Jump to content

rasputin

Regular Member
  • Content Count

    997
  • Joined

  • Last visited

  • Days Won

    2

Everything posted by rasputin

  1. Wow... can you imagine a Compositor built into C4D ! That would be amazing... To do all your video work "in house", so to speak...
  2. Cerbera, am I the only one who's noticed that Align Normals almost NEVER works? On an extremely simple low-poly mesh, it might, but if the model is even slightly complex, Align Normals does not yield any appreciable effect at all. And my copy of Selections Suppletives gives me a menu... in POLISH! Which I can't make sense of at all, and I don't know how to change its language to English....
  3. Hey guys, You know what I would like to see in a future release of C4D? Better selection tools. It should be easier, I think, to make sophisticated selections of all kinds. Above all, I'd like to be able to, on a mesh object, select only Front-Facing Normals Polys... Or only the rear-facing normals. Also, the Phong Selection mode should be perhaps more effective... making it easier to select polys that fall along a slope. As the years have gone on, I find that my objects are often very high-poly... especially with the advent of the Volume Building features. With high poly objects, it's tedious and frustrating to make precise selections, especially if there is a jumble of Front- and Rear-facing polys in evidence. So, I say, I'd love to see R21 have some more advanced, clever and sophisticated Poly Selection tools. Thoughts? ras
  4. I'll prolly go in and tweak some edges to make that cushion look separate... usually with a finished VB model, I'll run it through the Polygon Reduction process.... You're right, some of the smooth perfection is lost that way, but not too much of a problem if the model is not sharply focal in a complex scene...
  5. Here's a 1950's Moderne-type sofa chassis I designed this afternoon. It was all done using the R20 Volume Builder. Basically a thick Torus primitive with various chunky shapes Subtracted from it. A shape like this would've been a nightmare to model using just the old tools, I think you'll agree. But it was not too hard to do with Volumes. I'm lovin' the buttery-smooth shapes you can make with the VB. I'll have the rest of this room scene built in the next couple of days and will show it to you when I do. ---ras
  6. Thanks, cerbera. I should RTFM, hm?
  7. A bit of a dumb n00B question: What exactly is an irradiance cache? And why would you want to cache any aspect of your GI solution? And what happens if you DON'T cache your GI solution? Thanks in advance, ras
  8. Hey guys, I'm curious as to what's causing this: I've got this image currently rendering in the Picture Viewer, using Standard Render with Global Illumination. As you can see, I'm getting some "fireflies" that aren't being smoothed out. I've never seen Standard Render produce fireflies before... Wonder what's causing it? How can I eliminate them (or get them smoothed out) in subsequent Standard renders? The gold and blue colors of the fireflies are probably related to the two Infinite Lights I've got stationed to the left of this scene, outside the walls of this "building": one is gold, the other blue. I just don't know why my interior textures here are picking up those lights with fireflies. Thoughts? Ideas? Thanks, ras
  9. Hey guys, I've set up a scene using Standard Render, to which I've added pronounced DOF post-effect, in both blur and in tint. All good and well. In the picture viewer, it rendered like this, which is what I want. Out of curiosity, I also rendered a Multipass Image, with all the available layers checked active, just so I could study them in Photoshop. To my surprise, the Multipass PSD Image-- with all its pertinent layers active in P'SHOP--- ended up looking like this. Not only considerably darker in appearance, but my DOF was nowhere to be found amongst the layers... neither the blur nor the color tint. QUESTIONS: Why did the Multipass PSD look much darker and over-saturated in P'SHOP, and where did my DOF go? Do none of the Multipass layers contain that DOF information? Thanks, ras
  10. Hey guys, This may seem like something of a n00B question, though I've been doing C4D for over a decade. It's like this: When you select a Cubic Projection mode for an object's texture, sometimes you are going to get visible seams, like the one you see in my screencap here. Question: What determines where those seams will appear on any given object? Can I hide/remove/disguise those seams.. and still keep my Cubic projection? I like the look of a Cubic Projection, it looks better than any of the other default Projection modes. Frontal projection will indeed give me a seamless look, but as we know, it risks looking unnatural inasmuch as the texture is distributed uniformly over the object, and changes in appearance with camera angle and focus distance. Thoughts? Hints? Tips? Thanks, ras
  11. What Cerbera said. This is a job for Marvelous Designer. Not only because of the restraining cords, but because the fabric in your picture buckles in a very attractive, natural way, that I'm not sure C4D Cloth can simulate.
  12. Hi gang, Would you believe, I've been doing C4D since 2003, and I still don't know what an "isoparm" is? I have created a spline-sweep object today, and one of the parameters is, of course, Isoparm Subdivision. Increasing or lowering that number seems to have no visible effect (that I can see, anyway) on my object's appearance. So what is the importance of Isoparm Subdivision? In what 3D situation would this parameter be of great importance? Thanks, ras
  13. Thanks, srek, very helpful. I am preparing to render a very complex scene, filled with many objects... I'm trying to pull out every trick in the book for making the render go faster.
