rickyjohnson

I like the bits of debris and dirt around the rooflight edges. It's looking good. Any plans to expand this to include the rest of the space shown on Dezeen eventually? That dark corridor to the side of the main room looks great in contrast. With the splashes of white paint on the floor - maybe you could paint a larger texture mapped over the entire floor area (or repeating once or twice) so that the splashes aren't suddenly cropped at the edge of each board. Also, I know you're applying artistic license with the levels but I'd agree with Francisco that the steel to concrete beam interface as you have it stands out a bit. If you threw some endplates on the steels with bolt heads (as they would be anchored into the concrete beam) then it would look more plausible.

you can add a UVW Xform modifier on top of your UVW Unwrap and use the figures in the tiling to scale your UV's by an exact amount relative to the size they're already at. I don't think you can scale UV's to an exact factor within the UVW Unwrap modifier's UV Editor itself (other than resorting to judging it by eye from the checker pattern in plan).

I'm not confident if this will be of any help because your buckets appear to be of a reasonable size and the light samples available should be good also but if you are stuck for other possible solutions: I had an issue previously with darker buckets in Lighting and GI elements that was caused by the bucket sizes being too small in relation to the domelight subdivs (if buckets are too small for any significant light contribution to be detected in initial samples then the entire bucket is considered to be in shadow. Something along those lines). There is a better explanation in the thread here if it's of any use: http://forums.chaosgroup.com/showthread.php?77462-Dark-buckets-Rendered-on-single-workstation-linked-to-HDRI-in-Vray-Domelight It doesn't really explain the inconsistency in your results either. There are a few 'dark bucket' threads on the chaos group forum that are specific to distributed rendering as well but I guess you'll have looked to those for a solution already.

I did try that with the substeps, yes. I set the figure for substeps as high as Max would allow me in the end (159) along with the highest allowable solver iterations. Still it's no good. I don't mind how long the simulation takes (within reason). I didn't experiment with any particularly high damping. I will give that a go.

I was hoping that somebody could give me some advice on how to achieve a good collision detection/physics model in this scenario: I'm trying to model tassels on the end of a throw. They're supposed to be a pretty chunky knit so my intention was to get a spline prepared for each one and then use them as a path for a Path Deform on the knit geometry (whatever that ends up being). The only thing is I can't arrive at a decent simulation for a bunch of splines (with a given thickness for the purpose of the simulation) that need to respect collisions with their neighbouring splines and also with the geometry that they're being draped over. So far I've tried the hair & fur modifier in Max: the physical simulation of the Guides just goes haywire when they meet a collision surface in anything other than the most basic of scenarios. I've also tried setting up 'ropes' in Mass FX using about 15 bone segments with constraints as joints: Again, in a really basic scenario (1 rope being draped onto a plane) this works fine. If more than one of the 'ropes' is placed side by side anywhere near one another the simulation just goes haywire again with ropes flying all over the place like a hydra! Also, it barely seems to cope with any collision meshes much more complex than a box (I'm aware of the 'hull mesh' creation process within Mass FX so think I'm doing this appropriately). With the settings on the Max simulations - I'm no expert but I've spent a day on this looking at various tutorials and the available help and I can't see any way to push the settings that would help to solve these collisions any higher than I have tried. I've also tried using inflated tubes in Marvelous Designer: The collisions are much better but the set up is very awkward and extracting splines from the resulting triangular mesh central to the volume is going to be pretty painful. Surely this type of simulation is possible? Especially when you see the level of simulations that people working in VFX create for instance. Is there particular software that might be better for this? Not that I can justify much expense for the sake of tassels but it would be nice to know if the 3ds Max systems are a bit useless rather than the way I'm using them! I'm curious as to why the collisions in Max seem to buck so violently when an intersection is detected. Wouldn't you try and develop the simulation so it slowed right down in this situation and would try and gently move the offending elements apart rather than rapidly flinging them in opposing directions? I've noticed this violent movement at times in Marvelous Designer also although it generally behaves a lot better than the options I've experimented with in 3ds Max. Anyway, and ideas or insight would be much appreciated. Thanks!