• Hey, guest user. Hope you're enjoying NeoGAF! Have you considered registering for an account? Come join us and add your take to the daily discourse.

Naughty Dog animator explains ME's animation

Kremzeek

Member
We might start to expect higher quality sure, but to see MEA go backwards from what ME had already is tough to look at, let alone all the bugs and poor performing segments.

yeah i agree, it just stings that the facework in the previous ME's just looks better than this new one, despite being made years later on more powerful hardware.
 
Whatever. Witcher 3's combat has it's detractors and those that like it. Overall it's just one slice of a very large pie.

Meanwhile you look at the animations in ME and they are objectively bad.

The general animations in combat are pretty well done, and a large improvement over previous installments. I'd also say that the general locomotion of Ryder is more impressive than Geralt's. Not perse the combat animation, but just the way Ryder walks.

"The animations" in ME:A are, as a whole, not objectively bad. Undercooked, and some of the scripting clearly went wrong, though.

Point being, just because most people know better than to share that opinion in threads like these, lest they have to explain themselves, that doesn't mean that your opinion is the only one.
 

CamHostage

Member
Not an animator, but what i think this guy is saying is that by focusing on mocap you get the majority of the movements needed, so only having to touch up the eyes (Due to reading a teleprompter) is much easier to do then have to tweak every Body/Facial animation by hand.

Eh, that's sort of what he's saying, but not really. He's more saying that the systematic stuff isn't cutting it and so companies just have to bite the bullet in buying mocap equipment and suiting up actors if they want that level of performance that mocap can provide despite its drawbacks, and that if he ran the department, he'd push for as efficient and cost-effective mocap tech in this field as possible because it's qualities and drawbacks are better known right now. He saw more successes in Horizon by doing facial cap than other methods, and in his experience, "AAA story-heavy games can't skimp on the animation quality with a systematic approach alone". It's promising tech, but we're not there and may never be.

The problem with mocap versus VO is that a mocap actor looks like this:

Actor-1.jpg


All those dots. All those cameras. All that gear strapped to them. All that warehouse space that they're working in. It's a lot of stuff to pay for, and a lot of stuff to put up with. Get a great actor, get a talented director, get a space with the right layout to allow play, and you may well get a fantastic performance; it's an extension of theater, and it's full-body performing all "captured" in one. But there's a ton that can go wrong, a ton of restrictions on what the current tools can cope with, and a ton of fine detail that goes uncaptured.

If you really want the fine detail of the face to be captured, you need a dedicated facial capture rig. There are a number of ways of doing this, and some of those methods are as comfortable as this:
Actor-2.jpg

...but most are as uncomfortable as this:
Actor-3.jpg


Either way, you have a big kit of equipment, a limit of mobility for the actor (projects like HellBlade and the Apes movies take face capture out of the chair, but you still have a camera sticking out of your face and gear strapped to you,) a requirement of space, and a lot of weird tech to play with. (BTW, you talked about having to "touch up the eyes" for the actors reading off the script... it's actually a lot more work than "touch up" when you really think about how the face moves when eyes look around, and once you include those subtle movements, if the face is moving that much all the time, what's the point of fine facial capture anyway?) Mocap is a different kind of acting, and the tech will continue to improve, but there's good reason Andy Serkis keeps getting high-profile mocap jobs, because he's one of the very best in the world at this emerging new acting challenge, and not many others are in his league yet. Mocap takes different talents, and it produces different types of performances.

Now, let's look at a voice over artist at work...
Actor-vo1.jpg


Hey, that dude looks comfy! He's in the t-shirt he wore into the building, he's wearing his reading glasses, he's got a script on a music stand, he's got cans on his ears so he can hear how it's coming out, he's making wild gestures with his hands and making weird expressions with his face that may or may not have anything to do with what the character would look like once drawn, and he's doing it in a warm little wood studio. That's a guy that could rip through pages and pages of content, and the only limitation imposed on him is how long his voice holds out and how many water bottles are stocked in the back room. Sure, some animator is going to have to spend backbreaking work animating the sounds this actor is making (or some tech wizard is going to spend the same amount and more backbreaking work designing technology that can systematically re-create human expressivity,) but damn if this guy isn't free to give it his all!

