Jump to content

Cerbera

Cafe Crew
  • Content Count

    12,202
  • Joined

  • Days Won

    417

Everything posted by Cerbera

  1. Yes it's hard not to look back fondly at 'The Stack' and wish we had something similar over here, particularly in regards to symmetry and edit poly etc. But with each new iteration good strides are made both in regular poly modelling tools and in Parametric workflow, so I think the future looks generally optimistic in that area. CBR
  2. Your model is very badly made, doesn't follow any polygon modelling rules, so the selections can't work properly. CBR
  3. No, exactly like it was before they changed it - internal, no browser ! CBR
  4. Yep, I agree entirely - I came from Max, and stayed for the nicer environment, and workflows etc, but shared a lot of your frustrations when I first got here. That was 10 versions ago when Cinema's modelling toolset was considerably more lacking than it is now, but we must remember it has always been trying to catch up with Max's and Maya's serious head starts and years of modelling development. Cinema's too has got orders of magnitude better over 10 versions, but we are going to need another few revisions to come close to the programs that had such a massive head start. Generally speaking, and as someone who has done mainly modelling for a living in Cinema I can honestly report that there are ways around almost every issue you mention, but they often involve slightly different workflows or approaches. Just very occasionally I do reach a brick wall where I can't find any way of doing something I need to, but in 95% of cases Cinema's modelling set (with HB modelling bundle and a few other small plugins) is enough to get most jobs done. In a lot of those cases Blender or something else can fill the gap. CBR
  5. If you really do need all those features you are better off in Max or Modo. Both are still superior modelling toolsets although MAXON are working hard to change that. CBR
  6. No, yes (Code Voncs Selections suppletives). Yes, restriction tag. Because you can't have degrees of bevel - it is either bevelled or it isn't / use component selections instead. Take the relevant value in the coordinates manager and add '+1cm' to the end of it, or use Move Quantize, which you can set to any value you like in modelling options.. All numerical fields accept mathematical input. If you need to move the Axis only, enable Axis Mode (L). Q by default turns parent generator on and off. Designed primarily to work with SDS, and therefore one of the most important things to have short-cutted. Yes. Make edge selections, and then either use '.' key or call Weight Subdivision Surface and type the value in you need. Works best with OpenSubDiv and subdivision 1 or 2 levels higher than you would ordinarily need... No, but HB modelling Bundle does, which is essential for anyone serious about modelling in Cinema. It also does quad-caps, comes with a very useful retopo setup, and features hundreds of modelling additions / alternatives. No, not the newer ones. CBR
  7. To be honest it is probably quicker and easier to create this whole model again from a regular helix spline, some instances of that and a few deformers and generators. For example the base shapes for both tower and track could be made as follows in about a minute with this sort of setup... So here 2 helix splines (can be parametric) are offset on Y slightly, then placed under a Loft with a Taper deformer to taper the resulting mesh. That goes inside a connect with a high enough weld threshold to join the surfaces. Now you could copy the splines / Taper, make the connect editable and continue modelling, and use the splines you saved earlier to create the outer track, also via a loft. Subdivision and the loft settings give you immaculate control of segment count and resolution. CBR
  8. Welcome to the cafe I'm all about the modelling, so our paths shall probably cross again. Anyway happy learnings and rewarding poly-pushing ! CBR
  9. No, probably not me - my specialism is more on the modelling side - I'm quite weak with effectors and priority problems as it goes, but have become used to seeing the type of problems that these issues cause, which is why I can speculate here. But I can't solve it so far, so looks like we're waiting for the people who can But at least those guys will have everything they need to find the answer, so that's a good thing... CBR
  10. Just upload to dropbox or similar and post the link here - bingo no size restriction ! This sounds like it could be a hierarchy order problem but would need file to confirm that. Feel free to strip out all the unnecessary other objects and materials before you upload it ! CBR
  11. Hierarchical order very important in Cinema. Your delay effector needs to be at the end of the effector chain, as Cinema scans downwards from the top of the list, and executes in that order, so in your version the delay was only affecting the Push Apart and not the effector after it in the list.... CBR
  12. Glad you solved it, and thanks for posting your solution. For future reference, we would have been able to see this immediately had we been given a screen grab of your project, or if you had uploaded the file... conversely, without those things we could have been guessing for hours and still not got the reason ! CBR
  13. Please update your profile to the version you are actually using - that way we won't not recommend components we don't think you have... How did you duplicate the Xpresso tag ? If you didn't do it with copy and paste, try that, and it might well work. CBR
  14. Cerbera

