Cerbera

I would guess that this level of polygon density (almost none) is not enough to be to be able to produce a vertex map usable for keeping the scattered instances far enough away from the edges of the surface not to intersect the adjacent geometry (which has no proper real world thickness)... If you add a decent number of loop cuts into that your vertex map can be correspondingly a good deal more accurate. CBR

Field Forces, Volume Builder improvements, Remesher, massive UV improvements, updated modelling tools, updated IO modules, improved Motion tracking, much superior viewport, sculpting enhancements, place and dynamic placement tools, scatter tool, better spline pen / spline tools, better primitives and options within them, Asset Browser with unified filter presets, object presets, new GUI, Direct X VP, nodal materials, scene nodes / capsules, Magic Bullet Looks, intel de-noiser, the list goes on.... those are the ones that immediately spring to mind, but there are a lot of smaller quality-of-life enhancements I'd struggle without now. CBR

I don't have Arnold, but I can get it working with physical, which may help you regardless of renderer... But without it, I can't come up with a solution without making the hair physical, but if you choose circle or similar type in the generator tab of the hair object you can force it to create hair geo (and can turn hair render off), which you can then texture using regular projections, which should stick to the hair geo, like so. There is no reason to think that wouldn't also work in Arnold, though I wouldn't be surprised if you had to actually make the hair editable, and bake down a cubic or flat projection by deleting any existing UV map, selecting the projecting material, and doing 'generate UV cords' which would then the projection into the mesh. So, perhaps a fall-back method if there isn't a forthcoming answer directly within Arnold hair / material settings... CBR

My question would be - if it doesn't move, why can't be baked in ?! You could always keep the original in a new file so you had it to go back to ?! Unfortunately the displacer does update per frame, and that can't be turned off, so it is a bad idea to keep one of those live during an animation when it is applied to a very dense or excessively large piece of geo. The best you can hope for is some minor mitigation, theoretically achievable by dropping the Level of Detail Setting to Lowest, and if possible reducing the amount of geometry in the affected mesh, possibly via a remesher, although of course that must also calculate... but the point remains - by a long mile the most effective thing you can do to solve that is to apply the displacer. CBR

I wouldn't take out the edges you have marked, but rather these ones instead... By doing that we eliminate: The pole on border at A The triangle at B The over-density at C Before doing that I would bin off the cap end down by D, which I would re-patch later when I know how many edges will remain. As for the fact that 2 of the edge loops I want to remove run all the way down and into the cockpit that doesn't matter; it will leave just enough loops there to describe that shape once the neighbouring loops have been slided (preserve curvature on) in to fill the gaps. Quite a lot left to discuss, so on to the videos... CBR

That's OK - you are on the path to training your brain for working in the way it needs to for good modelling, but that path is quite a long one, and there are a a lot of bumpy, windy bits along the way ! 🙂 It takes a very long time before you can see ALL the things, and consider ALL the factors at every stage that leads to making reliably good modelling decisions. But like everything else in life, the more you do it, the easier and more reflexive it becomes. Re-builds are not to be feared or avoided; they are the most valuable learnings you can get. Sometimes you need to see stuff wrong to know what is better and why and that is more true while you are learning. I still re-build stuff sometimes, which surprises some of my students, but I am quick to point out that given such a vast array of topological choices we can make at every stage, and how much every stage affects all the subsequent ones, any and all of which will be somewhere on a gradient between flat-out wrong, through 'serviceable-but-not-ideal' to 'flawless victory', is it any wonder that we don't always make the perfect decision first time, even when we are quite good at this ?! For example, shortly after I posted my 'A*' solution above, I continued to find things I could improve about it for quite some time afterwards ! Were any of these things strictly necessary ? Arguably not, but I did them anyway, because I enjoy the process as much as the result 🙂 So here I have improved the red section by adding a perimeter loop at the front, but taking out 2 others nearer the back and redistributing things more evenly so that mesh resolution more consistently steps down the further we get down the nose... CBR

Dude - they went to SPACE with the second one (spoilers) !! And did a clever switcheroo with a giant alien roadrunner at the end ! I'd say as far as expanding the concept went they did OK ! 🙂 And moreso I think they HAVE to stick fairly close to the original plan of 'coyote-tries-and-fails-to-catch-roadrunner-with-ever-more-inventive-array-of-machinery' or we lose the central premise on which its success depends and was built. And imagine how disappointed we'd be if there wasn't at least one 'delayed-gravity' effect ! But that small point aside, thank you for a fascinating read on the subject - some very informed and salient points there... CBR

