Death Strandings Designs are weird to me

MoCap is a giant and expensive meme.

It isn't just the female characters. This is what happens when you use the faces of actors in your games. I wish this never became a trend. It's just distracting seeing a celebrity in a game.

There is something wrong with their skin, they look like they are made of wax or some shit. Really uncanny and not very nice looking.

They all look uncanny as fuck. Del Toro looks the worst to me actually. The only good one is Mads, he really looks real.

This is a recent trend in cinema; did you think Kojima wouldn't notice or follow it?

We have not yet conquered the uncanny valley
>It's the anime's fault
Sure mate

It's cause the engine is DECIMA = Horizon Zero Dawn

Attached: youtube_screenAloyMouth.jpg (1280x720, 346K)

>I fucking hate these character designs for some reason
wew

Attached: Shinkawa MGSxDS.jpg (850x1200, 123K)

Those actually look good

they look plastic

They're the actual character designs for the game, by the art director and character designer.

Attached: Death Stranding poster Yoji Shinkawa.png (1080x1920, 892K)

>del taco in that artstyle

Right is the worst offender. The other two are somewhat alright.
The guys looks okay I think.

Middle one looks like a lifeless sexdoll

Middle looks alright.
Seydoux looks like shit though.

Death Stranding is Silent Hills

Attached: dsh.jpg (3840x2160, 868K)

have sex

My Autistic waifu looks perfect. Don't bully.

>can’t even get the trip right

i'm having sex right now and its still weird to me

They looked fine in motion

yes

I think skin texture is too rubberish or plastic, but rest is fine.

Telling someone to "have sex" needs to be filtered or a bannable offense. Substitutes for conversation like that and "cope" are such a fucking waste of a post. In b4 being told to "have sex" or "cope".

Attached: 57092745_10100809610132331_1736621450057482240_n.jpg (475x441, 26K)

It's because they have no visible wrinkles, lines; anything.

They simply ugly.

Honestly I feel like Death Stranding does it much better than other games. Characters don't really feel uncanny valley, at least to me, yet Yea Forums thinks that Devil May Cry 5 faces look very good when to me those look like the definition of uncanny valley.

Attached: nico dmc5.jpg (1024x766, 86K)

What do you mean?

gorilla games did many things wrong but one thing i did like was how they understood that not every single fucking woman is perfect

Attached: C89D94E6-32E8-4497-BA32-25D82DBCAADE.jpg (1200x628, 112K)

It's especially odd considering how good the male faces look.

Nico. Is. Cute!

Art good. Game bad.
Why it always have to be this way?

She's fucking ugly.

They are too close to real people but you can still see they aren't real.

The guy actors also seem off, some uncanny valley feeling shit

Ummmm have sex

Cope by having sex, for your own benefit.

Not with that fucking goblin.

is margaret qualley crosseyed in real life?

Yeah, she's mine

Kill fragile
Fuck mama
Marry Amelie

Keep her to yourself then.

is he crosseyed in game?

It's typically the other way around though. Adding more "realistic" details to these kind of designs make them feel wrong and creepy.

With pleasure

who /threads their own post?

Example? Dante and Nero in DMC5 look fucking fantastic. Vergil has the same problem of no real wrinkles or lines.

nice digits

Death Stranding character designs
>Sam/Fetus
8.5/10, working man look is reminiscent of Isaac from Dead Space, but not as cool as the engineer suit, seemingly multiple different suits to wear as well. If Ludens Space Suit is available at some point then 10/10
>Cliff/Mads
10/10, looks menacing and the weird strands push the normal looking military up a notch and makes him seem even stranger than Higgs
>Amelie
Looks a bit generic. Woman in a red dress, 5/10
>Die-Hardman
Concept art is great, but because they use actors likeness in the game the concept art doesn't come off super well on the actual 3d model since the mask doesn't cover the chin completely 6.5/10 in game 8/10 concept
>Del Taco
Based/10
>Fragile/Seydoux
9/10. Looks mysterious and makes you want to know more about who she works for
>Higgs
10/10. Egyptian influence and gold detailing is kino.

desu it's one of the rare cases where facescan seems to work if they don't have a wonky neck like most "realistic" 3d models.
I'd say Kojima and Naughty Dog(still don't like their games) have the best grip on realistic characters.

DMC did it almost perfectly

Attached: 1544831834958.png (1280x1210, 911K)

I already made the example of Nico. Trish also looks wrong. Vergil too.

