I think mozart number 1 looks better cause it runs on my dumbphone and forces my high IQ brain to fill in the missing...

fuck you and fuck everyone in this thread and fuck me fuck this board it is shittier day by day but this image always gets me every time fuck
do you understand that textures can be baked from high-poly meshes on to low poly meshes? and from insanely high poly meshes too thus resulting in all sorts of multi-purpose maps for lighting interactivity irrespective of how high poly the mesh itself is
do you understand that certain level of facial animation or animation in general is literally impossible at low poly
do you understand that your average mobile game mesh for some League of Legends mobile port shit game has 40000 polys at least?
BUT MOST OF ALL
Do you understand that multiplying polycount when the detail isn't there will not result in anything, because it is the same fucking SHAPE RIGHT? THE SAME FUCKING MESH, RIGHT?!
DO YOU GRASP THE FACT THAT IF YOU TAKE A MESH THAT HAS BEEN CREATED WITH A CERTAIN LEVEL OF DETAIL IN MIND AND JUST DOUBLE IT'S POLYCOUNT THAT WON'T CHANGE SHIT BECAUSE YOU ARE JUST INCREASE THE NUMBER OF POLYGONS THAT MAKE UP THE SAME FUCKING SHAPE WITHOUT USING THE EXTRA POLYGONS, YOU GET THAT RIGHT?!
IF YOU HAVE A FUCKING PYRAMID AND CUT EACH OF IT'S SIDES INTO SMALLER TRIANGLES THAT WILL NOT CHANGE THE SHAPE OF THE PYRAMID, YOU GET THAT RIGHT?
YOU WOULD NEED TO ACTUALLY MOVE THE SUB SECTIONS OF THE PYRAMID TO MAKE A NEW SHAPE, RIGHT? LIKE MAYBE A PYRAMID WITH LINES OR SOME SHIT, RIGHT?
YOU DO UNDERSTAND THAT? OR IS THIS BOARD FILLED WITH LITERAL 30 IQ FERRETS?
HOW DID THIS IMAGE
EVER
GET ANY TRACTION
YOU DO NOT NEED TO UNDERSTAND 3D
YOU JUST NEED NOT TO BE A RETARD

settle down

Go take a look at a PS2 game and then rethink that.

>fully modelled lapels and shit
>not just getting a flast chest with textures representing them

No listen this is some fucking long running ruse
Imagine you have legos, right? Like maybe one of the giant legos for retarded small kids, right? And the retarded child builds a building with those big blocks, duplo or whatever
Then you get your usual legos and the retarded child makes lego bricks the size of bigger, retarded duplo sized bricks out of smaller bricks, then he proceeds to build the same structure that he built with GIANT bricks and he complains to you
>FATHER! WHY CAN'T I MAKE A MORE DETAILED LEGO BUILDING? I USED SMALLER BRICKS AND USED THEM IN THE EXACTLY SAME WAY YOU WOULD USE BIGGER BRICKS!
Would you not have wished for his post-birth abortion at the spot?
HOW
DID
THIS
IMAGE
GAIN
TRACTION?!!!!!!!!!!

dude are you defending 4 like a loser? 3 is superior

shhhhhhh its ok user

Depends you fucking retard.
There are meshes with up to 12k polys on PS2. Leon in RE4 has 10k polys as his mesh, but if the image is trying to demonstrate only head that is 10k, that is transitional period between ps2 and ps3 because ps3 instantly had higher polycount than that.
But a single mesh with that polycount was entirely possible.

Opinions are for uncalms

>HOW
>DID
>THIS
>IMAGE
>GAIN
>TRACTION?!!!!!!!!!!
Because it's correct and the retards who keep going
>b-but you can just use more polygons to make it more detailed!!
Are missing the point. You don't necessarily want the character to be more detailed if the character already looks like how it's supposed to look like.

