Essential Maya Tools

As part of an ongoing series, I humbly present Lightgirl.

This lighting GUI, built in MEL, is absolutely insane and one hell of a lifesaver if you have a scene with more than, uh, one light in it. It was built by Ben Wronsky, who I don't know personally but is apparently a Senior Technical Artist over at EA Games. Looking at this, I wish I was a Senior Technical Artist but I have to realize that for all my many gifts, I just ain't smart enough to pull off something this rad.

It's available both on his site & on highend3d.com. It's a must have if you care at all about being able to easily control the lights in your scenes. I found it while looking for a tool that would allow me to change the intensity attribute of a cluster of four lights. Without a tool like this, it's a matter of selecting one, changing it, then selecting all of the others and repeating - one at a time. This makes for lots of steps and lots of preview renders if you're trying to find a setting that looks "just about right."

Bah! All that crap is behind me now. The one caveat is that the UI isn't perfect in the OSX environment, but that's what I get for using Maya on a Mac.

Cheers.

Tricksy Camera, version 01

So all the futzing and time wasting and random code typing is done for now. I've got a camera rig that I'm going to live with for a little while, see if my half-baked ideas for what would make a workable set up make any sense at all.

Here's a copy of the Maya scene file (Maya 8) and the script. To try it out, import the camera rig file into any Maya scene, position it roughly, then run the script.

(screenshot of the rig and the GUI)

Now that this is in a workable state, I'm moving onto the animatic for the film, just blocking out shots and figuring out rough timing. I'll probably have stills to show as I go along but not much very exciting as models I'll be using in it are going to be very, very basic.

Cheers.

Maya Tip of the Day #2

This one is a little MEL script that's far from finished but that's gonna make it into my camera GUI in one form or another.

window -title "test Field Slider";
showWindow;

polyCube;
float $currentXLocation = `getAttr pCube1.tx`;
columnLayout;
attrFieldSliderGrp -min ( $currentXLocation - 15 ) -max ( $currentXLocation + 15 ) -at pCube1.tx;

This creates a slider that slides the cube on the X-axis +/- 15 units based on wherever the cube is when the script is run. Wow. Now that is...uh...what exactly is that?

The way I see it working is as follows:
The camera and aim are blocked in roughly using the 4-pane orthographic & perspective views. Once the basic shots are set up, I'm going to need to refine them while looking through the camera. It's at that point that I'll call up the GUI and there I'll have sliders and keyframe buttons to dial in the actual shots. Because the transforms are locked on the actual camera in my rig (to avoid gimbal lock and throwing the camera totally out of whack) you can't move the camera using the usual pan & dolly mouse controls.

This rig is taking me about nine years and seventeen days to build but an end is in sight. I think.

Cheers.

MEL Script = Headache

Ugh. I just spent the last several hours puzzling over how to do a bunch of different things for my camera rig. I've come the conclusion that a lot of the things that I put in my original rig & UI aren't really necessary any longer, particularly having two Aim Constrained locators and animation sliders for both of their weights.

I liked the idea at first. My plan was to aim the main locator at the main target of the camera and then have the secondary locator pulling it off target, bringing a little life to the camera. But if you think about how many possible animation tracks there would be in the graph editor with that kind of set-up, it gets ugly. Also that way of aiming the camera has no real relationship to the way an actual camera works. Not that a CG camera has to mimic a real world camera but I should probably have a decent reason for making it drastically different. Plus there are plenty of ways to bring life to the camera that would be more direct.

Alright. I'm tired. That's it for me. A copy of the rig will be up here eventually.

Moving Onto The Animatic

Looks like I'm on a weekly update schedule, which I'm not very proud of but it's a schedule at least.

So, the brief is in the hands of an artist I've worked with before. He's giving it a quick pass for clarity and to assess the reasonableness of what I want to get versus what I'm paying. It's not that I have no money to spend on an artist but it's more along the lines of, I'm offering one day's day rate and asking for, probably, more than one day's worth of work. Nevertheless, it's very important to me that I am offering paid work. All of the time in this industry you see "offers" for people to do free work because it'll "look great on their reel/portfolio." Now, I've done free, low-budget and work-in-trade but on this thing, I don't want to go that way. If an artist I already had a relationship with offered to do the concept art in trade I'd probably go for it, but I'm not going to approach someone I don't know hoping that they're in the mood to throw me some charity. This film is, for me, a real project and I intend to treat anyone who works for me on it professionally.