They're all thots is the problem, with their 1000000 cock stare

DMC5 characters look fantastic, you're crazy. Even Nico, despite her ridiculous outfit.

That's not how that works, that's not how any of this works.

Mocap uses "bones" which actuate weighted vertices on a gradient to deform the polygons of the mesh that they form. These are meant to follow the major anatomy of the face. Mocap just works to align them, it requires cleanup itself. Moreover it's got virtually unlimited resolution but must be capped at the polygon budget of the game/engine for a given scene. Most modern engines can handle 1m triangles or better, so a closeup shot could use that 'infinite' resolution.

Image related, it's a face rig for Quiet, showing the
bones" of the face. These are what the mocap dots are tracking, the orientation of the bones, both relative rotation and x,y,z to a lesser extent. You could track as many of these as you have vertices and patience.

youtube.com/watch?v=smj8i1__bmo

Attached: quiet.jpg (763x568, 80K)

Hey, it's your opinion and I respect that, I just disagree with you on that point.

Attached: 1551952367949.png (1067x625, 524K)

But Nico looks great, as does Trish. The argument for Trish was always that she was unattractive, not that the product wasn't good.

Nico has a problem that's as bad as the DS faces. Her completely unnatural movements in the shop. She moves like she's underwater, or like she's a fucking worm. It's supposed to be "sexy," oohhhh she's got her hands behind her back, face downcast while she waits for you to buy. It's fucking retarded because humans don't move that way.

>blank expressions, the stupid slight mouth gape
That's not the issue although those could be better. The issue is the facial structure. The bones are weird which is why you're getting the uncanny valley effect from them.

Girl on the right have the same nose as me, is it too bulbous?

The girls look fucking disgusting.

Marry left, kill right, fuck center

It's fine and sorta cute.
If you're a hobbit.

only perfect women are allowed in video games, imperfect thots belong in the gutter

I feel the men get it just slightly better because they have a particular distinct feature like a beard or glasses (Mads doesn't per se, but he's often covered in that oil-like fluid, and you could argue the "uncanny" look works for his character). Mama looks the most "normal" of the three, and incidentally, she's the only one wearing glasses

I love Shinkawa but he still can’t draw female faces.

It's the subsurface scattering which makes the skin look way to waxy and washes out all the sharper details. The corners of the mouths and the lines around the eyes really suffer for this, and we use those features to read human expressions.

It's the combination of real faces with motion capture.
Motion capture in general, but especially for faces, suffers from "lost in translation" syndrome.
Motion capture only covers the basic motions, it does not translate things that are important for us to avoid entering the uncanny valley.
Motion comes with so many consequences, and if they are off or completely missing, you enter uncanny valley.
You perceive a other perfect looking motion, but the effects that are caused by the movement dont fit.
Things like skin/cloth folding on various areas areas to different degree on a large scale, depending on the movement.
And they all feel kinda random too, doing the exact same motion 10 times will result in 10 differently looking pictures with different folds and related effects of motion.
So even if you somehow manage to perfecty recreate those effects via manually hardcoding them into the game, you will will not be able to simulate the continous unique pictures our brain perceives with every moment, no matter how perfectly equal the actual motion was.

And then we have motion capture on faces which doesnt even capture 100% of the basic motions, I dont see actors wearing some kind of beauty mask capturing device. There are multiple trackers, and all the spaces between them are missing and have to be interpolated. Be it because of technological reasons, costs or whatever.
But they have to decide what parts to capture and what parts to interpolate/manually recreate or flatout ignore in the final translation.
So not only is the face a way bigger complex of motions that cannot all be captured, the sheer amalgamation of motions cause an even more insane number of unique effects that also have to be sufficiently provided for us to avoid uncanny valley.

And to make things worse human brains and sense perception exerts a huge amount of effort on the processing of human faces. Body language is at large facial language after all.

Higgs is fucking based such a iconic character
He has a huge meme potential

It's more that it's an idealization of female faces as not having wrinkles holding him back. The art for Sniper Wolf, for The Boss, EVA, other select female characters look great.

continued

We dont perceive these problems in the same kind of way when the Character is crafted 100% in the game. The missing contrast of captured movement and unfitting/missing consequences keeps us out of uncanny valley.

