Andy Serkis on animators


#21

Interesting comparison, but really if you think about What Serkis does, that’s essentially someone playing a keyboard synthesizing a violin. You could argue it’s better than someone manually inputting each note for synthesis, but it is still synthetic and almost always goes through a cleanup phase. It’s still not the same as a live performance on the real instrument.

But these amazing moments that people like Pyke mention, suggest (not stating I know peoples experience) they haven’t been either on set or in the editing room. I’ve watched actors do 30 takes on a scene, and the little nuances are often not there in each shot. A great actor can not only reproduce those nuances but also make them “seem” natural although intentional. But A lot of it is also pure luck, something that suits the character and moment well, but isn’t in every shot, it might have been an experiment, it might have simply happened that way. It’s like thinking that every single goal a hockey player scores is a perfectly planned goal, not a single one was a surprise tip in, or that one instant distraction to the goalie or a knee jerk reaction when the puck flew in front of the net. You give them far to much credit for those nuances.

And for the animator, he performs over and over and over, and studies plenty of reference and discusses all the same things an actor does. The difference is that he can refine or tweak instead of doing more and more takes until he and the director feel they have one that works. Nothing sucks more than being in the edit room when the director loves an actor’s performance but some nasty light glare or extra screws up the scene and they have to choose a less ideal performance because it’s the only choice they’ve got.

Not saying what actors do is nothing important or remarkable, but motion capture performance, at least to date, is extremely reliant on a team to make it work at all. I mean motion capture is only as good as the modelers, riggers that make it a believable character visually, and the animators who tweak refine, redo etc. If you ask me, Davey Jones was far superior to Caeser. And Hal hickel does a pretty good job of acknowledging the Animators as “actors” in indicating the importance of their contribution on the Davey jones character.
http://www.youtube.com/watch?v=H3jnx3mXhVo (near the end 3:50)


#22

I agree that Andy did not say anything against animators yet he didn’t say anything to support their work, I mean after all CGI departments are still credited below catering stuff, drivers and electricians.


#23

It’s hard to figure out if you are being deliberately provocative, or somewhat sarcastic within an odd context :slight_smile:


#24

Could not have said it better!

While in school we saw quite a bit of raw mo-cap data and that data was no where near ready to be put on the screen.

To me Serkis is down playing the role the numerous animators have on the final look and feel of his characters. And that is quite a low ball move in my opinion.

It is also pretty easy to compare against… find the raw mo cap data and the final tweaked animation and compare how much of an influence the animators have over the animation. The evidence is there, why take someone’s word for it when you can just compare and contrast the actual work.

His comments are pretty far reaching as well. The general public is going to look at what he says and continue in the belief that CG is easy. That CG is basically pushing a button and out pops the good stuff!

It is irresponsible and just plain wrong!


#25

That’s the most logical and well constructed answer, which represents my thoughts on the subject matter.

If there ever has to be an award for a special category it should be one that celebrates the joined collaboration of animators / actors / and mocap artists into bringing a digital character on film. It’s not a one man show and it never will be, unless technology leaps into a new era of automated performance capture that prints directly on film, were no human input is required for a believable act.

Until that day, Andy should give more credit to people that like himself, put their hearts and souls in their work (in a creative and work ethic manner). With no disrespect to Andy’s face, because in my opinion he does a phenomenal job all these years, but recognition belongs to more people than perhaps he realizes, or in his effort to gain his own, treats poorly.

I have friends who worked with Andy on game projects and had nothing of a superb experience with the lad, as a person and as an actor… and I’d like to think that most animators like Andy.
But he is trying to get justified the wrong way on this matter, if you ask me…

Just my 2c :beer:


#26

I’m not saying that the animators aren’t important, because they are obviously are. I’m also not saying Andy is an amazing actor. I think most good actors with a strong physicality can do what he does. He’s a pioneer and a public face for that style of filmmaking, but not necessarily some kind of untouchable genius.