  14. waitaminnit--- am i fundamentally misunderstanding the nature of LOD? Does it only apply to your model's polycount... not affecting the textures in any way?
  15. I'm curious: Am I right in thinking that C4D uses a kind of built-in LOD (Level of Detail) for objects rendered, depending on how close they are to the camera? For example, in an ultra close-up of an object, the C4D renderer will work hard to capture (and Antialias) every single nuance of your texture... Whereas, if that same object is far away, the renderer will not "work as hard"--- and go more quickly--- because it knows that the tiny details won't be perceptible anyway? Is it a good strategy for speeding up rendertime to assign different objects different LOD's? Like, if I don't want the renderer to "work too hard" on some objects in a scene? As it stands, I have a textured stucco wall in my scene (including some Normal and Displacement)... and the renderer is spending an eternity on it, capturing every tiny little nuance of the stucco... when it's the foreground objects I'm primarily interested in, not the wall. Thoughts? Hints? Ideas? Thanks, ras
  16. I'm wondering: Can R20 still do that old rendering method that used to be called "Camera Render"? This is where you lock the position of your scene's Camera, do a scene render, then, in subsequent renders it can simply apply an identical bitmapped "image" to some of your scene objects, sparing your computer's resources for really complex scenes? I used to use that approach sometimes, but it was a long time ago, with earlier C4D releases... Like, R8 or R10. To tell you the truth, I don't even remember how I used to set up that kind of render. Am I losing my mind, or did this feature once exist? Is this technique still do-able in R20? If not, when was this feature discontinued? Thanks, ras
  17. Hey gang, I'm trying to create a crumpled candy wrapper, as for the waxed-paper on unwrapped bubblegum. I want to get a realistic crumpled paper effect as seen in my attached (real-life) photograph here. How would YOU go about it? Thanks, ras
  18. Hey, VST, I have encountered much the same thing when working with volumes: sometimes my computer would be made very slow indeed by a subtraction or union. And my machine is fairly powerful, too. I found that the problem was greatly alleviated if I: made sure that the primitives or mesh objects I dropped into the Volume Builder did not already have an extremely high polycount. For example, I have my sphere primitive bumped editable with a Segment Count of 60, rather than the usual 90--100 I usually prefer. With your Cubes, make them Editable with a polycount setting of 8---8---8. The trick is to make your Volume Smoothing do all the smoothing work, rather than putting high-poly objects initially into your Volume Builder. switched the viewport's current Display Shading away from Gouraud and down to Hidden Lines mode, at least while I am doing the object translation needed for a Subtraction or Union; you can even leave your scene in Gouraud, and instead apply a custom Display Tag to your Volume Mesher to temporarily reduce the demands it's making on your computer resources. and of course, as you know, the whole Volume operation is made more CPU/RAM intensive if your Voxel Count is dialed very low, or your Smoothing Iterations is too high. One trick is to perform your Union or Subtraction while the Voxel Count is set to something loose, like, say, 5cm. Then, after you've gotten everything the way you want it, dial the Voxel Count down to something finer, like 1 cm. And, as you know, the Meshes you get as a final result of Volume Building are invariably crazy-high in polygon count. I almost always then apply a Polygon Reduction to that mesh; you can reduce your polycount up to 75% or more, with no discernible loss in visual quality. All in the name of safeguarding your computer's resources and preventing a huge hang or crash. And, vis-a-vis your texturing your objects: Bear in mind that, once you drop a mesh (or other object) into the Volume Builder, it ceases to be a "mesh". It has become a voxel volume, (a whole 'nother thang) and, as such, will not take a Texture. Only the Volume Mesher (in your OM) can receive a texture. And that Texture will apply to the whole stack of objects in the Volume Builder. Somewhere online, there is a tutorial of how you can use the Correction Deformer on the Volume Mesher to enable you to assign different textures to the different objects placed under the Volume Builder. Though usually your goal is to create a whole new mesh with a single texture, as the goal of Volume Building is not so much to allow you to to do Boolean-type operations, but rather to yield brand-new objects which possess the unique "edge-smoothing" look that Volumes afford. Best, ras
  19. Here's my current WIP: a human skull. This was done entirely using the SCULPT feature in R20. I started out by using Volume Building, where I united a Cube and 2 Sphere primitives, to create my basic "chunky block", roughly in a skull shape... Then the rest was all done with Sculpt, looking at photographs. As you can see, because we're not working with voxels, there eventually occurs some nasty "bunching and stretching" of the polys. Increasing the Sculpt fineness/density does not really alleviate the artifacts, sadly; nor does using the Smooth feature. Maybe future releases of C4D will feature true voxel sculpting... What it needs, of course, is a Retopo, and I don't really understand how to achieve that in C4D. I never did figure out how to use the Polygon Pen to effect a Retopo. The skull isn't done-- I still need to etch in those teeth! Organic forms like this are jolly hard to sculpt. ras

Latest Topics

Latest Comments

×
×
  • Create New...