PDA

View Full Version : Best practices for a good 3D Next Gen Interface...Tell us your ideas.



robertoortiz
03-31-2019, 05:39 PM
As we all know the LW interface needs a lot of love. A LOT OF IT.
Frankly I feel strongly that things that as thing stand now, things that should not take a long time, like setting up BASIC shaders, looking for an object, doing BASIC poses to pose key frame animation or hell setting up a animation curve, take too much time and frankly are not intuitive for users fresh off the boat.

So lets imagine we can start from scratch...

And instead of playing catch up, or trying to be something lw cannot be(Houdini Lite or Maya lite), why not try to aim in a different, maybe better direction.

So tell us:
* What do you consider to be some of the best practices you expect to see in a MODERN NEXT GEN app?
* What concepts would be more useful to follow? More power and Accessibility or a more cleaner interface?.

Tell us what you thinks, and maybe show us some examples.

thanks,
-R

AnimeJoex
03-31-2019, 09:53 PM
Ease of use is the most important thing to me. After the trying ordeal that I experienced with Martin Hash's 3D Animation, I swore up and down that I would never waste my time with software that tried my patience like that. Which is why I really like LightWave.

OlaHaldor
04-01-2019, 12:00 AM
Sorry, I know how hard this is to many LightWave users, but having used several other 3D apps, I find it hard to go back to two separate applications. So my first point would be a unified application.

Next up is pie menus. Whether it's like Modo, Maya, or something else. Pie menus are far easier and faster to use than list menus for quick swapping between tools and features.
And that brings me to... Context sensitive menus. Right-click a vertex gives you access to the tools that can be used on a vertex, and so on for other primitives. Being able to chose something that does not work with given primitive, only to get an error obstructs the flow.

I also think customization of menus/tabs/viewports should be among top prio. And make it easy for the user to make their own panels, menus and so on.

Icons! I never thought I'd say it, but icons are actually not as bad, to a certain extent. Starting with LightWave as a teenager I thought LightWave looked cleaner and was less confusing than Maya, Max etc. which was bombarded with icons. But guess what; icons makes a lot of sense if you let them. For those who struggle with the concept; make it possible to disable icons to get that nostalgic LW text only look. Everybody wins.
Not all tools and menus need icons. Some are just fine as text buttons.

A modern viewport mode (think Blender or Marmoset Toolbag etc)

In general, refined tools which gives a more pleasing user experience. Better feedback in the UI that you are in fact hovering over a primitive for instance.

That's what I could think of at the top of my head (late to work already, gotta run! maybe I'll come back later to fill in some more)

Ztreem
04-01-2019, 02:02 AM
Ease of use, some said. A unified environment is just that, it's easy and fast to work with. Besides the need for unification, Lw need to update all things to be more interactive.
We can't have things like this.
144638

gar26lw
04-01-2019, 04:33 AM
yeah. it’s old skool in a bad way

gdkeast
04-01-2019, 05:00 AM
Since LW is the only 3D program I've ever used, except for a couple of tries at Blender, I am not really unhappy with it. Some of the tools are more complex than others, but overall, I can generally find what I need or figure things out. One of the strengths I see in LW is its ability to be customized to each users' specific needs. Anything that makes customization easier I think it would be a plus. For instance, I recently got OD Workspaces and love it. It seems to me that that type of functionality would be good to have native in LW. Giving users the ability to control their workspace would seem to be a good direction for LW and for me, anything that gives the users more control is an advancement overall. TBH, I'm so used to the programs being separated, it doesn't really bother me, but I understand why people like integration, but then again, I've seen people using multiple programs to accomplish things, then pulling them together in another program. I just watched a tutorial series on 3D scanning, and the guy used 5 different programs.

RPSchmidt
04-01-2019, 06:52 AM
I think the single biggest upgrade they need to make is to make all of the windows dockable.

Personally, I would like to be able to dock the image editor and the node editor to the surface editor so that when I open the surface editor, those items are also already open.

I would also like to dock the scene and graph editors together.

