Saturday, January 10, 2009

To Cap or not to Cap ? Part 1/3

That is a question I used to answer "hell no" to without any further thought. My motion capture phobia had grown over the years for all the abuse of the tool I had seen. I recently realized I had unfairly blamed the tool for the abuse of producers and artists in search of a quick bypass to animation. I even heard some cretin once say that "Animation was on it's way out and that motion capture was taking over..." or something equally naive.
Any way I wouldn't have been caught dead near a motion capture studio until recently when circumstance compelled me to explore the medium.

After the horrendous crunch of 2006 on God of War 2 most of us swore never to go through that again. Now as the specter of crunch lurks a few pages down the calendar, I figured motion capture might be a healthy alternative to depriving the Cinematics team of sleep for months. That said, my skepticism for mocap wasn't down for the count, rightfully so, God of War must keep it's trademark animation, otherwise the fans would get angry, we wouldn't like them when they are angry.

I almost feel the presumption of a boy who just lost his virginity and is going to talk about sex ed. but our shoot went so well I want to share the new found wisdom:

Part 1/3 preCap

I) What to capture ?
Common sense: if you have kratos Grappling through the air and swinging his blades like the tazmanian devil, leave that to the animators. Dialog on the other hand, loooong scenes of dialog where characters stand, walk around each other etc... are perfect candidates for motion capture. Your actors can try multiple takes, vary the performance and you'll get the natural weight shifts, steps etc... which, although a great exercise in animation, is a worthy sacrifice when you are contemplating crunch. Save your animator's energy for the money shots that can not be motion captured !

II)Casting your actors.
Having some instinct about people helps here.
Select actors who's reel shows performances that are close to the ones they will be doing for you. If they have somewhat the same build as the character they'll be lending their performance to the results will look more natural. Focus on performance, you can mix it up and have a woman play an effeminate man or vice versa, make it fun.

A nod to our god of war actors, we had a great time and they all got along and were very generous in their performances :D

III) Scheduling.
A mistake some studios make (or so I hear) is to cram retarded amounts of scenes to capture each day. If you wear out your actors (physically and morally) on the first day, the vibe on your set will go down and the results will testify to that; Your animators will then have to reanimate the scenes thereby defeating the whole purpose of saving time.

Plan every single scene or move you need to capture.
Start on the first day with the easier scenes: those with fewest actors, no props and no environment. It'll give you a chance to ease into things. Also consider that the energy on a set fluctuates throughout the day so keep a pulse on your crew and adapt... (Don't feed your actors anything too heavy, so they don't get belly aches while performing)

Include in your plans some time to rehearse scenes, do multiple takes and a buffer for minor technical difficulties, because they will arise.

In trying to be considerate to the actors (and incidentally to your producers) schedule so that each actor has consecutive days. Some actors may travel to get to the set and you don't need actors waiting around for nothing, paying for hotel rooms etc...

Have some bonus scenes to pad the schedule, scenes that would be nice to capture but won't be missed otherwise; Those were scheduled to be animated manually anyway :p

IV) Preparation !!
Too often people show up unprepared. Preparation may prove a tedious task but getting all the minutiae pinned down will allow you to focus on the performances when capture starts and everybody will have fun. Unless you like to direct a la Stanley Kubrick, in which case you should still be prepared so you can focus on making everyone's life miserable.
Either way, be prepared.

Things to consider when you select your scenes: the environment. If you don't know where the scene takes place, or if the environment is not locked and might change later don't bother capturing the scene. If you captured your actors standing on flat ground and later your director says he wants them walking down a flight of stairs you will be sad.
Avoid scenes with overly complicated environments. If your characters need to interact with intricate objects leave it to the animators. Recreating something too specific on a motion capture stage won't save you time. Stick to things you can recreate with boxes, mattresses, rope, etc...

Have all your scenes blocked out, with audio if applicable !
The more final your audio and models are the better. If nothing is ready to go hold off on the mocap until you have solid direction to work with.
Having an animatic video of your scenes is a tremendous help for your actors to visualise the scene. When actors can see the model of the character they are impersonating it helps them understand the performance and they get pumped when the model is badass. So also include some key poses and basic acting ideas in your blockouts as a general guide.

Send your scenes to the mocap studio a few weeks before the shoot is scheduled: Let them take a look, have notes for them if you have any and let them get back to you with questions. They will probably catch something you missed, tell you if a scene is going to require special planning, safety measures or anything else you haven't anticipated.

Man this is long. I'll break it up in parts... and add pictures :D

Next up: Directing on the mocap Stage !

Wednesday, January 7, 2009

Folding a Shirt

So I am now working on normal mapping folds on the shirt for the girl below. You know, to make it look like she bought a size too small... Any way the front folds look ok but still not convincing...

The side folds look like crap. Where they start and end makes no sense, nor does the shape.
I need to study my reference a bit more attentively... Sadly I'll have to leave it at this for now, it's past my bed time.

Tuesday, January 6, 2009

Couple more Characters...

This is the first character I modeled... he was born around 2001 and went through many, many iterations. Everytime I learned something new about rigging or modeling I had to update him and the result is I hardly ever animated him. So I need to cut myself off. At this point I am happy with the way he looks.... Except that he'll need normal maps... but that's it !

Then there's also this fatty, her's more recent, simpler design but I like him. He was rigged once but I scrapped that. I swear it's the last time. He's going to need some pupils and a few normal maps too but other than that he's ready to go.
I have a few more characters at the same stage but I don't want to spill everything out right here...

Saturday, January 3, 2009

Happy New Year 2009 !!

This blog is hereby born on January 3d at 3 a.m. because I have no life.
It has been a while since I've published any of my personal work, and you might be wondering what I've been up to.
I've been casually working personal stuff throughout the past decade but never finished anything.

This break I decided to sit at home and try to wrap a few things up while keeping up with new and exciting advances in the computer graphics field such as normal mapping.I've been playing with a trial version of Mudbox 2009 and I'm absolutely buying this thing because it is sweet !! The interface is very intuitive, especially if you're used to Maya.
Here's what my first shot at detailing a leg looks like: this is a day's worth of trial and error. Anatomy is very not right and I did go a little crazy on the muscle bulging. But I believe when you start you gotta make it fun first then reel it in...
Below is the character the above leg belongs to. Man after working with Bosses on the god of war project I can't help but realise what an amateur I am when it comes to modeling...