    Grass in wind

    Cloth systems could maybe work for this, and it is easy to manually fix points (see dresser options / fix points). You'd have to turn gravity way down, or even to positive values so the grass didn't collapse. Normally, for grass, the Hair system is most ideal, which has fixed bases by default, but I suspect that won't work with megascans because the texturing options with hair are rather more limited. CBR
  15. Welcome to the cafe Are you not defining where textures go with polygon selection tags ? Because I rather suspect you should be, although I can't tell for sure without seeing at least a picture of the setup, the polygon structure of the object in question or some reference of what you are aiming for... CBR
  16. Please use the cafe's image upload system rather than bouncing us out multiple times to external sites. Yes you pretty much require Sub-Surface Scattering for this sort of thing, and that needs to be setup in quite a specific way (see SSS section in manual)... But I think you want to do the entire background in post, and use a shadow catcher to produce shadows that will later work over any background. So you need one set of lights that illuminates from the front and sides and casts shadows how you want them, and then you can make additional specific omni lights in the scene positioned behind your gummies (and affecting ONLY the sweets, and with shadow casting off) to illuminate them nice and brightly which will best catch the light with the SSS in your luminance channel. CBR
  17. Speaking as a guitarist and music producer among other things, when a guitar is strummed you wouldn't see any sinusoidal movement in the strings, even the lowest ones - the overall perception to the eye is only that if it is slowed down massively, or the note being played is ridiculously low. So for that reason I would instead animate an 'up' frame and a 'down' frame, where the strings deform from the point the pick hit them, and then motion blur between them to get the overall effect of a string being strummed. But if sine movement is what you want, then I don't think you can use your existing strings - you need new ones that are based on splines that can be displaced by, say, a formula effector. If you could make that work for you then you'd only have to keyframe 3 parameters therein to get the effect (scale in 2 directions and falloff). Then you could just sweep those splines to make them renderable. That works just fine on a straight string made from a helix, but of course guitar strings don't do that because they have to go from tuning pegs through the nut and over the saddle and terminate in the bridge, which requires a lot of angles. However, the only place the string vibrates is between the saddle and the nut, which, in theory should be a straight section of spline, so the cheaty way would be to simply do those bits separately, and hide the joins somewhere in the hardware at either end (there are usually grooves in both nut and saddle which can help with this), but if you don't want to cheat and instead choose to make it form a single spline, then you can use secondary fields to further restrict the area affected by the deformer - just bear in mind you need to subdivide the spline segments you will be deforming so they have at least 100 points in them, enough to support the sinusoidal deformation... There may be other ways too, so will be interested to see what anyone else contributes... CBR
  18. What is the collision shape set to on the cylinder's dynamics tag - Automatic is it ? If so, try changing that to cylinder, or even moving mesh. CBR
  19. Yes it is Here's Stage 2 of the modelling approach, which consists of taking that surface we made earlier, temporarily removing the bevels and SDS, spline wrapping it via a circle and a rail, and FFD-ing it into the right general shape like so. This requires quite a complex and specific hierarchy to achieve, although it is remaining nice and parametric at this stage... That gives us this sort of result once bevels and subdivision are re-applied, at which point we'd be ready for Stage 3, isolating the patches of this we need, applying symmetry, and modelling the rest of it etc etc... CBR
  20. Just to give you some basic insight as to what is involved if you try and model this, here is the first stage, in which we design a low poly scale that is eminently tileable in a honeycomb array to produce a usable surface... So on the left is our source tile, middle is what that looks like cloned and connected, and on the right bevelled, subdivided and ready for further deformation... stage 1 of many... CBR
  21. No, that's simply not true when you need to use parts of the material workflow in order to get your modelling done, which is what Dan is wisely suggesting. If you had the level of skill required to be able to model this by hand, you wouldn't need to be asking this question. And back in R18 you don't have the volume builder, so no lazy, cheat-y substitutes like cloned scales or spline wraps are going to work here either because you need a solid mesh for printing. So your only options are 1. modelling it by hand, which is very difficult, very time consuming, probably beyond your skill level, and out of the scope of what can reasonably be achieved in any single forum post or 2, with sculpting stamps / bump / normal / displacement, which is what Dan is suggesting, along with his exceptionally generous offer to even make you the map ! I could tell you how to model this properly step-by-step, but that would take around 4 hours, and would be a chargeable event. CBR
  22. Why are you subdividing this via sculpting ?! That workflow doesn't make any sense at all. Lose the sculpt tag, and just use Sub Poly Displacement in the material. Now we know you are rendering with Arnold, I don't have that, so can't help with the specifics of how that works in an Arnold material. But this is no longer about your phong settings, so probably fine to leave those as they are, although you can turn off edge breaks, as there shouldn't be any in an organic model like this. CBR
  23. Ah yes if it came with a displacement map, then it will require that in order to work. If that is a greyscale image, you can use it in the displacement channel in your material, and you should probably turn on some level of sub-poly displacement as well. CBR
  24. I'd be checking to see if it requires subdivision, then any normal or phong tags it imported with. We can't see any details from the shots you provided so it is impossible to confirm what is wrong without the file. CBR
  25. Oh that's annoying. I was hoping we could use a formula spline to control the bevel, but we can't - that field only takes editable splines not parametric ones. So with the external profile spline option being out of the window for that reason, and if animating the user curve is unpredictable and weird, then I think the answer to this, to all practical purposes in your current situation, is no, other than what can be achieved via something like a post-bevel falloff-enabled displace deformer - perhaps that can help here ? I am by no means an expert in Xpresso, but looking at the available ports for a bevel deformer the user curve is not an addressable property - only Shape, which is an integer, so I am guessing that option is out too. Failing that, if animating the bevels is the main focus here, then you might need to rebuild the text from an entirely parametric setup you can alter after the fact, like making it out of lofts for example - much more work obviously, but sometimes there is no other way... So here, it does seems like animating the user curve on a frame-by-frame basis IS the fastest way to get this done. CBR
×
×
  • Create New...