4 AND 3 ARE ALMOST LITERALLY THE SAME MESH THE ONLY DIFFERENCE IS THAT BECAUSE OF THE SHITTY LIGHTING AND MATERIAL YOU CAN SEE THE FACES EXTREMELY SHARPLY/DEFINED

>let's ignore improvements in lighting
>let's ignore improvements in textures
>let's ignore improvements in shading
>let's ignore improvements in animation
>let's ignore improvements visual effects
I hate boomers, old people just need to die.
Of course there's never gonna be as much of a jump in graphics as the one from SNES to N64 again, but that doesn't mean there aren't still immense, important improvements all the time.

There's not actually anybody who still believes this image, right? I mean, nobody can actually be that retarded, can they?

No, retard. 4 looks chubbier than 3, especially in the face. Like Mozart just gained some extra weight. Also, the suit looks worse, and the hair looks like it's made of plastic.
3 has some of these problems, but everything looks more real.

You know, someone posted a "debunking" of this image some time ago, but after thinking about it a little bit more, I think the image still stands. Even if he took the model at one angel and lowered the count down, you can still see MASSIVE diminishing qualities just from that, whereas the one with more triangles doesn't really add much to the overall quality, and you're definitely spending WAY more effort to produce more triangles than you would just settling for triangles that look well enough.

I think this picture actually still stands.

Attached: Face shifting.gif (637x358, 2.34M)

>DO YOU GRASP THE FACT THAT IF YOU TAKE A MESH THAT HAS BEEN CREATED WITH A CERTAIN LEVEL OF DETAIL IN MIND AND JUST DOUBLE IT'S POLYCOUNT THAT WON'T CHANGE SHIT
if you're referencing tessellation it actually does assuming it has a baked map from a high poly model to work with, other than that you're talking nonsense, no one ever said that

>HOW DID THIS IMAGE
EVER
GET ANY TRACTION
cause any normal human being with 2 eyes agrees that 4 looks better but that seems to trigger non humans

You are a fucking moron holy shit you fucking retard, the mesh itself does not even need to be high poly do you understand that more than half the detail is baked into texture maps of all sorts? Even when creating diffuse maps for stylized games it is beneficial to bake from high poly to low unless you are outright making a game that looks like Fighter Z and even in that case it is still benefitial to use high poly mesh because it gives shaders more room to work with without fucking up and creating all sorts of weird ass seams and edges
Look at Okami from PS2? See her? Her mesh and her effects can be done 20000 times better and closer to her intended concept art today, you have to be a literal fucking MORON to think that there is any benefit to older tech
I CAN UNDERSTAND deliberately limiting yourself to extreme low-poly indie look but EVEN THEN you could actually have it secretly be HIGH POLY which would grant you benefits FOR SHADERS

that example is so fucking bad, poor topology in each one of them, would be way bigger difference between 6k and 60k if it was made proper

>cause any normal human being with 2 eyes agrees that 4 looks better but that seems to trigger non humans
The problem isn't that 4 looks better, it's the fact that 4 only looks slightly better, but has a massive amount more work and system requirements than settling for 3.

60k has been an industry standard for many many years, we are past that actually

>if you're referencing tessellation it actually does assuming it has a baked map from a high poly model to work with, other than that you're talking nonsense, no one ever said that
I am arguing for higher poly fucking brainlet, the image is retarded
did you even read the subtext
>Diminishing returns
The image implies that increasing the polycount has diminishing returns

>The image implies that increasing the polycount has diminishing returns
But it does.

Post the one where its 6 billion triangles and its an unrecognizable blob

This is almost entirely a nyquist limit thing. The 'source' mesh is low frequency enough that 6000 tris recreates it accurately, but this is not true for every mesh.

Yeah we should ignore them, because if anything they've made games look worse. You used to be able to simply see what was in front of you. But with these so called improvements, games now require crutches like entirely different view modes that highlight things or 2D markers all over the screen.

