11.15.2007
quarter final WIP
I really really really would have strongly loved to start this post with "I present..." but evidently that is not the case. I'll get to do that soon enough, but for right now here is the final WIP of the effect shot as of the end of Fall Quarter. I'm not going to let this project take up an inch of my Studio II space, so I plan to crank the finishing touches out during the first week of break or so.

Anyway, the WIP:



You can also view the QuickTime here (20MB)

The fluids suck, and that is going to be the primary focus of work along with putting together the final breakdown. So on the upside, there isn't much work left to be done. But on the downside, there's work left to be done.





11.08.2007
up to speed


I have a working comp of the final effect. But first, some history.

So last time I posted, I left you with a test render of the fluids-based smoke trail. The step after that was to get a sweet looking explosion to rip the wing off. For this I consulted the Gnomon Fluid Pyrotechnics DVD. I created a simple explosion using the steps from the DVD as a foundation, and then tweaked the attributes to suit my needs, i.e. a tighter more controlled explosion, high speed wind blowing it back, etc. etc. Stuff specific of a jet engine explosion.



Next up on the list: The Shoot.

Yes, I finally got around to doing the shot. My cameraman, Josh Feezer, had trouble locking down a steadicam from the film department but was able to pull off a reasonably stable shot for the effect, not to mention trackable.

The weather and lighting for the shot could seriously have not been better. It was about noon on a beautifully crisp Friday, barely a cloud in the sky, and the sun was in the perfect position.



I had given some thought to who the actor/actress for the short should be; everytime I would talk to someone I know, I'd stop and imagine how they'd look in the scenario. Eventually I decided a friend of mine, Sam Rossignol, would be the best call for the spot. Sam was reluctant to do it at first because "she can't act" but fortunately for her its not hard to act like you're walking down a sidewalk. I also humbly explained to her that she didn't have to do it, but if she said no I was just going to talk her down over it until she agreed.

Ultimately, I'm sure it could have gone to anyone and worked just the same, but Sam was a good choice. What a good sport.



In the above video, it shows the actual plate shot footage and the tracked shot as a playblast from Maya. I used Boujou to do the tracking for the shot, and its a lucky thing I did. I hadn't taken a single class or read a single thing about tracking before this project, despite knowing that the project would definitely entail tracking a full shot. I figure the shot isn't insanely complex, although it doesn't include an element that will remain in the picture for the entire shot (the camera points to the sky and then to the other side of the street) but I also just figured I could just learn what I had to learn to complete the shot. Boujou was perfect for this. After reading up and doing several test tracks on old footage and different takes of the shot, I managed to get a rather successful track out of the fourth take.

So. After getting the track, and getting the elements rendered out, its time for the final stretch, which is the comp. I put the elements together and comped them into the footage using Shake, which was a sort of simple endeavor but still required a bit of work and still will, as I've got a list of fixes and tweaks a mile long that I'm planning to do to both the comp and the render. There's also a minor breakdown to show the separate elements of the shot.



You can also see a better quality version in QuickTime here:

The shot also comes in HD 720, as the entire final project will, but I think I'll wait for the final project before I start throwing HD rez around.

Like I said, theres still a huge list of things I plan to fix. These things include:

- Compositing tweaks: There's a shift in exposure and focus at the beginning of the shot I plan to compensate for, and I could do a much better job on color correcting the elements into the shot. There are also some masking tweaks with the trees and the horizon to do.

- Effect tweaks:
The smoke and fire look terrible! I think realistically the smoke would not be as plumey and thick for that long, and probably start dissipating a bit sooner. I'm gonna play some opacity changes, and try to bring out the fire and hotspots in the engine burn more. The explosion is nice, but it still looks a bit bizarre as it doesn't really look like its moving at high speed like it should. Not to mention on the list of stupid idiot mistakes, the one of the pieces of debris geometry I was instancing onto particles is gleefully hanging out in space by the broken stub of the wing as the plane goes down. Doh.

- Filming things:
The whole plane doesn't look like its moving THAT fast either, really. This might be something I'm just gonna have to deal with, as a reshoot is pretty out of the question at this point, but I think I can fix things by working with the animation. Also, I have to shoot the rest of the short!

That about wraps it up at this point, its getting reeeeeally close, and with one week left in the quarter, the full thing will come to closure here in a few days. I'm excited. Lets do this.





about portfolio extra links contact