Having used Bryce for a couple of years, I sort of expected Carrara's Boolean operations to look the same. When I punch holes in things (like a rectangular window in a wall) I get strange artifacts, like misplaced facets, which also show up in the render. Am I doing something wrong? JWB

No, nothing wrong. According to Antoine, one of Carrara's creators, boolean's are notoriously problematic, even in other modelers. His suggestions include: increasing the poly count of the objects and then decimating after the operation; Increasing the Fidelity settings for the operation; trying to use the spline modeler if appropriate, and finally, avoiding booleans if possible. Most of the funky artifact issues are a symptom of too low a poly count on the faces being changed by the boolean, however. Mark

NarkBremmer wrote: Most of the funky artifact issues are a symptom of too low a poly count on the faces being changed by the boolean, however. If that is the case, then one would expect that the more complicated the objects (which are the subject of the Boolean operation), then the more likely it is to suceed without errors. Or have I misunderstood? In my experience (limited though it is), the trick to succesful Boolean operations is to keep the objects simple or at least reduce the number of verticies as low as possible in the area of interaction. For example if you try to do a Boolean subtraction/union of two significantly overlapping Poser body parts, then the result is almost always a crash. However, if you delete most of the parts that are not going to appear anyway, then Carrara may be able to cope. I would be more than happy to have Antoine correct me if I'm wrong, or to make some suggestions for reducing Boolean problems (other than "don't go there!"), as I seem to use them a lot - they are very useful critters (potentially)! IsaacNewton

The booleans in Bryce do not act the same way as they do in most modelers. Bryce Booleans only act upon how the light & material react to the shapes, they don't really change the shape but only "hide or reveal" the intersected space. Carrara actually tries to guess the intersected area and omit unrelated vertices and faces. Yes, if you manually remove vertices beforehand and then add more subdivisions in the area to be acted upon then your boolean will be more true. And, it will take you much longer to accomplish the goal (especially if you have to redo a couple of times. Frankly, for the extra time you will spend in fighting with booleans you could learn to use the Spline modeler to create that wall with a compound shape to make the window. I once did a boolean that took the machine about 1 1/2 hours to do. If I knew then what I do now about the Spline modeler it would only take about 15 minutes to create the object I needed from scratch. You are "stuck" with objects that you bring into Carrara though (like the Poser parts).

Thanks everybody for the help. As I read through this archive I found others were having the same problem. I did eventually create the wall in the spline modeler, which accomplished the goal. I will also try increasing the poly count, etc. to see if that improves the process. I especially found the Bryce explanation helpful as that made Carrara seem less clunky. JWB

Most of the problems with booleans in Carrara are due to bugs in the vertex primitive and really cannot be overcome... you can work around them, for instance there was a specific bug where if you placed a cube and a sphere at the origin then did a non-uniform scale on the cube(there was a specific range but I don't remember what), did a subtract of the sphere from the cube, it would crash. But only if the fidelity was cranked to %200. Set the fidelity to %199 and it didn't crash. As for the old flying polygons, that's a different issue. The first thing to do is a test render on the area to see if they're actually there and not just being drawn by the interactive renderer. If you see them in the vertex modeler, they're there. Undos and booleans don't mix because half of the time information doesn't get written into the undo buffer properly, if at all, causing many a crash on undo. I think the key is high polygon count and medium fidelity. But there's no real cure.