...And that's why professional animator Cooper is saying that if he ran the room, he'd probably put his chips on mocap since it is paying off more than other solutions in projects he works, but even he doesn't see this route as "100% proved". The integration of systematic approaches to animation (of having syllables translated to mouth movements and of having animation routines defined for gestures and for blending animattions together and for whatever I can't even think of that they've tried in order to make a digital actor act like a human ... I'm also not an animator) hold promise, but they're not cutting it. A combination of these techniques, as well as new techniques, will be needed in order to do 100+ hour games with animation that's not slapped on Youtube with LOLs for every mistake, but so far, the only real way to get the levels of quality audiences demand today is to invest time and money.
 
He excused nothing despite the very well-detailed explanation.

Bioware had this game in development for FIVE years. It should have been ready. Clearly multiple people worked on the core functions of this game for years while demonstrating unchecked incompetency at their jobs.

I don't buy that they planned to go back and fix everything all along but just ran out of time. The game has far too many bugs for that long of a dev cycle with as large of a budget as it had.

And the fact that they released it thinking it was anywhere near acceptable for a AAA game is extremely telling.
 
Actually really cool for the guy to come out and say this. I'm not a Mass Effect fan and never was, but I can understand the simplicity of "Making animations for huge games isn't exactly an easy thing when there's so many options for the player to make".

Either way, good game or not, it's extremely stupid to think 1 single person was responsible for the bad to mediocre animations in the game. That's honestly one of the stupidest things I've ever heard.

At the end of the day, it doesn't meet expectations in the animation department, but hey, that's not even a main thing that makes a game good in my honest opinion. At the very worst, it's good for a laugh.

well there's this and of course, the wrongly identified individual was a woman which made it even worse with the online trolling.

I think the "tweeter" should have written an article for a good site instead of those tweets. as it'd make for an easier read. That said, very salient points from a professional!
 

jdstorm

Banned
Eh, that's sort of what he's saying, but not really. He's more saying that the systematic stuff isn't cutting it and so companies just have to bite the bullet in buying equipment and suiting up actors if they want that level of performance, and that if he ran the department, he'd push for as efficient and cost-effective tech in this field as possible. He saw more successes in Horizon by doing facial cap than other methods, and in his experience, "AAA story-heavy games can't skimp on the animation quality with a systematic approach alone". It's promising tech, but we're not there and may never be.

The problem with mocap versus VO is that a mocap actor looks like this:

Actor-1.jpg


All those dots. All those cameras. All that gear strapped to them. All that warehouse space that they're working in. It's a lot of stuff to pay for, and a lot of stuff to put up with. Get a great actor, get a talented director, get a space with the right layout to allow play, and you may well get a fantastic performance; it's an extension of theater, and it's full-body performing all "captured" in one. But there's a ton that can go wrong, a ton of restrictions on what the current tools can cope with, and a ton of fine detail that goes uncaptured.

If you really want the fine detail of the face to be captured, you need a dedicated facial capture rig. There are a number of ways of doing this, and some of those methods are as comfortable as this:
Actor-2.jpg

...but most are as uncomfortable as this:
Actor-3.jpg


Either way, you have a big kit of equipment, a limit of mobility for the actor (projects like HellBlade and the Apes movies take face capture out of the chair, but you still have a camera sticking out of your face and gear strapped to you,) a requirement of space, and a lot of weird tech to play with. (BTW, you talked about having to "touch up the eyes" for the actors reading off the script... it's actually a lot more work than "touch up" if you're thinking about how the face moves when it looks, and then if the face is moving that much all the time, what's the point of fine facial capture anyway?) It's a different kind of acting, and the tech will continue to improve, but Andy Serkis keeps getting mocap jobs because he's one of the very best in the world at this emerging new acting challenge.

Now, let's look at a voice over artist at work...
Actor-vo1.jpg


Hey, that dude looks comfy! He's in the t-shirt he wore into the building, he's wearing his reading glasses, he's got a script on a music stand, he's got cans on his ears so he can hear how it's coming out, he's making wild gestures with his hands and making weird expressions with his face that may or may not have anything to do with what the character would look like once drawn, and he's doing it in a warm little wood studio. That's a guy that could rip through pages and pages of content, and the only limitation imposed on him is how long his voice holds out and how many water bottles are stocked in the back room. Sure, some animator is going to have to spend backbreaking work animating the sounds this actor is making (or some tech wizard is going to spend the same amount and more backbreaking work designing technology that can systematically re-create human expressivity,) but damn if this guy isn't free to give it his all!