This is fucking wrong, the 4 looks slightly better because the 4 in that image has bad topology and both of them are covered in generic stock shit material with flat lighting
there is a gigantic difference between what you can do with 6k and 60k polys
It doesn't even have to do with just how the mesh "looks" to you, do you understand that it also affects how light and shaders interact with it?
Leave topology and polycount to people that actually understand it
Also who spread the myth that making 6k mesh is easier than 60k mesh, it is not, difficulty is not measured in polygons especially with modern sculpting software where you can have gigantic polycounts just out of nowhere
the difficulty is in what you are TRYING TO ACHIEVE
If you are trying to achieve HYPER REALISM IT WILL BE HARD
If you are trying to create some sort of weird super anime shader like fighter Z IT WILL BE HARD
If you are making generic low effort shit IT WILL BE EASY
Making assets became easier than ever, not harder, the budget is increasing because of the ASSET COUNT not because of how difficult the asset creation is, though there are more complex assets today of course

600 has the most soul.

>the 4 looks slightly better because the 4 in that image has bad topology
Sorry, I'm going to stop reading there. Not because I think you don't have value, but because I'm very much a bystander and amateur in this. Can you instead demonstrate what you're talking about with visual representation instead so I can follow along and confirm you're not just throwing out graphic terms to confuse the issue?

I had the same thought when I saw the debunking image.
>Heh you thought there were diminishing returns but there's actually slightly less diminishing returns than you originally believed

No it doesn't, look at Kratos from new God of War, that sculpt is literally unachievable with lower poly count and it is very clearly not a small jump from God of War 3, it is fucking night and day

Attached: duc-phil-nguyen-kratos-002jpg.jpg (2180x936, 1.23M)

>massive amount more work and system requirements than settling for 3
lol 60k is nothing
characters were starting to use a 60k count back in like 2014, we are now way past that

and they definitely look much much better

Just add more triangles.

What happened to his sick abs?

I'm gonna be honest with you, that's not that big of a jump as you're advertising. It just looks like the fur and the hair improved. That's really all I see. Was that really enough to warrant jumping system requirements up so far? And then look at the gameplay between the two games as well, and I think that speaks massive volume in and of itself.

>and they definitely look much much better
Eh. I mean, I can see some improvement, but I don't think that's really enough to warrant the other areas where games like this suffer the most, which is gameplay, good writing, development time, and cost.

What he is basically trying to convey in laymen terms is that the model number 4 does not have enough detail to utilize 60k polygons. In reality a model like that would not have that many polys in the first place because developers constantly try to optimize.
So what usually happens is that an object that is not detailed enough to need that many polygons stays at lower polygon count just with nicer textures than before, while models that need higher polycount, like characters with complex clothing and folds get 60k and much above.
The image in the OP can not demonstrate the difference between 6k and 60k accurately because no matter how much you increase the polycount the intended detalization of the mesh will not change, the intended shape if you will. If the intended shape would be more detailed, with wrinkles and shit, it would be obvious.

Souless - Soul

the problem is that they just added a subdivision modifier, they could have spent the 60k tris much better, but really what happens is you you use something like mozart no. 3 and then make a sculpting mesh with 3 million polys with ridiculous detail, then map it onto mozart no. 3 with some maps. You wouldn't need to have your program draw that many triangles.

Be original.
boards.fireden.net/v/search/image/jKojaoiQmANsVnCN-EkcQw/type/op/

Attached: 1445991820518.png (1006x1007, 492K)

Even for shaders you will reach a peak eventually where you get the game looking like you want it to look like and there's no benefit to adding more polygons.

You have to be literally blind or lying.

Attached: duc-phil-nguyen-kratos-002jpg.jpg (1257x559, 347K)

Consider it this way: If you have an infinitely tessellated box, then however many polys you use to represent it it will always look the same, but if you have a sphere then the recreation of the shape will gradually improve as the number of polys increases. The value of adding extra polys diminishes for the cube immediately, but diminishes after far more polys for the sphere.

Something like a palm tree would have an even higher limit.