And to answer Raff’s question: no, high-end VFX was never my industry–I’m primarily a games guy–a 2D artist who dabbled in 3D. But I’m not questioning whether the animators are important, or whether the raw capture was 100% usable, or how much cleaning up and key-framing needed to be done.

My logic goes like this:

Before mocap/facial-cap was invented, we had some fine animation. We all grew up on similar stuff, so we can all point to the classics and make our case. I don’t think anyone can dispute that. But in the realm of live-action, I don’t recall seeing any complex, nuanced, natural, and realistic digital character animation before mocap/facial-cap was invented. Or maybe there were, and because they were so realistic, I just didn’t know it wasn’t real. But AFAIK, the really amazing stuff we’re now being spoiled by are directly because of the advancements in mocap/faicial-cap. If we had achieved the same level of realism/nuance with just key-frame animation, where are the stunning examples before mocap/facial-cap became popular? Can you guys point out some examples (preferably humanoid characters, displaying realistic and complex layers of emotions)? I can’t think of any off the top of my head.

Somebody might point out that Gollum’s animation had parts that were totally key-framed. I have no idea because I haven’t seen any behind the scenes videos pointing out which parts were totally key-framed. If any of you worked on Gollum, I’d love to know which parts were key-framed.

In the end, we all know that actors and animators must work together to achieve the kind of results we see today. Actors without animators will not result in a digital character, and animators without actors will have to key-frame until the cows come up to match the realism/nuance/complexity of a good mocap/facial-cap performance. Even if they need to clean it up and tweak it, it’s still based on the performance.

Of course, animators can just mocap/facial-cap themselves and say screw the actors; then they can take all the credit. :smiley: That’s probably more likely than actors learning to become proficient in VFX. But it’s not like acting doesn’t require any talent at all, so you can’t just put any person in a mocap rig and expect brilliance.


#27

I would argue that the improvements in CG animation have developed in tandem with the use of performance capture. It’s a bit like saying “I haven’t seen any decent CG lighting before HDRI - therefore all good CG lighting must be HDRI”

I can’t link to any examples of Gollum as I heard it first hand from people working on those shots but another good example is Davy Jones

http://www.cgsociety.org/index.php/CGSFeatures/CGSFeatureSpecial/a_treasure_chest_of_techniques

From the article : “Animators worked with mocap data for body performances in many shots but hand animated all facial expressions and lip synch”


#28

they used this for promotion?

isnt this proof that its not his performance in the movie?


#29

Facial motion capture has only recently (think Avatar and Ben Button) reached a stage where it’s become more of usable visual effects technique - and even then the data is massaged to varying degrees - travelling from Motion Editing (where the data is filtered, edited, transformed, scaled, rotated) to Animation (where it may be scraped) and then the occasional Shot-Sculpt (where the baked animation is refined even further) before it is handed to Lighting for rendering.

Prior to that facial animation was primarily done by “hand” using the actors performance as reference - Gollum, King Kong and Davy Jones all being examples of this. Using the actors performance as reference didn’t necessarily mean that the reference was followed - much like a director requesting a change from the actor onset, a director could very well ask the animator to change the performance to better suit the character/film.

Andy Serkis may have well have been the main driving force behind the character’s performance on set, but so was Tom Hanks in A Polar Express and it’s quite easy to see how dull and lifeless a virtual character can become on screen - even with a fairly excellent actor (in the case of Tom Hanks) driving the performance.

The problem with giving actors Best Oscars for artificial characters is that it’s too hard to tell how good the acting is on it’s own right. If you strip away the art and science that goes into a virtual character and presented just the actors performance - would that performance still hold up?

For an example of complex, nuanced animation not using the actor as reference - (from what I can gather) Frank Oz only provided the voice of Yoda in the last two Star Wars films - the body and facial performance of the character came entirely from CG animators.

As for his live-action roles I thought Serkis was pretty good in Einstein and Eddington - he also got excellent reviews in Sex and Drugs and Rock and Roll - although I’ve yet to see the film, so I can’t really comment on his performance.


#30

It’s been said in multiple articles covering the process, Gollum’s hands, feet, and face were entirely hand animated. So all the nuances and incredible acting in his face that people credit to Serkis, were animated by hand.