There are many other items I would like to have the ability to dock together for ease of use.

Right now, that's about the only thing that slows me down... opening multiple windows and their habit (and my habit) of closing them to reduce clutter.

If I was going to make one single other change, I would make the graph, scene, and item windows poll for changes more often. It's a pain to add an object to Layout and the item window doesn't update with the item. You have to close it and open it again.

Anyway... most of this has been said before.

Marander
04-01-2019, 09:04 AM
LightWave development needs an architect/developer with strong and up-to-date UI/UX skills. It's obvious to me that there is currently no one with this knowledge and dedication in the development team, otherwise new tools would not look like they do in 2015/2018/2019.

On the other hand I don't think the LW Modeler and Layout applications allow the implementation of state-of-the-art UI features. Adding tools on top is not the solution.

In LW there is currently no customization possible (except some colors and menu entries) and I have to use it as it is. I purchased all the OD tools (including OD Uberpass, OD Workspaces and OD Pie menu) as well as the new db&w QuickPanel but that only makes it a bit more acceptable to work with.

In my other main 3D application I can work the way it fits my workflow best, modify or place any UI element where I want it, even simulate the grey-in-grey Text-Only LW interface if it has to be, because it's fully customizable and supports individual workspaces. So there is no benefit in the LW UI towards others in my opinion.

I learned to accept the way LW is and don't expect drastic UI or workflow improvements anymore, however appreciate the small additions like Undo (finally, even if still not complete), CTRL+Space and the drop down filtering in 2019.

OK here you go (note that all this would just be catching up since that's part of other applications since long):

- Node editor to expose / propagate parameters into the (layer based) Uber Material
- Other node editor workflow improvements
- Context-sensitive and dockable property panels
- Get rid of the countless and ineffective drop down menus
- Consolidate duplicate tools
- Implement (optional) industry standard navigation
- Non-destructive workflow
- Collapsible / expandable sections in property panels
- Duplicating, locking / unlocking property panels
- Setting as new root, arranging / docking / undocking of property panels and menus, custom tabs
- Free choice where to use icons, text or both
- Adding tools, parameters and other HUD elements to the viewport
- Workspaces
- Easy filter for UI elements, for example for hiding bones, grid or other viewport items
- Get rid of all the floating panels and sub dialog boxes
- Correct the fixed-width dialog boxes with truncated text in the tabs
- Find a way to get rid of all those More... sub menus
- Find a way to consolidate the numerous panels: Scene Editor, Surface Editor, Image Editor, Motion Options, IK Booster, Backdrop, Legacy Volumetrics, Pixel Filter, Image Filter, Volumentrics in Render panel, Object properties, Light properties, Camera properties, Node Editor, FiberFX, Dynamics, several Option panels and so on...
- Better integration of 3rd party and internal plugins
- Context sensitive Attribute Manager / Properties dialog
- A proper Preset system, the current one is incredible poor in its implementation and default content
- New viewport option (to create a new quad / single / camera / perspective / ortho viewport and drag it anywhere, independent from others)
- Change existing UI elements from input boxes to sliders or radio buttons etc.
- Highlight new tools
- Multi scene support
- Take system
- Layer system in Layout
- Multi parameter select options (drag over parameters like XYZ to mark and change them at once)
- All parameters animatable (including on/off states) with a simple keyframe click, not only a few via the E button

When I watch LW tutorials, I see users spending a lot of time moving around dialog boxes because they're often all over the UI and in front of the viewport. On the other hand, I cannot even have the properties of two separate objects open at the same time. Because all of this, LW is a very "clicky" application and feels like a giant patchwork of tiny tools to me.

gar26lw
04-01-2019, 09:25 AM
100% agreement on ui/ux.
these threads are good but i think i’d like to see lwg initiate them so that there is the smallest chance something will actually happen.

Marander
04-01-2019, 10:07 AM
... and if it belongs to the UI topic..

