolving, never-to-be-finished) rig, you would attach it to the geometry hand, which blends from IK to FK quite smoothly. EDIT: I see you've just addressed that in reply to KenH. It looks good! Things I like: * The new 'acting'. * The detail of the scar in the palm of his hand. * The design of the room - the interaction of light and shade is particularly apparent in this shot. Things I'm not so sure about: * The sudden movement of the camera at the end. I realise you're trying to give a sense of sudden surprise from an "almost POV" shot, but I found it jarring. I wonder if keeping it moving slightly during the head shot might be better - I find accelerating from a slowly moving shot easier on the eye than from a static shot. YMMV of course! One question: how long are those strides? He does seem to take a lot of steps to cover that distance. This is really excellent work, and I look forward to the next clip.

That's a lot better! The background hills look quite bright where seen through the front and rear windscreens. I'm sure this must be due to bright sky reflections adding to the transmitted light. I'm not sure what to suggest to fix that - as it is I'm feeling guilty for picking holes in your work when I know I wouldn't have the patience to get it looking this good myself.

Where you're data's concerned, never ever ever rely on anything! Take back-ups all the time, and then take another.

Yes. If you single-step through it then you'll see that on odd frames where a wing overlaps an arm or a leg, the semi-obscured part of the arm or leg goes very bright and uniform. It also appears on some frames of wing-over-wing. It's not on every such frame - it looks as if it happens when the offending wing is square-on to the camera.

<grin!> Mostly when our government agency spending gets like that it's only an aspect of internal accounting, and not real.

It's 'Mylar'. We used to have a stock of it in our lab - our remote sensing department used to use on occasion as they were involved in satellite instrument calibration. When our branch secretary retired we wrapped her leaving presents in Mylar (we were in a hurry and couldn't find any wrapping paper in time) making the wrapping far more costly than the contents.

You think that's bad? I'll be no help too *and* I'll risk sending the thread off-topic by pointing out that there's also some weirdness going on with wing transparency about two thirds the way though...

I like the more saturated colours - it makes much more of an impact. Excellent. The front windscreen still looks too clean. How about adding a dirt map outlining where the wipers have wiped? You'll probably need transparency, diffuse, reflection, and specularity maps for that (that soundds a lot but you can derive them all from one master map).

I've seen this. It usually crops up when I'm reworking splines that have been previously joined together with the shift key held down (so that smooth flow from one side of the CP to the other doesn't occur (on purpose, I might add)). It's as if the CP retains that mode even when detached from that connection and reattached to another spline. Or something like that - it's one of those things that hasn't happened often enough for me to clearly understand the problem, so I've yet to file a report on it. (I last saw it on v12.0 but I haven't done enough modelling in v13.0 to see it again.)

That's a nice running-in-water action. The ripples and water droplets look good - hopefully you'll be able to get them together, along with lots of much smaller ripples for the drops that hit the surface. I think you have too rapid variability in the mist at the bottom of the waterfall, though.

Oh we can do better than that... The car's going fast enough to kick up stones yet there's no motion blur on the wheels... The front windscreen is spotless.... And... there's no driver! <grin!>

I like the mud along the bottom sill. Are all those stones on the ground modelled? If so - wow!