The mozart mesh or whoever it is is based off some model that doesn't have a continuous amount of detail. As some point you've exhausted the frequency of the source surface and adding more polys isn't increasing the detail, but a better mozart head model might have a higher detail surface in which case more polys would add more.

Yes, we are not there right now, and we are just entering the real time raytracing age which will change shit a bunch when it will become more common in coming five years +.

This is Beethozen not Mothart.

Basically this

runescape through runescape 3

>What he is basically trying to convey in laymen terms is that the model number 4 does not have enough detail to utilize 60k polygons.
The problem with this statement though is that if you look back at the lowered count, you can see the quality shift downwards by a massive amount, more equivalent to their x10 loss of triangles. And then pictures like don't really do much to convince me that the 60k+ really does anything other than jack up costs for hair quality.

Is it possible to take a 60k triangle object, downgrade it to 6k, and show the same level of quality drop that you would see at the 600 and 60 levels?

I mean, fuck, look at . The drop from 20k to 2k is nowhere NEAR as significant as the drop from 20k to 200. Isn't that evidence enough?

Hair. Looks fatter. I guess the textures have more shadows which give him weight. Honestly the increase isn't impressing me.

But the point is that there are definitely dimenishing returns. Most of the "debunking" is focusing on the quality of the top model, but almost always fail to address the massive inequality of the drops. Dropping from 20k to 2k isn't as significant as the same drop from 20k->2k->200 in that picture. How is that not diminishing returns, exactly? I can look at the 2k and the 20k picture and tell every single part of his coat apart, but I can't do the same in the 200 picture without some guessing.

The biggest thing that gamers need to understand is that what is making development longer and harder is the asset count, not the asset complexity. The complexity of overall scenes and how they interact with each other/player.
In old times a lot of detail of the background would literally be a flat texture or just not present at all. Rocks on grounds? Random buildings? Varied pedestrians? Extra idealization of the interior? None of that.
So while making 3d models became easier and more intuitive in a lot of ways, not to mention that we can replicate previously impossible effects for all sorts of stylized looks, what used to be implied detail now has to be modeled and created.
Background, foreground, character interactions, random assets, overall asset count etc. that inflates more and more because people expect more and more of the world to be tangible.
If the asset count was the same as in the old games it would be easy and wouldn't increase the budget much since making higher detailed meshes is easier than ever now if you are a skilled artist, but it will look jarring when you have a super detailed character walking in an empty world where half of the details are flat textures.

based furious t/3/apot

"But the point is that there are definitely dimenishing returns."
There are diminishing returns for this model, but that's dependent on the model. The poly limit for different models will be different.

This is something game artists, programmers, and hardware vendors are well aware of. Artists won't waste their vertex limit on things that don't need it.

Be original.

First of all, no offense but you are blind.
Second of all read
The increase to 60K+ has almost no impact on the budget, default 3d sculpting software makes you model at a resolution much higher natively and has brushes with in-baked details like automatic pore brushes and scars, that is an on-issue.
Why would you not detail hair and character more when it looks better and isn't much harder since production technology and artists themselves got better?
The issue is that overall size of the projects, not the fucking individual asset detalization. Games are bigger than before, and those already bigger games have even more assets in them than ever before. In addition to that reducing the asset detalization but keeping modern polycounts etc. would not do much to decrease the budget unless you decrease all tech including lighting, shaders, physics whatever, which would be horrible. The key issue would still be the ammount of individual things to do, not how hard one thing is to do.
What happens then is that usually our leads start hiring more and more people for minute small things, for example on Witcher 3 there was a team of people that animated exclusively character faces during dialogue choices, which was a different team from guys that animated their bodies and all of them were a different team from the guys that animated the cutscenes and more impressive stuff.
The individual ammount of labour has increased.

I sort of get what you're saying, yeah.

>but that's dependent on the model.
Is it? I feel like this would affect every model.