- Workplanes
- Snapping, Guidelines, Measure tools
- Onion Skinning, Animation Layers
- Better Wireframe Antialiasing
- Better, unified and hideable Gizmos
- Improved OpenGL
- Improved Sliders
- Better IDE / Python editor
- Better Graph, Dope Sheet and Scene Editors
- Modifier (Shift/Control/Alt) keys to Copy items / Quantize translation etc.
- Spline preview (realtime preview how splines looks like while modifying <-- ok for that LW would need proper Spline support in the first place)
- Better realtime Dynamics and non-destructive Fracture
- User friendly openVDB workflow
- Better UI representation of Primitives in Shaded and Wireframe mode
- Realtime parametric (Boolean, Mesh Generators, Cloners etc)
- Simple brushes to correct geometry in Layout (without needing to add Metamorphic or other modifiers), in order to quickly correct intersecting polys like cloth, smoothing problematic areas etc.
- Modifier / Deformer OpenGL preview
- Displacement viewport tesselation
- Ability to display / connect / disconnect Graph Editor Curves automatically to Scene Editor and/or Objects in Viewport (clicking on one of these automatically displays the corresponding Curves or Objects in both ways)
- Proper tangent controls in Graph Editor

Marander
04-01-2019, 10:37 AM
...One of the strengths I see in LW is its ability to be customized to each users' specific needs. ...For instance, I recently got OD Workspaces and love it. It seems to me that that type of functionality would be good to have native in LW...

That's one of the biggest weaknesses in LW, its lack of customizability. Workspaces offer much more in other applications than what OD WS does (the same for OD Uberpass), if that would be the native implementation it would be typical LightWave style but not state of the art.

hrgiger
04-01-2019, 11:58 AM
pretty much everything olahaldor said.

erikals
04-01-2019, 02:58 PM
Sorry, I know how hard this is to many LightWave users, but having used several other 3D apps, I find it hard to go back to two separate applications. So my first point would be a unified application.

Next up is pie menus. Whether it's like Modo, Maya, or something else. Pie menus are far easier and faster to use than list menus for quick swapping between tools and features.
And that brings me to... Context sensitive menus. Right-click a vertex gives you access to the tools that can be used on a vertex, and so on for other primitives. Being able to chose something that does not work with given primitive, only to get an error obstructs the flow.

I also think customization of menus/tabs/viewports should be among top prio. And make it easy for the user to make their own panels, menus and so on.

Icons! I never thought I'd say it, but icons are actually not as bad, to a certain extent. Starting with LightWave as a teenager I thought LightWave looked cleaner and was less confusing than Maya, Max etc. which was bombarded with icons. But guess what; icons makes a lot of sense if you let them. For those who struggle with the concept; make it possible to disable icons to get that nostalgic LW text only look. Everybody wins.
Not all tools and menus need icons. Some are just fine as text buttons.

A modern viewport mode (think Blender or Marmoset Toolbag etc)

In general, refined tools which gives a more pleasing user experience. Better feedback in the UI that you are in fact hovering over a primitive for instance.

That's what I could think of at the top of my head (late to work already, gotta run! maybe I'll come back later to fill in some more)

Excellent! :king: https://i.imgur.com/fMb2Mjm.gif

COBRASoft
04-01-2019, 05:45 PM
Start with the base... Dockable panels, non-modal windows, non-destructive workflow, interactive tools, fluent transitions between different stages (model, anim, ...).

In short: UX, UX, UX!

rdolishny
04-01-2019, 07:18 PM
After today's news you can post your wish list in any Blender or Maya forum.

OFF
04-01-2019, 10:57 PM
User friendly openVDB workflow - by and large everything rests on very slow calculations, as a result, almost any operations with VDB lead to an extremely slow playback. I specifically watched a number of videos on YouTube related to the construction of VDB geometry in other 3D programs - all operations are done literally on the fly, you do not need to wait for your result to be drawn for 10 or more seconds, even on the simplest scenes. VDB tools are absolutely necessary part of modern tools in the field of 3D visualizations. It is good that it is in LW. But the level of implementation is very similar to the good old Newtek traditions - when the in principle availability of a tool is more important, but its development leaves much to be desired.