That is one LOW-ass fender / grille thing up front there. Isn't that gonna be dislodged by the first stone on the road ?! 😉 CBR

Righty - on to a bit of cockpit planning. Let's get a little temporary guy in there so we can visualise scale properly, and then we can work out what edge flow we need for the inside of the ship. Initially that should be just the wider 'cockpit seat area', and assuming that all consoles, the seat itself and other inner detail will be separate meshes, we only need to define the general bowl shape as it comes down from the rim. Initially I would plan the edge flows thusly... ...but also keep in mind that probably need some sort of footwell also, depending on how humanoid whatever is driving might be ! So that might end up looking vaguely like this... And if needed we can further extrude the section behind the chair back into the hull to make additional space. I reckon we should think of this ship as being slightly bigger than it is now, in relation to the pilot(s). For example - maybe this is more like it... if it was that scale, then the canopy would probably fixed, and entry into the ship would be via the rear, whereas if you go for the much smaller scale above and single pilot, we might need to worry about canopy opening mechanisms - if indeed you need to show the inside at all !! Perhaps you don't ! 🙂 CBR

Yes yours is still a bit lumpy up by cockpit there isn't it, so a bit of 'looking along the lines' and checking for contiguous curves is required there. There will be some points that break that continuity, and therefore add the lumpiness. But the edge flow is now more correct than it has ever been, so that will help in your fixing of that. Now, the next thing we need to look at in our quest for perfect topology is what damage has been wrought by the staged subdivision we added. Sure, it has given us the resolution we need up by the cockpit, but of course it has doubled it everywhere else as well, which increases the potential for lumpiness down the nose section. The orange crosses below show where this is most likely to occur, and, as usual will be the result of all those pesky poles that have crept in, complex or otherwise. BTW, the red crosses there highlight triangles where we have forgotten to dissolve the centre edge into kite quads. So, the final part of doing staged subdivision is to consider taking loops out in areas where we need less detail, so that SDS can reassert itself nicely, and resume doing most of the work for us. I went on a war mission against all those polar junctions, and eliminated them all, to arrive at what I consider to be the holotype topology for this model, which is thusly... This has given us a result that is what I would call optimal, and as far as I can tell there is no lumpiness remaining. So, with that done, now is indeed the time to be splitting off, and adding thickness to the canopy section.... CBR

That's looking pretty decent now I'd say 🙂 But of course there is always stuff to improve... let's have a look at this little area... Technically, your bevels there have redirected the edge flow around those sections, but look at the SDS stress inside and outside those, which gives a warning that this is not enough... so let's turn SDS off and look at what the issue is... So what we need is the inside edge flow doing something better than that (green lines) so we can avoid those simple poles on the borders (orange cross) and fix that degenerated (concave) poly (red cross) which is actually in an error state currently... A little bit of re-cutting gives us what we should have there... So now we see a contrast between the inner section that we have fixed, and there is now no SDS tension, whereas on the outside, which we haven't fixed, we still see the edge flow being wrenched off to the left where the pole meets the bevel. So, to fix that other side, we just need to select the whole edge loop the other side of the bevel, and ctrl-slide a copy of it away from the bevel, then redistribute the points more evenly. Now we have a mesh where the SDS tension here has been all but eradicated, or lessened and moved to where it doesn't matter. The result ? Sexier curves, and yes, that will come through in render ! CBR

If you want to increase your Octane skills in a serious way, may I suggest New Plastic on Youtube.... CBR

Oh good - glad it helped ! Now you just need some lights so you have some decent shadowing in the scene to separate the watch from the background ! HDRI based lighting alone isn't usually enough to provide this. Also I would have to question the wisdom of putting a mostly shiny object into a mostly contained box, where the reflections from the HDRI can't actually reach it except on 2 sides, which is why so much of the strap is just pure red and it seems to merge with the BG. CBR