Higgs is probably be the least offensive in that regard. His motion may be captured, but the face model, the aestethics, the look in generell is original. We can tolerate the contrast much better if what we perceive isnt based on a real human.

The bumps on Fragile's shoulders extend out into spikes sometimes, looks like in response to the BT nearby. She's going to be really cool.

I wanna fuck mama in the pussy and cum on her face so bad AAAAAAAAAAAAAAAAAAAAAAAA

How do I learn to draw like this?

retards who reply with "have sex"

There's a lot of issues with just throwing photoscans in. The human biology is very dynamic. Things like blushing aren't easily captured, and using them contextually is even more difficult. The way that textures are handled in instances like these, is they're shot through a polarized lens in indirect lighting to try to remove specular lighting. You end up with a very tonally accurate albedo texture, but the end result is also perfectly flat. More importantly though, is the behavior of light below the skin, subsurface scattering, which can't really be accurately reproduced due the a multitudinous number of factors. You can get a nice pink to shine through, but factors like skin tone, pigmentation, fliushing, light color, source, intensity, ambient temperature... All effect the skin and the interaction of light beneath it.

Then there's the resolution of the texture. You've got to consider that light is a sub-atomic phenomena, photon rays bouncing hither and thither. Resolution in the real world goes down to microscopic levels. That can't be replicated digitally, at least not practically on the scale we're talking. There's a lot of features that will simply be overlooked by photoscans. Then there's various maps you've got to work with, roughness being a key one that as far as I'm aware must be "faked" it dictates the specular/diffuse behavior of the shader. Then there's the normal map which is a coordinate system that can define relative light behavior from the normals +/-x,y,z. Both of these have practical limits in resolution, as does the albedo. Quiet's textures for example have a 2k size for the retail. Cinematics can go well above that, though 16k+.

Finally we have to come to lighting. Lighting in games is hacked together, it simply produces a noisy approximation of reality. With things like the RTX package we're coming closer, and further (more realistic, but with a necessary compromise to resolution to maintain performance).

This looks the same as the MGS4 art released before the first trailer, god I'm getting old.

I’m not gonna lie this post was way more based than it should have been.

It's the lack of imperfections.
Kojima can't stand to give one of his models a mole or asymmetry, and the result is uncanny.
This is a well-understood and basic character design trick.

Their SSS effect is way overblown. I think it's on purpose to give it a more stylized look but it ends up looking kind of weird. Almost like they're covered in a thin wax.

It's weird because I remember having the same problem with Until Dawn and this is on the same engine. It's as if they never bothered to update anything over the years.

Attached: Norman.png (750x421, 321K)

At least the game will have pretty women, I'm so sick of the goblins.

Learn the lighting of 3D shapes and human anatomy. Get so well practiced in your translating of those things into 2D iconographic representations of the interplay of light, form, and shadow that you can simplify it all with basic brush strokes.

Done.

My only problems are that the nostrils on all the characters look a little too open and cavernous, and the lip syncing looks a little unnatural sometimes (even though it is super clean and detailed animating). I mostly noticed the lipsyncing thing when Mads was singing to the baby at the beginning of the trailer. Like sometimes the facial animations look too twitchy or something and I feel like I can see the transitions between different animated expressions.
Overall though I think it looks quite good

thats just because you've never seen a woman in real life

Kill marry fuck you retard

>Is it the blank expressions, the stupid slight mouth gape? The soulless eye?
That's why they're realistic, but I suspect this is just subtler-than-usual bait.

I suppose I should also mention that the framerate of the animation data they're using appears to be too fast, this causes a stuttering effect which is massively offputting. Lindsay Wagner's character is probably the best highlight of this and it goes back to a question of resolution. It doesn't reflect the mostly smooth motion of the musculature of the human face. Then there's micro expressions that aren't properly captured, just very small hinting at inconsistencies in interpretation and executive action. Gaze is another. Contraction and dilation of the pupils is arguably one of the biggest incidences of uncanny valley. There's just massive amounts of information lost moving from reality to digital, and then even more from the raw digital to an optimized format for a game engine.

Attached: 1553651028428.jpg (1448x2048, 351K)

Brittany Venti on the right

>Brittany Venti
Fuck, I shouldn't have googled that.

>photo realism and face scanned hollywood trash

and people thought indie pixel shit is the worst thing that happened to video games

Attached: 3634733.jpg (211x203, 15K)

Poor lighting and skin shading.

They are
Not your kind of people