Limitations in existing Blender features you would like to see resolved.

General Forums Blender and CG Discussions
21

Be able to edit the parameters of primitives after you move them (this is crazy), they should stay parametric until I decide otherwise (with a button or when moving a point etc).

And more, they should stay parametric even if I enter the edit mode and do nothing with them and go back to object mode.

22

+++++++++++++++++++++++++++++++++

23

What I mean is this…

Say you have a rectangular solid that is divided into a grid of squares, now let’s say you want to use the bridge tool to punch a grid of holes through that solid (as in, the equivalent of doing dozens of bridge operations).

The Bridge tool as it stands now will try to treat all those face pairs as a single loop and make overlapping faces to connect all of them in one go, what the looptools bridge does differently is see them as individual pairs and do an individual bridge for each one instead of doing one long one.

I hope that’s more understandable, the basic thing is that based on the conditions seen in the selection, the bridge tool will either choose to do a lot of smaller, individual bridges instead of one long one.

Be able to edit the parameters of primitives after you move them (this is crazy), they should stay parametric until I decide otherwise (with a button or when moving a point etc).

And more, they should stay parametric even if I enter the edit mode and do nothing with them and go back to object mode.

This can be argued as something that can be applied to all operations (ie. something like a limited history stack that resets itself when you do something that changes the topology).

24

Other small features:

  • in cycles/node editor, the possibility to keyframe directly the values (without having to go to the outliner/datablocks)
  • in cycles/node editor, the possibility to access every value in a node via a connected input value
  • @doublebishop post #11 about object pass index limitation to 32767:
    I don’t think the object pass index is stored in rendered images.
25

I’m going to mention something that I’m sure used to work but now doesn’t (or at least I can’t get it to)

I’m sure it used to be that when I clicked on an image texture node the UV/image editor would update automatically and show that image. Now it does not do so.

so I would like that back please. (If it was there before, and I’m not going mad…:wink: )

26

I`m certain that it did do that before .68 or .67 and I too miss this feature. Takes too much time now if you have to adjust textures for bigger projects.

27

Glad to know that I’m not going mad. Sad that it is now gone. Is it a bug does anyone know?

28

How could I forgot bout normals editing! For game art its tremendously serious aspect. I mean vertex normals of course.

29

1.) I will echo the cries for real grid snap. I really, really like Maya’s method. If I remember correctly, I believe you just hold down x and whatever you’re moving will snap to grid increments. Hold down v and you can snap to vertices. Hold down c and you can have something follow an edge. I would adore seeing something like that employed.

2.) The style of artwork I do relies heavily on GPU rendering so I’m quite biased but I would really like to see a lot of focus on GPU rendering, lighting, and materials. This is really, really important to me as an artist, perhaps more important than anything else. I can’t even properly explain how revolutionary GPU rendering has been to me and how Blender, specifically, has become one of my most important, indispensable tools. I need very accurate lights, different material options, and good, fast, realistic rendering. Blender is a good GPU renderer (and it’s free so yay!) but I would like to see it become one of the best GPU renderers. I want to see it compete with or even outperform other quality render Engines like Vray and Mentalray. I especially want SSS for GPU.

30

For me blender need only few improvment in some area for improving workflow

  1. Edit multiple objects as some time
  2. A more usable and functional rich outliner. A good improvment in outliner (with grouping object without parenting like photoshop layers) will make (finally) obsolete Layers
  3. Improving gizmo. Actually for moving or scaling in between two axis we need to press shift+the excluding axis, cannot do using the handles, this is fine when the object/vertex/faces is in global coordinate, but you must press twice the key for normal or local transform orientation (and local and normal orientation axis is not the some of globlal, so in many cases you need to test what is the right axis to exclude), so in my opinion working with manipulator in some cases is a better option. in this video a proposal for improving manipulator (min 2.27 ):

what happened to this project?

31

Multilayer exrs store object passes?

32

Modes for the propotionnal editing, surface and volume.

33

You are right, I understand the limitation. (in fact very powerful for compositing to isolate an object).

34

Two probably little features:

When doing remove doubles should be nice to have a option, checkbox to let merge only vertices non connected (by edges) between them.
An example where i need to merge only vertices not connected between them:

In cycles can be interesting to have a “negative” holdout material, that is hiding object in front of it and not in the back of it. This probably can be done with osl materials but the mayority of us don’t know how to code them.

Then should be nice to see improvement in the cloth sim, the self collision part especially; particles, etc…

35

we already have that…

36
  • possibility and shortcut to jump directly in the middle of 2 keyframes, like in nuke, would be cool for animators.
  • display of grease pencil keyframes in the timeline & possibility to see the opengl-scene-render of grease-pencil-lines in the VSE.

shouldn’t be that hard to accomplish and would boost up some workflows.

37
  • ability to have influence over “most” modifiers, slider from 0-1
  • decimate modifier can use texture not only vertex groups to affect.

there´s so much stuf, but mostly for me I would like to see modifiers become node based, and have way more inputs. great for 3d motiongraphics pieces.

*** already possible!

38

Indeed, my bad !!

39

Array / Mirror modifiers using instancing instead of generating extra polygons.

40

I second that!