...And that's why professional animator Cooper is saying that if he ran the room, he'd probably put his chips on mocap since it is paying off more than other solutions in projects he works, but even he doesn't see this route as "100% proved", and that the only real way to get the levels of quality audiences demand today is to invest time and money.

Thanks for the detailed post. I really appreciate you taking the time and effort to explain this part of the process.
 

Outrun

Member
He excused nothing despite the very well-detailed explanation.

Bioware had this game in development for FIVE years. It should have been ready. Clearly multiple people worked on the core functions of this game for years while demonstrating unchecked incompetency at their jobs.

I don't buy that they planned to go back and fix everything all along but just ran out of time. The game has far too many bugs for that long of a dev cycle with as large of a budget as it had.

And the fact that they released it thinking it was anywhere near acceptable for a AAA game is extremely telling.

It is the first game in a while that I have felt profound disappointment. I am returning it tomorrow. I never thought I would say that about a ME game.
 
I'm playing ME1 for the first time this week (played 2 and 3 already) and it's made me want to play the new one more.

ME1 has got some wonky stuff it in. Of course it's a much older game and you would hope to see the progression shown in 2 and 3 continue but I'm really enjoying it and I don't think I care as much about the animations in Andromeda now.
 

CamHostage

Member
Thanks for the detailed post. I really appreciate you taking the time and effort to explain this part of the process.

To be clear, I'm not an animator either, just an animation fan -- anybody who has experience, please kick my ass and add some more info where I'm not on it.

But I did do a double-take when I first saw this post, because before clicking, I psyched myself up too much. "Oh, awesome, a working animator from a studio on the bleeding edge, here to give insight on what cool animation tech is out there but why it can still let down even a massive project like Mass Effect!" that's what I had in my head, and Jonathan Cooper did hit those points. But unfortunately, he's not as hyped in the future of systematic solutions as I was wishing he would be (though the way they're doing sequencing and blending is pretty rad, and I'm not sold on mocap solving everything but even if I'm on Team Systematic, I can't deny that mocap is truly space-age coolness as well.) No matter what, he's saying if you really, really want lifelike quality, you'll have to work for it.

He excused nothing despite the very well-detailed explanation.

That's not what he's here to do. He didn't work on the game, he doesn't speak for the team that did work on it. He's just giving insight into his analysis of its animation system and the problems that have been identified in its presentation.
 

Kinyou

Member
I'm playing ME1 for the first time this week (played 2 and 3 already) and it's made me want to play the new one more.

ME1 has got some wonky stuff it in. Of course it's a much older game and you would hope to see the progression shown in 2 and 3 continue but I'm really enjoying it and I don't think I care as much about the animations in Andromeda now.
I played ME3 again before I tried Andromeda and despite the awkward animations it still felt like a big jump, so you should be good.
 

HariKari

Member
There's other obvious, obnoxious shit in this game like your entire arm clipping through what is presumably a plate of glass when you're talking to a prisoner on the Nexus. There's no way nobody noticed that. They just didn't care to fix it, or didn't have time.
 
Actually really cool for the guy to come out and say this. I'm not a Mass Effect fan and never was, but I can understand the simplicity of "Making animations for huge games isn't exactly an easy thing when there's so many options for the player to make".

Either way, good game or not, it's extremely stupid to think 1 single person was responsible for the bad to mediocre animations in the game. That's honestly one of the stupidest things I've ever heard.

At the end of the day, it doesn't meet expectations in the animation department, but hey, that's not even a main thing that makes a game good in my honest opinion. At the very worst, it's good for a laugh.

This world needs more people like you. Very well said.
 
Basically. The VA is straight up bad, matched by equally bad writing, who the hell says my face is tired in conversation? It's like the writer is being self aware of the bad animation. In fact the animation is the only excusable thing as it is.

I cannot stop laughing at that "my face is tired" line lol. It's like you said - it's almost as if that line is breaking the fourth wall rofl.


"Sorry, my face is tired."

