AVATAR and performance capture: is it, isnt it?

[brianct]

confused?...youre not alone....
this fellow complains the problem lies with the eyes....maybe the next step is an eyeball camera!


'Actors with faces but without voices can’t be nominated for Oscars. Actors with voices but without faces can, but don’t get nominated for Oscars.

Ingrid Thulin couldn’t have been nominated for The Four Horsemen of the Apocalypse because she was dubbed by Angela Lansbury. The same goes for the widely praised Philippe Noiret, dubbed into Italian in Cinema Paradiso, or Mel Gibson, dubbed into American English in Mad Max. Robin Williams could have been nominated for his Genie in Disney’s animated 1992 Aladdin, but wasn’t. Nor was Meryl Streep or George Clooney nominated for their work in Wes Anderson’s Fantastic Mr. Fox.

The rationale seems to be that if you have a voice but no face or a face but no voice, you’re giving only half a performance. (Dubbing singing voices, however, is acceptable. Even so, Audrey Hepburn didn’t get a best actress nomination for My Fair Lady in 1964 chiefly because it’s Marni Nixon’s voice you hear coming out of Hepburn’s mouth.)

Now, what about performance capture? You have a voice, a face, and computer drawing. How much of that is acting? How much of that is animation?

Avatar director James Cameron has stated more than once that performance capture, especially as done in his 3D sci-fi blockbuster, is real acting. Actors emote on an empty soundstage; visual effects people draw on top of them while preserving "every nuance of the performance," as explained in an Associated Press piece about this issue.

"It is a performance. It hasn’t been animated on top of that. That to me was a big thing, that they don’t enhance my performance in any way," remarked Sam Worthington, who goes from human to Na’vi during the course of Avatar. "Whatever we did does translate exactly 100 percent. Maybe my nose is animated and my tail. That’s because I don’t have a tail. And the ears are a bit different, but those are about the only things they’ve changed."
etc

AVATAR: Performance Capture and the Oscars

[PaTRioT1987]

"When I see Zoe Saldana’s eye expressiveness in Avatar, I don’t know how much of that is a result of Saldana’s performance and how much is a result of the computer animators."

This person should look at some of the performance capture footage of Sam and Zoe (split screen) when they are acting out the 'I Trusted You' Scene.. and then see how much of it is Zoe's performance or not!

[palulukanmakto]

This person should look at some of the performance capture footage of Sam and Zoe (split screen) when they are acting out the 'I Trusted You' Scene.. and then see how much of it is Zoe's performance or not!

Agree. Zoe was awesome.
But things are changing. I think we'll see such nominees in the not so distant future.

[Dusso Janladde]

Here's what they should have done as a tech demo to show what performance capture is capable of...

1. Create human (not avatar) CGI models of Jake, Grace, and Norm.

2. Choose a scene where only these characters appear, all as humans. A good example would be the morning after Jake mates with Neytiri, where Jake is about to get into the link, Norm is balancing in Jake's wheelchair and showing him how to run the link, then Grace comes in and shuts down the link to feed Jake his breakfast.

3. Shoot the scene twice. Once live, and once using performance capture.

4. Use the aforementioned human CGI models (not the avatar models) to show the captured performances.

5. Show the live and CGI versions of the scene side-by-side. If performance capture is all it's cracked up to be, they should be identical in performance terms. It's a given that the CGI models wouldn't look completely identical to the actual actors, but it would be very close.

[Olo'eyctan]

"When I see Zoe Saldana’s eye expressiveness in Avatar, I don’t know how much of that is a result of Saldana’s performance and how much is a result of the computer animators."

This person should look at some of the performance capture footage of Sam and Zoe (split screen) when they are acting out the 'I Trusted You' Scene.. and then see how much of it is Zoe's performance or not!

Yes. Plus in some of the other behind the scenes videos and stills you can see how while the cast was in Hawaii, she had to act some parts out. there is split screen footage for Sam too. Sam and Zoe both did a great job and my personal opinion was that both of them deserved nominations.

[Olo'eyctan]

Here's what they should have done as a tech demo to show what performance capture is capable of...

1. Create human (not avatar) CGI models of Jake, Grace, and Norm.

2. Choose a scene where only these characters appear, all as humans. A good example would be the morning after Jake mates with Neytiri, where Jake is about to get into the link, Norm is balancing in Jake's wheelchair and showing him how to run the link, then Grace comes in and shuts down the link to feed Jake his breakfast.

3. Shoot the scene twice. Once live, and once using performance capture.

4. Use the aforementioned human CGI models (not the avatar models) to show the captured performances.

5. Show the live and CGI versions of the scene side-by-side. If performance capture is all it's cracked up to be, they should be identical in performance terms. It's a given that the CGI models wouldn't look completely identical to the actual actors, but it would be very close.

That would have been a great idea. If anybody can find a scene that was filmed like this, pease post a link. Also, you can tell that some of the human-action scenes are at least partly CGI because Sam's legs aren't really atrophied and useless like human Jake's are.

[Dusso Janladde]

That would have been a great idea. If anybody can find a scene that was filmed like this, pease post a link. Also, you can tell that some of the human-action scenes are at least partly CGI because Sam's legs aren't really atrophied and useless like human Jake's are.

I really doubt that they did, as the cost of creating photorealistic CGI models of all three of them just for a tech demo was probably too high.

[PaTRioT1987]

Here's what they should have done as a tech demo to show what performance capture is capable of...

1. Create human (not avatar) CGI models of Jake, Grace, and Norm.

2. Choose a scene where only these characters appear, all as humans. A good example would be the morning after Jake mates with Neytiri, where Jake is about to get into the link, Norm is balancing in Jake's wheelchair and showing him how to run the link, then Grace comes in and shuts down the link to feed Jake his breakfast.

3. Shoot the scene twice. Once live, and once using performance capture.

4. Use the aforementioned human CGI models (not the avatar models) to show the captured performances.

5. Show the live and CGI versions of the scene side-by-side. If performance capture is all it's cracked up to be, they should be identical in performance terms. It's a given that the CGI models wouldn't look completely identical to the actual actors, but it would be very close.

That would have been an awesome idea for them to do.. but as you said, very pricey just to prove how realistic it is

[Treefriend]

Honestly, look at any of the major Na'vi or Avatar faces and tell me that's not a real person. It doesn't look animated at all to me.

[Human No More]

Zoe and Sam each deserved to win. Unfortunately, the oscars are stuck in the past.