>Artists won't waste their vertex limit on things that don't need it.
But then why is it that AAA titles of the last 10 years have been creating really beautiful movies about hallways instead of the older AAA titles that would actually have branching paths and gameplay, and the decrease on content while the costs of development have gone up, if not the fact that more assets are being spent on graphics and art than they are with actually creating a good game?

Generally you expect a certain level of detail per inch, this scales with the size of an object...so imagine if you expect an object size of a finger to have 100 polys, a building which is millions of fingers can have 100000 polys if it's extremely detailed...like a gothic church, though even in that case a lot of it will probably be done with normals and other maps to optimize.

>First of all, no offense but you are blind.
I'm not though? There's not really that much impressive between picture 1 and picture 2 other than the things I mentioned. I can see that picture 1 obviously looks better, but it's failing to impress me as much as a simple jump from PS2->PS3 era.

>The increase to 60K+ has almost no impact on the budget
Sure, now. What about when 60k was cutting edge though? What about the push from the default 60k to where we're at now? You're also ignoring the processing power jumps to render these things as well. PS1 only needed 2MB of RAM and a little 33 MHz processor, and then the PS2 jumped up to near 300 MHz with 32 MB of ram. Now we're on multiple gigabyte systems, and most games only look marginally better than the PS3 days? That isn't diminishing returns to you?

Read.
It is because each square inch of the game is expect to be detailed and have more things. Look at Resident Evil 2 remake to understand.
The level design is comparable and often times similar since it is inspired/copies it, yet there are just more objects and more of everything everywhere.
There is more light, more clutter, more interact-able objects just more everything.
If you make one corridor that leads somewhere, and you want to make branching paths for multiple outcomes or whatver you would need to do the same thing but three times.

...

It's a specific case of this thing en.wikipedia.org/wiki/Nyquist_frequency
If you're interested, this is also the reason you get aliasing. In fact, a polygonal meshes inability to recreate a surface IS aliasing.

> But then why is it that AAA titles of the last 10 years...
I would say that in general games don't waste polygons that much, but modern game development works to some budgets better than others. The content production of a modern AAA game is well aware of the limits of rendering a nice frame at ~30/60FPS, but isn't as good at building the larger more varied world. This is due to technological issues with loading / streaming / memory, as well as focusing on making a smaller number of reusable high quality art assets instead of many lower quality art assets.

what the fuck

>What about the push from the default 60k to where we're at now?
Which came from the fact that we started actively utilizing sculpting software as an industry standard for character models, which allows developers to have more detailed mesh very quickly from the get-go. Z-Brush is so easy to use for character modeling compared to how usual 3d modeling works that any good traditional sculptor can become more efficient than any oldschool veteran 3d modeler in about 3 months to a year max, the software is so good that it became just about the skill of the artist basically.
>You're also ignoring the processing power jumps to render these things as well.
Yes, and you know what had an even bigger impact? The overall increase in asset count.
Basically if you want game world and decals to be empty again, sure. Have a stroll through jungle in MGS 3, then forest in MGS 4 and then do the same in MGS V and count how big was the asset increase in each case.

Basically what I am trying to convey is that the increase in budget is due to increase in expectations and demand of players, not because of detalization of individual assets. Everything has to be more and bigger and more pretty, every single asset, not just meshes.
Basically you are demanding a global downgrade of all modern visual tech that is used in games if you want budgets to deflate.

>The Nyquist frequency, named after electronic engineer Harry Nyquist, is half of the sampling rate of a discrete signal processing system.[1][2] It is sometimes known as the folding frequency of a sampling system.[3] An example of folding is depicted in Figure 1, where fs is the sampling rate and 0.5 fs is the corresponding Nyquist frequency.[note 1] The black dot plotted at 0.6 fs represents the amplitude and frequency of a sinusoidal function whose frequency is 60% of the sample-rate (fs). The other three dots indicate the frequencies and amplitudes of three other sinusoids that would produce the same set of samples as the actual sinusoid that was sampled. The symmetry about 0.5 fs is referred to as folding.
Sorry captain, this is all greek to me.