But if you follow this tradition - LW will never come close to the level of its competitors. Everyone tries to provide if not the best solutions, then at least convenient for work. I can not call the current implementation of VDB in LW convenient - first of all, precisely in the matter of speed of miscalculations.
In addition, all of yesterday I observed surrealistic situations - when one or another simple function fell off one by one while I was working with VDB and a character in my scene - various menu elements disappeared, the ability to select several items disappeared, hot keys ceased to function, etc. etc. As a result, I now had to completely reinstall LW 2019.03. But I'm not sure that this will not happen again.

https://www.youtube.com/watch?v=KvvmHJYbwjQ
https://www.youtube.com/watch?v=-CQf8caAEOU
https://youtu.be/R8Y9VqU_YTI?t=90

robertoortiz
04-02-2019, 05:39 AM
After today's news you can post your wish list in any Blender or Maya forum. Besides the Newtek sale did anything else happen?

- - - Updated - - -


After today's news you can post your wish list in any Blender or Maya forum. Besides the Newtek sale did anything else happen?


- by and large everything rests on very slow calculations, as a result, almost any operations with VDB lead to an extremely slow playback. I specifically watched a number of videos on YouTube related to the construction of VDB geometry in other 3D programs - all operations are done literally on the fly, you do not need to wait for your result to be drawn for 10 or more seconds, even on the simplest scenes. VDB tools are absolutely necessary part of modern tools in the field of 3D visualizations. It is good that it is in LW. But the level of implementation is very similar to the good old Newtek traditions - when the in principle availability of a tool is more important, but its development leaves much to be desired.

What would you consider to be the best implementation? Houdini maybe?

OFF
04-02-2019, 06:11 AM
As I have already said, it’s not so important that the level of convenience was at the highest level (yes, Houdini is probably the leader here for many years) - it’s more important that the end user is just comfortable. But when you can't even play the simplest animation on the fly and you have to start recording a preview every time for this - this is not the most convenient option.
Maybe this particular problem could be solved by multithreading of calculations (most of us have powerful multi-core processors, but only one thread is involved here - this is just awful). But the current incarnation of VDB is just a hint of a good tool.

Chris S. (Fez)
04-02-2019, 08:20 AM
After today's news you can post your wish list in any Blender or Maya forum.

Actually, this seems like a great fit. Companies change hands all the time. Vizrt is all in on Unreal and Lightwave 2019 integration is excellent.

robertoortiz
04-02-2019, 11:29 AM
Actually, this seems like a great fit. Companies change hands all the time. Vizrt is all in on Unreal and Lightwave 2019 integration is excellent.
Thanks for confirming something I suspected...Unreal support is THE future.

erikals
04-02-2019, 01:34 PM
Thanks for confirming something I suspected...Unreal support is THE future.

+10

ConjureBunny
04-03-2019, 03:57 PM
**Watching this thread to shamelessly steal ideas and later try to pass them off as my own.**

vncnt
04-04-2019, 04:25 AM
- Zoom/Pan for Camera viewport, similar to Fusion viewports. To monitor small details as they appear in the final output.

- OpenGL Preview viewport mode, without Overlay/Wireframe/Grid/Handles/Icons/MotionPath/IKchains/SubPatchCages/SplineTargets/FogCircles/TargetLines. For instant Preview that resembles the final output more closely.

- Separate resizable synchronized viewport. To be used (permanently) on a 2nd monitor.

- Multi-track editing, including sound tracks.

Elmar Moelzer
04-04-2019, 12:13 PM
I would really like to see dockable windows, which I think is one of the biggest issues with LWs UI.
Other than that, I think workspaces are a good idea. LW sort of has the basics for that already.
Another good idea is to have the ability to dock and undock viewports. That is particularly useful for second screens.
Other than that, I like a clean and well structured interface. Controls well aligned and logically clustered, things orderly and spaced evenly. Thankfully, LW already does a lot of these things.

slv
04-04-2019, 01:03 PM
to unify the modeler and the layout is essential and will allow other few possibilities like the recording of a sequence of modeling for example.