- said no one, ever (until ME:A)
 
i don't think anyone is comparing ME:A to UC4...hands down the pinnacle of animation in games at the moment......

if anything else, they are comparing ME:A to ME1 - 3.........which doesn't suffer from animation problem walking up a stairs.
 

FLAguy954

Junior Member
None of that explains why characters couldn't even walk properly in scripted cutscenes, which are just as controlled and contained as in Uncharted.

This. None of what this animator says excuses the poor end-result of a multitude of animations in ME Andromeda vs other recent games.
 
i don't think anyone is comparing ME:A to UC4...hands down the pinnacle of animation in games at the moment......

if anything else, they are comparing ME:A to ME1 - 3.........which doesn't suffer from animation problem walking up a stairs.

We just had a thread where the OP postulated that every game should aim towards Uncharted 4 level of fidelity in terms of animation.
 
Good for him to speak on behalf of the developers.

Sure it's not great and critique is warranted but there is no reason for people to harass them or for example be elitist and demand animation as in uncharted for every fuckong game. Absolutely disgusting and elitist.

Critique and hope they are able to fix it or do better for the next game and hope ea gives them enough funding to be able to do so.
 

ElFly

Member
I was wondering that too. F4 did get some flack about the visuals though after the reveal trailer. I think people didn't tear it apart because the usual response to any Bethesda Game Studio RPG game is that "mods will fix it".

I think the real difference is that Fallout is partially satire, so any jank adds a little of unintentional comedy to the intended comedy. ME is a serious series

on top of that, the fallout jank isn't that bad because when it goes bad, it normally is more dramatic and drastic. so you have people in fallout moving either robotically but as expected, or spastically and obviously wrong

while ME:A does have people walking funny, the worse part is the expected movements are slightly wrong. handshakes failing a little, eyes jittering, etc. while Fallout will have that kind of thing from time to time, it doesn't seem as pervasive as in ME:A

honestly the animations don't shock me that much, cause animation is _HARD_ and it is not unexpected that an AAA game will be rushed out of the door with systematically bad animations, but the graphics for the faces in ME:A are horrible, and dunno why, given they were largely fine in DA:I, a game by the same company, that was launched what...two or three years ago?
 

Falchion

Member
Really cool learning about the quality threshold, not sure why you wouldn't want it as high as possible before you come back and manually tweak whatever you can.
 

Tosyn_88

Member
I will still buy the game just to support Bioware, they have earned that much trust from me. I hope they can learn from this with regards to how much they manage their future projects because from all I have gathered, it seem to boil down to a Project Management issue. They had a set deadline, a lot of their initial estimated workflow didn't align with the timeline because they underestimated how much it required to actually do something of such magnitude, especially with impressive results. I mean, this is the first Mass Effect for the new consoles and everyone was very excited, so of course they wanted to impress and it didn't go well yeah, I'm sure they are gutted as are the fans but they will definitely learn from it, that is a guarantee.
 

Widge

Member
Yeah, I was thinking the same thing when I first saw the thread earlier today, but I didn't want to sound like I didn't appreciate him saying what he said or anything.

Not that I'm implying you're like that...

I may be fickle but if I have to click out of Twitter to read what someone is saying off the cuff, I just skip the entire thing. Same thing if I see a partial post with fb.me url following it up.
 

KorrZ

Member
There are ways of recording facial capture with a head set up that do not involve dotting the face. Just as an FYI

The Last of Us documentary was interesting for this. They did "facial capture" but without the dots, essentially they just recorded the actors and then hand animated based on video recording of the actors performing. I know this doesn't really apply for every conversation in a massive RPG but the quality speaks for itself.
 

gypsygib

Member
I hope developers don't have to record animations for each game and all have some massive database of animations they can use across games and just add and combine as necessary.
 

Majukun

Member
fair argument,but it only shifts the blame from the single division of the team to management/director and or publisher..either the director should have been able to understand that his/her approach was not possible,or the publisher should have delayed the game to increase the quality of the final product..yes it's not free, but if you don't care you get a situation like andromeda,where the game it's reaching the shelves with a lot of bad advertisement around it.
 

Pif

Banned
Thank you for the good explanation as why MEA animations suck.

Also, now it is also clear it is no easily fixable it seems.
 

13ruce

Banned
I agree with him but when games like Witcher, Horizon and Zelda and even the previous Dragon Age game and Mass Effect games have better animations why can't this one?

