On 14.02.2013 13:09, Brecht Van Lommel wrote: > Hi, > > On Thu, Feb 14, 2013 at 9:59 AM, Gaia <[email protected]> wrote: >> So i have to ask back: Where would it be bad to get a visual >> information about unreferenced vs. referenced verts using >> the black/blue color code ? > I would think that we should improve our tools so that there are never > any vertices part of a vertex group with weight zero, and not let the > user worry about this at all. Why burden the user with an > implementation detail? Because it is relevant for some usage scenarios outside of blender.
I am 100% with you regarding "hide the technical details". And indeed this specific detail might best be hidden from the user as long as we keep everything inside of Blender where this detail is unimportant. But actually it becomes very important when we look outside of the Blender universe. And indeed this little difference between unreferenced verts and zero weighted verts has a huge impact on the display of 3D objects in our game environment. > I can see a visualization being useful where vertices with zero/no > weight are shown in a different color so that you can quickly spot if > vertices are part of any group, but this should then make no > distinction between the vertices being internally part of the group or > not. We already have dark blue as the color for zero weight. No need to change that. "not referenced by any weight group" is a bit fuzzy, because weight groups can be used for very different and unrelated purposes, so this visualization would be pointless imho. But i still believe that getting a visual feedback for "not referenced by this bone" is a very useful information. (See next comment below) And actually i do not understand where that is a technical "detail" (apart from "blender doesn't care so the user should not care either" ... but the universe cares, so do users ...) >> There are situations where knowing if a vertex is referenced >> (even with zero weight) becomes relevant. It can even be >> wanted that verts are members of a vertex group but weighted >> to zero (or almost zero). And in such a case cleaning up low/zero >> weighted vertices with blender cleanup tool might be not practical. >> And seeing them all in blue doesn't help either. >> >> One case where it is mandatory to have each vertex be placed in >> at least one vertex group is when it comes to create items for >> game engines outside of Blender. Yes, we already have added a >> lot of helper tools to our own addons for finding unreferenced >> verts, visualising them, forbid to export partially unweighted >> meshes, etc. >> >> However getting this little extra visual control would help as well. > I can see game engine exporters need vertices to all be part of some > group, but if like Blender they only consider weight > 0 to be really > part of a group then this can work fine still without exposing > implementation details to the user. But it is not about automatically removing zero weighted verts. it is about feedback about unreferenced weights and feedback about "unintentionally belonging to this vertex group" verts. Exporters can not reliably handle that automatically. One very simple example: During weighting you unintentionally added some barely noticeable very low weighted areas in the head of your beloved character while you where painting on a foot bone. yes this can be avoided, nevertheless it happens. Now try to find these verts on areas which do not belong to the active bone... Try to find less dark spots of blue in a sea of dark blue. Mimicry works well here :) Ok, i hear you telling me i can use the "clean weights" tool and be happy. But what if i intentionally WANT to have some weights to be set close to zero (or even zero) ? Then cleaning the weights might ruin my weight maps. And again it sounds very appealing to me: If we had unreferenced weights be displayed in black, we could quickly step through the vertex groups and see blue spots in the black seas and that's much easier to spot. well... If that still doesn't make it clear why it is a useful improvement, then i am running out of argumentative ideas for now :) cheers, Gaia _______________________________________________ Bf-committers mailing list [email protected] http://lists.blender.org/mailman/listinfo/bf-committers