LWcad is an excellent tool: complete need integration in lightwave

painting and integration realist ecosystem
painting bump, or displace as a sculptor
Complete integration of Octane render
better FX tools for fluids
better interactivity with objects and natural forces (wind gravity etc) to easily animate the trees according to the strength of the wind.


Unreal bridge integration is great, go even further.

vncnt
04-05-2019, 05:07 AM
Im not sure about "dockable windows".
In applications like Vegas Pro I still keep fighting them and every Vegas project looks different because of that flexibility.

I think Im more into "slightly magnetic windows".
Windows that can be placed next to other windows, with lined-up title bars but without annoying gaps.

We should keep LWs shortcuts to access specific windows.

vncnt
04-05-2019, 05:13 AM
Id like to see Camera presets that are flexible / user defined / project wide.
To quickly switch quality/speed balance for rendering.

And a method to Copy/Paste the Camera settings all together.
In-application and between applications.

gar26lw
04-05-2019, 09:07 AM
Unity, Adobe Photoshop, Maya, bit of Max, some Modo and a little MS Visual Studio.

hypersuperduper
04-05-2019, 09:33 AM
A bunch of UI stuff plucked from various software isn’t really next gen though. It’s just current gen. Next gen should look farther ahead. Like do we need windows and panels at all? Would it be feasible to do EVERYTHING in the viewport somehow. Could one with perhaps envision a system where all of the connections and hierarchies and what not that make a scene could be viewed directly in the viewport like ephemeral threads and lines connecting the world together. Kind of Like nodes but taken to the next level by removing an entire level of abstraction. View and edit all animation curves directly in the viewport for example. That sort of stuff is what seems to me to be NEXT gen. Like when Zbrush and sculpting appeared and turned modeling in its head. That was next gen. Dockable windows isn’t.

gar26lw
04-05-2019, 10:21 AM
A bunch of UI stuff plucked from various software isn’t really next gen though. It’s just current gen. Next gen should look farther ahead. Like do we need windows and panels at all? Would it be feasible to do EVERYTHING in the viewport somehow. Could one with perhaps envision a system where all of the connections and hierarchies and what not that make a scene could be viewed directly in the viewport like ephemeral threads and lines connecting the world together. Kind of Like nodes but taken to the next level by removing an entire level of abstraction. View and edit all animation curves directly in the viewport for example. That sort of stuff is what seems to me to be NEXT gen. Like when Zbrush and sculpting appeared and turned modeling in its head. That was next gen. Dockable windows isn’t.

c'mon, we need a catch up from paradigms 10-20 yrs ago. Future ? Get with the present first.

hypersuperduper
04-05-2019, 10:34 AM
Don’t chase where the puck is chase where it is going to be. If lightwave spends years catching up it never will. Besides, The title of this thread is best practices for a 3D next gen interface. I take that to mean not current best practices. UI trends change constantly. The idea is to try to anticipate where they are going and cut the competition off at the pass so to speak. My examples were pretty out there, but do we really need another unification/docked window thread?

vncnt
04-05-2019, 11:00 AM
... Like when Zbrush and sculpting appeared and turned modeling in its head. That was next gen. Dockable windows isnt.
And I still dont understand the logic in Zbrush.

Maybe AI menus?

robertoortiz
04-05-2019, 12:18 PM
Dont chase where the puck is chase where it is going to be. If lightwave spends years catching up it never will. Besides, The title of this thread is best practices for a 3D next gen interface. I take that to mean not current best practices. UI trends change constantly. The idea is to try to anticipate where they are going and cut the competition off at the pass so to speak. My examples were pretty out there, but do we really need another unification/docked window thread?

And one advantage Newtek has is that to be frank, the interface HAS TO CHANGE. It really does.
Now try selling that the Maya or Cinema 4D crowd.
We have nothing to lose by going in a bold direction. Hell we already shot the legacy content horse wit the new render engine, so why not go all the way.