MH new GUI, BRAINSTORMING here

[SuperCGpro]

Does this mean you will be able to have multiple characters in a scene?

[Manuel]

Does this mean you will be able to have multiple characters in a scene?

Yes, it's possible, but I'm thinking to multiple mesh (clothes, accessories and hairs) instead multiple characters.
Have multiple characters is not so useful...MH is not a final composer...

[xpundx]

This is a concept in progress

This "J-widget" or something like it could replace the joint rotation flat selection panel . As the j-widget is pulled over the mesh it would realize the joint and resize and attach itself for that specific joint allowing rotation as defined by the constraints to that joint. The joint could then be articulated by the controllers in 3D space. When the tool is finished being used, it could then be docked in the toolbar area. This I think would allow for very exact positioning, save space, and be a clearer way of visualizing the rotations.

j-widget01.PNG (53.57 KiB) Viewed 8088 times

j-widget02.PNG (65.45 KiB) Viewed 8086 times

j-widget03.PNG (57 KiB) Viewed 8083 times

What do you think?

[Manuel]

This is a concept in progress

This "J-widget" or something like it could replace the joint rotation flat selection panel . As the j-widget is pulled over the mesh it would realize the joint and resize and attach itself for that specific joint allowing rotation as defined by the constraints to that joint. The joint could then be articulated by the controllers in 3D space. When the tool is finished being used, it could then be docked in the toolbar area. This I think would allow for very exact positioning, save space, and be a clearer way of visualizing the rotations.

What do you think?

Not bad.
But there there is a problem: j-widget" require to be morphed with the limb, and this can be a little complex to do.

The same thing can be done with a more simple widget, like this common one, but of course,
the j-widget above is more scenographic

[xpundx]

I agree, simpler is better. The widget would still have to have knowledge of the joint in its current rotation. It would have to resize and rotate itself to the camera position. It needs to be easy for the user to understand its use.

I think in terms of mechanical aparatus because of my background. I know other people visualize differently.

r-widget01.PNG (44.42 KiB) Viewed 8038 times

Tim

[pistacja]

That looks good. When moving/rotating things with the gizmo the axis/dimension in which we are moving/rotating should be somehow highlighted and above the gizmo a number saying how much did we change and what is the current value of the joint (eg. if it was 15 and we changed it to 40 it should display "40/+25"). There could be also some visual help to show the amount of change we just made, for example this gizmo in it's initial position would become grey or semi-transparent when we move/rotate and then fade or just disappear when we are finished rotating/moving.

[clausawits]

I am having trouble understanding the scope of what people are talking about. The first few posts I read had me quite scared that MH would go spinning off into deep corners of whacko-land... I pictured menus that were actually multi-sided objects:
a) I'm looking at the edit menu, I don't see the blah button!
b) Which side are you looking at? Look at the bottom!
a) The bottom row of buttons is (...)
b) No! grab the bottom edge of the menu, and rotate it upwards so you can seem the menu options on the bottom of the cube. If you click on blah, the cube will transform into an octohedron, and if you rotate that to the upper left three faces, you'll find the "smile" options.
a) ...

If we are talking about something like the joint thing, ok.. but how do you interact with it? Wouldn't there still need to be a 2d slider somewhere to grab, or to type text into? If you do the joint thing, one of the most important things you could do, would be to use it to make it clear which rotation is which between the joint and the place you are actually entering what you want.

I could be lame, but I fear that making people learn Esperanto to read the instructions to use a product is a good way to drive people away.

Must be that I don't have a fertile atmosphere in my head for this brainstorm yet. A fully 3d GUI when you only have 1D or 2D input methods can make for frustration.

Hmm.. think of some positive contribution... how much trouble would it be to replace the static pictures with slowly spinning, animated representations of the action being applied from 0 to 1 strength on the neutral mesh? This should not require a lot of manual effort to generate the pictures (the animations could be batch generated automatically), but it seems it would take some extra computation or disk space...

[xpundx]

The widjets would probably exsist in some sort of tool bar. The tool would be selected with a mouse click on the tool bar then applied to a joint on the mesh by clicking on the joint itself. A widjet like those shown in the examples would appear.
The rotation widjet for example would be controlled by clicking on one of the rivets holding the rings together. There should always be two visible from any point of view on the screen. Each one of these control spots can rotate the rings (and the joint of the mesh) on two axis. Between the two of them, they would give you control of all three axis if that particular joint allowed rotation in all three. It would be interpreted from the mouse movements on the screen. Click and drag.

[pistacja]

I just took a look at the blog and I think I like the "Different solution" better.

I have just a small idea... You'd only ever modify one side of the mesh and then maybe click a "symmetry" button to make the other side exactly the same. You'd have a great preview of what you are actually doing since you can just compare the right and left side. There could be a "invert controls" to change the side you are currently modifying to make some asymmetrical meshes.

[Pete]

I think the new GUI approach has potential. My biggest concern with it would be usability - whether it's getting complicated to the point that it interferes with workflow. Hard to say until people have it in their hands. One thing I think would be useful is if the parts of the mesh not being worked on were visible. The concept on the blog with the nose for example: you'd want to see how the changes you're making affect the rest of the face. However, it would probably get in the way of using the widget. So what about normal mode for the part being worked on and "X-ray" mode for the rest of the model?