Now of course you can say it was still his performance the animators were referencing, but then you’re ignoring all the decisions and changes the animator made to improve it, or the fact that any animation is almost always based off reference either done by the animator or someone else.

Look, the discussion on who did what, or who is more important, etc. could go on forever, and the longer it goes on the more black and white the sides get. The overall point I think most animators are trying to make, is Serkis, stop playing it off like you are Caesar, you are Gollum, you are King Kong, and that you deserve an award for that performance. There are still too many people adding their touch to the performance, strictly movement wise, for one person to receive all the credit. (As so plainly shown in that side by side image of Serkis and Ceasar)

All other discussion is really besides the original point. Actors and animators are different. Mocap has helped the industry. Gollum wouldn’t be the same without Serkis. I agree! Still doesn’t change the original point.


#31

Even when he was wearing the suits, the nuanced facial expressions, eye movements, ear twitches, etc were done by other professionals.


#32

Somebody might point out that Gollum’s animation had parts that were totally key-framed. I have no idea because I haven’t seen any behind the scenes videos pointing out which parts were totally key-framed. If any of you worked on Gollum, I’d love to know which parts were key-framed.

Gollum was practically all keyframe because there was no reliable facial mocap at all in 2002.
The face of Davy Jones was all keyframe too (well the tentacles were partially simulated).

The change is in how much more realistic the facial deformations have gotten, and also the rendering part (SSS shaders for more believable skin). This was mostly the result of using detailed models and hand crafted blendshapes and looking into scientific research (FACS).
Faster computer hardware with better realtime feedback has also been an important factor. Avatar and Apes character heads are like 20-30 thousand polygons, that takes a strong machine to work with.

What mocap provides is a shorter turnaround on the first stage of the animation workflow. The rest is mostly PR talk clouding up the actual processes behind the results.


#33

QFT my friend, QFT.


#34

i agree with this.


#35

Not sure i agree. Sure if Andy just went in the mocap room all by himself did the clips and came back out again and said to the director:
“You now have everything you need-my job is done here!”
And it were true.

But he’s still under the direction of at least the film’s Director. I don’t think thats a lot different than an animator pleasing an art director. Most of the productions i’ve worked on
a single performance is handled by a single animator (unless they aren’t available for a retake).


#36

“People say, “You provided the movements, the emotional background for the role.” No! I played the role. If you give a bad performance, you can never make it great, no matter how much you layer and texture it after the fact. It will never be anything more than the original performance.”

I was on board until I read this.

Douche…


#37

PLING

'Nuff said.


#38

Ahh sweeping generalisations. Two can play that game Mr Serkis…

Created by Andy Serkis

Created by Artists


#39

DaveKW, he never said he modeled and rendered anything, it is all about the animation merits here. Your comparison of stills don´t prove anything, besides the fact that you didn´t get the point.


#40

Well… to simplify what I’m trying to say… if I was a Director… and I was talking to an animator I would be saying:

“The character has to have his shoulders back like this.”

However, I think if I was talking to a guy in the mo-cap suit I’d be saying:

“You have to have your shoulders back like this.” So the level of internalization is not the same.

BUT… That is only true while he is wearing the suit. And in reality once he provides the data, he is “technically discarded” (which would be an afront to Andy if he didn’t take this the right way, but it’s true). The capture data then becomes the total domain of the Animation Team.

Again, that picture in Time Magazine is really telling. It is clear that the CAESAR Performance (not the Serkis performance) is the one being lauded in Planet of the Apes because they are not the same performance point-by-point.

So that is to say that Andy Serkis did ACT (which is not animating) and many aspects of his ACTING provided reference for the final result even if in the capture process it didn’t appear instantly but instead required animators to approximate it instead. The manner in which his performance is used is different of course to say, the manner in which the Johnny Depp performance was used in Rango.

However, what is up there on screen is actually the CAESAR Performance (ie: with Lip Position changes hehe), which at its final level is not really that different in final implementation to the RANGO Performance.