Everyone’s talking about it so it would be rude not to post that the new trailer for Rockstar’s (Sydney-based team Bondi’s) L.A. Noir dropped yesterday and the facial performances look fantastic. Using Depth Analysis’ MotionScan technology, they appear to be going all out to capture the performances of apparently over 200 characters.
While the visual benefits are obvious, I can’t help but wonder just how much it cost the production for so many actor contracts when not only voices but likenesses are required. Additionally, the production task must be one of the hardest parts of the game, considering that at the time of writing they’re still hiring senior roles for the cinematics team around 6 months before the planned release date.
That said, I’ve really been keeping an eye on this since E3 so can’t wait for spring to roll around, plus I’m a Noir nut since a Humphrey Bogart stint last summer.
Read more about the MotionScan tech and process here.
UPDATE: They’re not still hiring as the front-page job posting on their site is dated 2008. Makes me feel better about the frequency of my own posts.
[...] This post was mentioned on Twitter by Jennifer Bourion and Jennifer Bourion, Jonathan Cooper. Jonathan Cooper said: New blog post: L.A. Noire – MotionScan Performance Capture http://www.gameanim.com/2010/12/17/l-a-noire-motionscan-performance-capture/ [...]
WOW!
It’s pretty impressive what they guys did here, but what’s the real difference between the depth analysis and the universal capture(ucap) technology, that was used for matrix II and the tiger woods 10 golf game?
You have a unique actor, who’s performance and only his is captured for animation. There’s nothing about retargeting.
If it’s an animated mesh with changing details in the textures it would be interesting to know how they build the processing/streaming of the data in realtime.
Chris
@Chris – That’s the first thing I thought of when hearing of this also. I went for a talk on the uCap method of facial video playback at MIGS a couple of years ago and I don’t believe they ever managed to get the compression to an acceptable level so it was never used in realtime beyond the Tiger Woods demo and some pre-rendered Need For Speed videos as far as I know.
You can see the before and after versions of the facial compression in this movie and they still look great despite losing some resolution. Perhaps that’s the only difference, not forgetting the clearly advanced depth-capture process no longer requiring spraying the actors’ faces with that pesky UV paint.
@Jonathan – The textures look very blurry. Maybe this has to do with the performance of the realtime algorithm that blends the different texture sets for each facial feature. Moderate textures sizes seem to work better than big ones.
Or the camera resolution is to small and it doesn’t produce higher quality regarding from the distance between camera and actor in the capture volume.
Have you ever worked with something like this for MassEfect or DragonAge at Bioware?