It feels like something went wrong during development and that's why it's that bad.
 
Jonathan Cooper said:
Andromeda seems to have lowered the quality of it's base algorithm, resulting in the 'My face is tired' meme featuring nothing but lip-sync. This, presumably, was because they planned to hit every line by hand. But a 5-year dev cycle shows they underestimated this task.

This makes so much sense. Especially because while they're very few and far between there have been some scenes where I was like "wait, that was really well animated". But those moments are like an oasis in a desert.
 
The animations do indeed suck. But just changing the eyeball textures for everyone would help this game a tonne. Even just for the main characters you meet and your squad. Surely that can't take too long?!

That and a performance patch on ps4 because my God does it run like dogshit.
 

Justinh

Member
I may be fickle but if I have to click out of Twitter to read what someone is saying off the cuff, I just skip the entire thing. Same thing if I see a partial post with fb.me url following it up.

ehhhh

I sort of get that feeling when I see someone tweeting 21 times in a row to do the same thing. Reading something on Twitlonger is much more convenient to read, and I'd guess to type too.
 
Cooper's tweets make a lot of sense and seem to play into the idea that this game just had way too much of a hill to climb with new premise, new characters, insane hype, and using Frostbite for the first time on top of all of that. I would have preferred EA hold the game back and give the team 6 more months to polish it before release, but these tweets make it seem to me like hitting a March release date wasn't optional.

This is.... like... this is running on a 10 year old alienware laptop or something right?

This shouldn't be real. Like the lighting and wash out and the eyes oh my god the eyes...

Wuuuuut happened. I enjoyed dudes tweets.. any behind the scenes on development is cool and I enjoy it. He especially seems like a cool guy for diffusing some of the snark around this conversation but ... like the animation isn't the only problem with Andromeda..
every ounce of writing and every image and every video has some glaring fault that i can't avoid locking into and it's just so fucking distracting.
Anyone that's put more than a few hours into it here have some kind words? does it develop past all these concerns? does the gameplay warrant giving at least some of this stuff a pass?
oh how the mighty have fallen... fuck..

Actually, yeah. This is my first Mass Effect game so I'm not holding it up against prior games in the series and I'm having a good time with it. Some of the facial animations are really bad, most have been mediocre, occasionally they're good. That having been said, the voice acting has all been awesome so far, which takes some of the sting off. I also play with subtitles on, so I'm often reading along with the dialogue rather than watching the characters' lips.

The combat has been really fun so far. Menus are a bit cumbersome and there are a TON of customization options for how you play, but once you get a loadout that you like, the moment to moment gameplay works great.

Honestly, I think the reviews have this right on the money. To me, it's around a 7 - 7.5 and I'm about 8-10 hours into it. The game has some issues (many of them with presentation), but I've been on an open-world / RPG kick lately and this is scratching the itch.
 

Tiandrad

Banned
What I got from this. Video game animations are hard, and even harder to do with a rpg. My issue isn't how difficult it is to do; my issue is them not putting in the extra time to do it well. If time was the issue, the game should have been scaled back to accomplish every part of it well. But instead, they tried to make a massive game and underdeveloped it in important areas.
 

CamHostage

Member
There are ways of recording facial capture with a head set up that do not involve dotting the face. Just as an FYI

uZKrP9R.gif


Right, the dots are probably the first piece that will eventually evolve in facial mocap. (I haven't looked into the tech but I assume a combination of more facial bone detail in models, increased resolution cameras, and improvements in middleware interpolation in stuff like FaceFX is making that happen? And I wonder how much less uncomfortable application like discreet facepaint and invisible UV markers can help get detail that dots are a kludge for?) I'm not sure what the differences are so far in marker-less mocap quality and range (if you have the markers, my guess would be that you'd get some movement data that doesn't conform to your model rig that'd then teach you what new bones/movements to add to the model?) but the fact that it has good lip/cheek/jaw movement and can even catch and simulate that unusual cheek-puff makes it look quite promising to my eye.

If you look at HellBlade by Ninja Theory, for example, even over the life of its development, they have changed how many dots are needed (or at least have occasions where they need fewer dots than other times.)

hqdefault.jpg

Mocap from a video in 2015

hqdefault.jpg

Mocap done live in 2016
 
Top Bottom