UV-mapping category

you select a few edges in the uv-editor and the move and sew tool will automatically find the uv-patch which also shares those edges. the found uv-patch will be moved/rotated/scaled to make the corresponding edges fit as good as possible for connecting the two patches. finally the vertices will be weld. many uv-editors like the ones in maya and modo have this feature.

Say you have a 3d model of a head, but the eye is in the wrong place. With this tool you could drag the eye down in the 3D view and the UV map would change automatically. I know you can do this in the image viewer but its not very natural.

(I don't know whether this is possible in the 3D view already, but I couldn't find it anywhere)

Weld all the selected UV-vertices together, that are within a specified range apart from eachother. ( like remove doubles )

This would be useful for some things that we have to do to save texture space in realtime application like setting up mirrored UVs.

Basically for meshes where there are two or more faces of the exact same size in 3DS Max you can get one of the UVs aligned to a face, then copy that UV and paste it to the other faces. Copy and paste in this way is FAR, far quicker than aligning each and every UV. Imagine the pain if you had half a dozen or more faces of the exact same size!

The best way to do this in Blender is to make it so only the size and positioning values are being copied from one UV, and pasted into the size and positioning values of the other UVs. This would ensure that nothing is messed up, because you aren't copying the UV itself, just the specs, and you're then giving those specs to the other UVs.

The result would be much quicker and more efficient UV editing. Combining this with my other suggestion, to have a keypress which will allow you upon typing in a value for all the selected vertices to move them to the same location at once instead of an average point, will result in the user being able to get their UV mapping done a whole lot faster.

I made this topic before and it was WRONGLY marked as implemented, so please make sure you understand precisely what I'm saying before marking it:

At the moment you can define how textures repeat only within the UV space but there's no way to define how they repeat OUSIDE the uv space. So if you lay a polygon so that half of it is outside the UV space, it will simply use the opposite side of the image, whereas I want to use the near side of the image flipped over. There needs to be an option to define how the image is repeated OUTSIDE the uv space. Here's an image to make it clearer:

http://i77.photobucket.com/albums/j62/drain-body-confirmed/outuvidea.jpg

Please see that this is DIFFERENT to the F6 "Map Image" repeat option, since that only controls repetition inside the UV space, not outside, as I want.

This feature is absolutely vital because it is used all the time when making low-poly games (DS & PSP).

Island packer is nice feature, but it would be more useful if user could specify a margin which will be used by packer to prevent islands to be completely touching other islands.

This would help the process of baking much. At the moment you really can't use Pack Islands and use baking with margin.

Sometimes when you unwrap you get some large island with big holes where other islands could fit.
Blender's current Pack Island does not take this into consideration when packing islands. It would make Pack Island really useful if this feature could be added to Blender!

Here is two demonstrative pictures if my description it too weird.

http://etana.kapsi.fi/uv1.jpg
http://etana.kapsi.fi/uv2.jpg

When you project a photo into 3D (ie. using UVProject), you may notice that the photo gets deformed.

If you don't understand the problem, see this example: I want to map a chess board texture from a photo (example) onto one square face. Using UV projection I get this image
The correct perspective transformation can be achieved in gimp, but not very comfortably: image
(A solution within blender is to subdivide the face many times)

The problem happens because blender calculates UV based on tris, not quads. I don't think that it's more CPU consumptive to use a quad-based interpolation. If so, it could be optionally turned off.

This feature useful if you would edit the mesh after made a uv set and recalculate the uv when you editing mesh and this way remain the uv in same position and dont need remapping. i.e. create a low poly mesh from high poly with remove loops, merge vertex etc etc and the uv will be same and dont destroyed :) .

Multiple UV island selection using the L KEY isn't consistent with the L KEY select in the 3d view. In the 3D view you don't need to press Shift+L to select more than "one element". You only need to press shift when deselecting elements. Pressing the L KEY over multiple UV islands should select them. Pressing SHIFT+L KEY to deselect.

This would be a time saver especially in UV mapping. Include some way via a simple keypress that instead of moving multiple selected vertices to an averaged point will place each vertex to the values you type in for X and Y.

It would be nice to have a notifier showing whether or not Proportional Editing is on or off in the UV window.

When an image is UV project by a camera on a mesh, it kind of "pass through" the faces and some parts can be displayed serial times.
If the goal is to project a picture, that would be nicer if the picture stops on the face (like a real light emited by a projector)
I made a screen capture to clarify:
http://picasaweb.google.com/braverdy/Blender/photo?authkey=8ll8Mc48wWU#52002670 44751118786

?in 3d max, UV unwrapping you can move vertexs freely without affecting or selecting other vertices,(just like now in blender) however when you select one, it becomes yellow, and highlights the shared vertexs blue .. but doesn't affect them when moving the first selection.

The ability to map (UV) one image to one set of normals of an object, and then map another image to the other set of normals.

For example: this would allow for "double sided baking" - where one side of a wall (single plane width of 0) is lit but the other side is in shadow. This feature would allow both sides to be baked onto separate images - one for the "positive" normals, and one for the "negative" normals.

Now we have the great tool "shrinkwrap" that allows us to retopo and keep the precision sculpting.
But we cannot yet get back the texture/mesh colors from the previous mesh to the new one.
That would not be too difficult to implement, is it?
E.g. using shrinkwrap, when the mesh goes to "the closest surface point", can we associate to it also the vertex color of that point?

Let's say you have a cube and want to paint the texture in Blender and then end up with an .obj and a .tga that is a beautiful lightmap of that cube with the painted texture.

It should be possible to jump into UV-paint-mode with a keystroke or a menu selection, and then bake the texture with good default settings by clicking the "Bake" button somewhere.

Blender has the capability for doing this, but I wish the workflow was smoother.

Ability to extend the UV Editor's Grid beyond the standard 0 to 1 space.