Action Node

Messages
30
Action Node

I thought a new topic would be better...

Just goofing around, a few beers taken, what do you guys think about this?

Action%20Nodes%20001.png


Based on the unbelievably easy to use materials node system.

Am I crazy or would this make our lives a hell of a lot better?

I think we might call it.... ACTION NODES

So you can see from the action time line a series of plugin nodes, so you can combine several poses in one sub action. The first column for example involves two poses. The you add a frame count for the transition from the previous sub action to the current action. So moving from a relaxed pose to a look left pose could be quick or gentle depending on the frame count added.

Frame counts would be prioritised according to hierarchy, so those at the top of a segment would be the total count, those lower would default to begin at the start of a segment, unless you added a frame delay value. Some thoughts here?

Once the action is finished we could then just drop it onto the time line. (As I mentioned previously, I think Hype handles this quite intuitively.) So if we want to tweak a character action its not the major ordeal like it is now.

Also, it would be cool if you could add actions to an action, so in this sequence, the guy looks around, raises his arm (to get a pass from another player), then shuffles his feet and darts off (using the run action). The direction of the run is controlled by a simple spline (called "Run1" in the example). If no splines are added to the actions, the character is static, just carrying out the pose.

Perfect. 5 minutes work, versus 2 hours.

Any thoughts on this guys?

Martin, do you think any of this would be viable?

By the way, the inspiration for this is thanks to Karpckie, an excellent Polish beer. And thats from an Irish man who was raised on stout. Ok thats it, I'm off to watch some TV with my beloved.
 
Last edited:
It's a nice dream. Personally I'd love to see something akin to messiah's expressions integrated into this node editor. I love the simplicity in the design of cheetah, but it definitely is lacking power at the moment. If power could be plugged into the simplicity we'd see a lot more use of Cheetah.
 
Seems like most of what you want could far more easily be achieved by adding sliders to the Pose tag (the way the morph tag has them). You could explore takes (or whatever) for the pose you want, save it, then when you wanted to interpolate between poses simply drag a few sliders around.

I would like something like Blender's NLA system (which lets you treat the equivalent of "takes" like film segments and layer them, crop them, compress them, blend them, and so forth in an Avid/Premiere/Final Cut kind of UI).

I just don't see what all this complexity is supposed to be achieving. In Blender's NLA system you'd just layer animations (takes?) on top of each other and they'd blend as appropriate. Want someone to wave while jogging? Overlay a wave segment on top of a jogging segment.

Not only is this a more intuitive way to handle multiple animations as reusable components, I think it also would map onto C3D's existing functionality much better.
 
Last edited:
yeah, hibernian's method was complicated. i didnt get it at all;)

Hmmm, maybe. I did say I was having a few beers didn't I?

How about this, I've been using Gantt charts like this for years in Project Management, is easy to follow now?

Action%20Nodes%20001.png


So the little slider bar at the bottom controls the frame count. An output at the end of each segment can be linked to the start of another segment to make it start in sequence. We could control alteration to frame rates dynamically this way.

I must check out the Blender system, but I already use C4D and the prospect of learning yet another program is quite frankly depressing.
 
Last edited:
I think you're obsessed with nodes — they're not helping.

I'll forget it so. Only trying to help.

Saying somebody is obsessed is considered rude where I come from. Sorry to have wasted your time.

It was the 900'th wish list post, I thought that would deserve a prize at least.
 
Last edited:
If you work in creative endeavors — 3d graphics, programming, and user interface design all qualify — I'd suggest you deal with criticism in some other manner than to snidely accuse others of being rude.

If you want to wish for something, I'd suggest reducing it to the essential thing you want and why you want it before jumping head-first into designing the user-interface.

Professional UI designers use "wireframes" and "lo fi prototypes" to mock up UI concepts quickly and cheaply, so that they don't over-invest in ideas too soon, and so that people they're asking for feedback from don't think they're looking at something more advanced than it actually is.

What do you want to do? I think you want to blend poses. OK so that's what you should wish for. Now, if you have some ideas on how to do that, that's potentially useful extra information. Instead it's "look, here's this amazingly complex UI I want you to add to your product" and when some of us gently push back you come up with something even more complicated.

As a side note, any UI design inspired by Gantt charts is off to a bad start, in my opinion.
 
If you work in creative endeavors — 3d graphics, programming, and user interface design all qualify — I'd suggest you deal with criticism in some other manner than to snidely accuse others of being rude.