Enough time on the soapbox. One place I am planning on going for some free help is the CG Society forums. Now that the storyboard is considerably tighter (though probably not "done") I feel like it's time to get the animatic cranking. Earlier, before our change of direction, we brushed past the animatic but I think that this time it's important to make one. It'll help us really suss out our shots, it'll suggest which shots are just plane lame and can be cut and it'll give us an idea of whether or not the structure of the scenes makes any sense. In order to work on the animatic, I returned to look at the camera rig I was building a few months ago. That's what I'm going to be asking for some help with. I'm pretty happy with that I've built but it's not working exactly as intended. I'm hoping someone up there can throw some pointers my way.

I'll post screens and the MEL script once it's in a place I'm happy with.

Cheers.

Broken Cam

It's no longer being called "TricksyCam." It's now "BrokenCam."

I've not been testing the changes I'm making to the script as throughly as I should've been. Now, several steps on down the path, all sorts of transform controls are conflicting with each other.

The major issue is that the motion path is messing with the pivot points for Roll, Tilt and Pan.
The pivots are staying at 0,0,0 - the place where the camera was created. When the camera's offset by the motion path, the pivots should leave 0,0,0 and go wherever the camera is. Centering the pivots doesn't work/change a thing.

This has to be a problem plenty of other people have dealt with, but it's incredibly frustrating to meet right now, this far into the process.

The GUI saga continues

This week has been just as busy as the last, meaning work on this thing is lots of little fits and starts and very, very little sustained attention. I'm really ready to get this GUI nailed down and move onto the next part of the shot, especially as I did have a little bit of a breakthrough while doing sketches for the central clocktower. Hopefully I'll have the chance to scan my drawings and post them up before too long. I wouldn't call them final concept art or anything, but they've got me excited about the tower. I'd been feeling lost on it for a while now.

As ready as I am to move past the camera GUI, one thing about it is really nice, and it's that I'm taking the time to build this thing correctly, right from the start. One of the difficult things about working in production, on "one-off" sort of spots, is that so many of the elements we build feel like hacks. They're built just for one shot, or to be used at only one angle - fake 3d stuff, for example. These hacks are fine most of the time, but they come back to bite you when the client wants your 2d ribbon, faked to look 3d, to wrap around the star of the show, or push in a completely different direction. Had it been 3d in the first place, no sweat, but it was built in 2d by the designer and used that way in animation.

Often enough to be frustrating, the hacks used because someone felt like there wasn't enough time to do it all-the-way, end up costing the same amount of time on the backend.

All that's a long way of saying that it's fun to put the time into this GUI right at the start and build it a bit more fully than it really needs to be for the shots we have in mind. I'll try to post a new screen of it soon, but this weekend looks very rough in terms of free time.

Too, Too, Too Busy

Work on this project has nearly, but not entirely, ground to a slow burn. Not to a "halt." I didn't say "halt" and I'm not going to.

I'm still in GUI mode and I'm somewhat embarrassed to show how little progress I've made in the last three days as it only represents about 3 hours worth of work. Nevertheless, I need to stay in the habit of posting here, so here's the updated GUI.


There's what, three extra controls on there?

The Wander on the camera is a really nice effect that I grabbed from Richard Morris over at The Jackal's Forge. He put up a great tutorial on building a very detailed, multi-variable camera shake utility. The Wander I'm using here is actually what he calls "Destabilize." Of the four shaking functions he describes, this is my favorite. It gives the camera a very natural-feeling sway that you don't really notice until you turn it off.

Moving on from here I'm going to add a couple of sliders to control the frequency and intensity of a more traditional "jitter" camera shake. That's not a ton of work, assuming that I can get a block of time to do it in, and from there it's back into the scene proper, using this rig to nail down the camera move.

Cheers.

I'm a Freakin' UI Designer

This week has not been the most productive week ever, but I did get busy trying to make a good control rig for our camera. It's been pretty fun and my crash course in MEL scripting is just impressing the hell out of me in terms of how usable Maya can be once you put in the time to figure it out.

The above rig is as simple as they get but it does have a check box that suggests that I may, in fact, be awesome. I'd check it.

This GUI and rig are, arguably, overkill for what we need the camera to do in this shot but this is an opportunity for me to learn how to make these things and make them useful. So, some overkill is just fine in the long run. It's all helping me in the end.

Crazy busy weekend coming up, but my goal is to make a much fancier GUI and then move on to dialing in the camera move.