Welcome to the Core 🙂 First things first - pls complete your profile so we know at a glance which version of Cinema you have, what you are rendering with etc etc. We need this information to be able to help in most cases. Fortunately we have the scene file this time, which tells us, but even so... This is currently what the UVs look like for that object, so Octane can't be set to be using them like we need it to. Tag should be UV mapping, and Octane Projection mode should be Mesh UV. In your Octane node tree, no Projection is assigned at all, which may explain your results ! So I would add that in first.. So to sort that UV mapping out we need to project all the polys frontally, from the Front orthographic view over in BP UV Edit. Then we need to select the polys that comprise the rims for the holes and relax them. That will look like this when done, but wait - what are those weird seam lines in the middle ?! They are caused by points not being welded in that mesh, so you need to optimise it, or manually weld those double points together before you start. Then you should get this result... So texturing is now fine, but we have some duff SDS technique going on here, so you need additional control loops to tighten that up, and to not have poles on borders, which is bad modelling technique generally. So each of the holes needs a perimeter loop, which we should get by insetting all the front facing polys like so... # ...and then we need control loops inside the holes as well, so that SDS rounding is controlled from both sides like it should be. K,L for Loop cut (preserve curvature off) to pop those in. Then we will get a much nicer SDS result because the edge flows have been redirected properly, and rounding is back under our control. Lastly, I notice you have some unevenness in the polys comprising that lowest hole, which is affecting the SDS circularity of that shape. To fix this you would need to make sure sure that all the edges comprising that are evenly spaced, and then conformed to circle once that is done. The plugin HB modelling bundle is your best friend to fix that (hb_even_distribution / hb_points_to_circle), but it can also be done manually. Note that if you do want to fix that, you need to do it BEFORE you project and unwrap the UVs. CBR

Yeah I get some tingle factor looking at the progress here - bravo ! Lovely stuff. Floor needs 10 years worth of jackboot scuffing though - I don't think that's as shiny in the films, even on Emperor visit day ! I always assumed those little zippy cleaning droids would struggle with boot polish 🙂 CBR

Wow - so everyone really hates that then ?! 🙂 Perhaps I'm easily pleased, or slightly misty-eyed with nostalgia, but I thought it was alright ! I agree of course it lacks the charm and absolute expertise in the older originals, but I guess it's all about budgets and mass production now, and presumably thinking that they don't need to work quite so hard for their audience of mainly small, distracted, attention-deficit children ! Perhaps that also explains why they don't feel the need to roll out a full orchestra anymore either, but again I'm in the minority, preferring pristine modern Virtual Instruments over the wow, flutter and epic noise floor of the original recordings from the 30s ! CBR

It's been a while since I watched any cartoons that weren't Simpsons, South Park, Rick and Morty, or Disenchanted, but the YouTubes found me this the other day, which I was very interested to see has progressed from classic line art / hand drawn cartoon to 3D. Some people not appreciating the transition in the comments, but I think it works quite well... and was fun to catch up with those guys and know the great war is still being waged after all this time 🙂 Also, if you are learning to score to film, there is little better example of cue-based scoring, so a very good lesson there too. CBR

Where on earth is Octane in that list ?! That can do awesome Arch viz renders !! Oh - it's right down the bottom there, with just 2% of respondents using that. Wow - thought it would be much higher up the list ! CBR

Drag it there directly out of Asset browser ! Or if following Chris's video it should end up in the spline primitives menu, which does the same. CBR

Lols - you beat me to posting that by about 10 minutes ! Great tool (I thought), great video ! 🙂 I think it's a LOT more useful than what we can do with Edge to spline, and (to a lesser extent) the cloner... did you watch that vid all the way to the end ? CBR

Yeah I didn't wanna use the term 'oldschool' in case it was your favourite plugin or something, but yes, sample tech has moved on a way since then 🙂 Nice to see Cinema in UI design before it took off in that field, and what you did looks great ! CBR

Yes - that one's not bad at all ! The Bosendorfer is bit ropey, tone-wise, specially at the extended low end. Overall, not as good as keyscape 😉 (IMO, but then I would say that) Annoyingly, the guy making that demo plays nicely, but does have his system latency set too low (hence occasional clicks) and his output set too high (distortion) in all those demos ! Promise those aren't an issue on a powerful machine. CBR

CBR

I did, but only cause DV explained it in another video 🙂 That is proper genius what they did there. Guess what colour that sand goes if you invert it ?!! 😉 CBR

Wow you got a REAL Rhodes ?! I take it the sound is worth the backache ? 🙂 CBR