Accusing somebody of being obsessed is rude. Disguising your single line quip as creative and constructive criticism does not wash. Perhaps it's just a cultural misunderstanding.

I am an Architect with my own practice, employees, office, set of Pantone markers etc, and well respected as such. I sketch and design for a living, the UI mock up was, I though easier to follow than a long list of cryptic descriptive text.

A picture paints a thousand words. It can be almost impossible to articulate an idea without a sketch, mock up, wireframe whatever. Would you not agree?

Despite your feelings, that mockup actually took some effort, I did it in an attempt to clarify how I think poses could be put together in a (I thought) simple but powerful manner.

If you want to wish for something, I'd suggest reducing it to the essential thing you want and why you want it before jumping head-first into designing the user-interface.

Professional UI designers use "wireframes" and "lo fi prototypes" to mock up UI concepts quickly and cheaply, so that they don't over-invest in ideas too soon, and so that people they're asking for feedback from don't think they're looking at something more advanced than it actually is.

Now you're being rude and patronising.

I would never say "I want", I would appreciate however a more powerful and flexible way of stringing poses together. I still haven't looked at the Blender suggestion you made, but as I mentioned, I already use C4D, ArchiCAD, Artlantis, AutoCAD, Pixelmator, PS, AE and several other programs. Learning another powerful but complicated UI is not an option I would consider lightly.

What do you want to do? I think you want to blend poses. OK so that's what you should wish for. Now, if you have some ideas on how to do that, that's potentially useful extra information. Instead it's "look, here's this amazingly complex UI I want you to add to your product" and when some of us gently push back you come up with something even more complicated.

Back to rude again.

As a side note, any UI design inspired by Gantt charts is off to a bad start, in my opinion.

If you don't understand their function, that could easily slant your opinion.

I do respect your opinion, not your coarse attitude. If I criticised your book regarding poor grammar, lack of clarity would you be receptive? (These things aside it is a good book, and I have positively reviewed it as such.)

My point being, relax. Life is too short my friend.
 
If you have instances of bad grammar or lack of clarity in my book please do not hesitate to forward them to me.

A picture paints a thousand words.

Of course.

I am not suggesting that you do not furnish sketched ideas, I am proposing that you use simpler sketches which show less investment rather than going straight to a complex and elaborate case and a complex and elaborate sketch that looks more developed than perhaps it is.

Accusing somebody of being obsessed is rude.

There are two things about forum conversations that tend to lead to bad feelings and "flame wars". One is anonymity — which I do not think is at work here — and the other is lack of social cues such as tone of voice.
 
If you have instances of bad grammar or lack of clarity in my book please do not hesitate to forward them to me.

As I said, you're book is very good, I have it on my Nexus at all time when I'm using C3D, its a hand reference.

I am not suggesting that you do not furnish sketched ideas, I am proposing that you use simpler sketches which show less investment rather than going straight to a complex and elaborate case and a complex and elaborate sketch that looks more developed than perhaps it is.
Thats the Architect in me, a sketch sometimes is not far off a working drawing...


There are two things about forum conversations that tend to lead to bad feelings and "flame wars". One is anonymity — which I do not think is at work here — and the other is lack of social cues such as tone of voice.

Agreed, I think we're both too long in the tooth for a flame war about nothing. I'm happy to leave it at that. (Picture a jovial Irish accent).

I just looked at the NLA thing, you know what, it looks like a gantt chart :smile: Seriously though, did you get a chance to look at Tumult Hype yet? I think if we had an animation system like that C3D would be mind blowing. Thats what I was hinting at, but with the POWER OF NODES...

All hail Nodes. :icon_thumbup:
 
In my experience when you build anything you really care about, the first time is just practice :)

I think Martin probably needs to build the functionality first and put a minimal UI on top of it so that it works, and in doing this will learn enough about the problem to produce a really good solution (which is, I think, kind of what the node-based material system represents).

There's a trap here too — you can spend a lot of time refining a UI and get left behind in functionality. I think there's a bit of that in the node-based shader system — we got a great system for manipulating the old shader functionality (and there's no question it can do things the old shader system couldn't do) but we haven't had any real new functionality (except, I think, for anisotropic reflections) in a long time.

I played with Hype shortly after it came out and wasn't impressed, but it may have improved since then. I've used a lot of animation software over the years, so I'm pretty jaded :)
 
Back
Top