>The content production of a modern AAA game is well aware of the[...]
Aren't the memory issues in and of itself indicative of the diminishing returns? My current toaster's graphics card has enough RAM on it to simultaneously load thousands of NES games fully and at once, but we've pushed graphics so far that we can't load a full map without load times?

>Which came from the fact that we started actively utilizing sculpting software as an industry standard for character models,[...]
Hm. That's interesting. Aren't there inefficiencies or unoptimized items in there? That's typically a drawback with programming engines.

>Have a stroll through jungle in MGS 3, then forest in MGS 4 and then do the same in MGS V and count how big was the asset increase in each case.
Sure, that's fair, but how many of those assets were someone putting in something interactable for the game's sake, and how many of those assets were like , modeling every single strand of grass?

It's not like I'm asking us to go back to the stone age and play around with Atari again. I just think that the value return on graphics pretty much stops at the point where we can read small-text on a book on a shelf. Pushing beyond that I think gets us nothing except additional surface area for the game to bug out and/or crash in, increases time to release a single game, and stops us from taking riskier ideas. I mean, who is going to want to pay for development for a 5 year-long game idea that's risky vs one that could be made 6 months - 1 year's time?

>modeling every single strand of grass
LMAO

> Aren't there inefficiencies or unoptimized items in there?
Yes there are, so after a high detailed sculpt is finished usually they go through a process of slight retopology to become optimized for game use.
>modeling every single strand of grass?
Most of them are just that, rocks and grass, but we don't know how to fake that fast enough and efficiently enough. There is stuff like speed tree for quickly generating beautiful varied trees, and terrain brushes in various engines but often times without personalized touch it ends up looking very cheap so people have to "model individual stands of grass" so to speak. It's a constantly battle of trying to do as least as possible without it looking really cheap, which still results in people doing a considerable amount of unique assets compared to old games .
AI generated stuff and photogrammetry will save a lot of time in future and game development should become even easier and faster, that includes AAA sized games. The bad thing is that by then a lot of devs will be in habit of focusing on asset churning instead of gameplay so even if it becomes extremely easy and fast gameplay might not improve until people realize that this should be the focus.
In reality there is nothing stopping people from doing good games now, game design itself is a ZERO BUDGET thing, it only takes the minds of game designers, animation etc. would stay largerly the same since it's already there.
>tl;dr
We have bad game designers in the industry right now.

>That's typically a drawback with programming engines
literally what

fuck you and fuck everyone in this thread and fuck me fuck this board it is shittier day by day but this image always gets me every time fuck
do you understand that textures can be baked from high-poly meshes on to low poly meshes? and from insanely high poly meshes too thus resulting in all sorts of multi-purpose maps for lighting interactivity irrespective of how high poly the mesh itself is
do you understand that certain level of facial animation or animation in general is literally impossible at low poly
do you understand that your average mobile game mesh for some League of Legends mobile port shit game has 40000 polys at least?
BUT MOST OF ALL
Do you understand that multiplying polycount when the detail isn't there will not result in anything, because it is the same fucking SHAPE RIGHT? THE SAME FUCKING MESH, RIGHT?!
DO YOU GRASP THE FACT THAT IF YOU TAKE A MESH THAT HAS BEEN CREATED WITH A CERTAIN LEVEL OF DETAIL IN MIND AND JUST DOUBLE IT'S POLYCOUNT THAT WON'T CHANGE SHIT BECAUSE YOU ARE JUST INCREASE THE NUMBER OF POLYGONS THAT MAKE UP THE SAME FUCKING SHAPE WITHOUT USING THE EXTRA POLYGONS, YOU GET THAT RIGHT?!
IF YOU HAVE A FUCKING PYRAMID AND CUT EACH OF IT'S SIDES INTO SMALLER TRIANGLES THAT WILL NOT CHANGE THE SHAPE OF THE PYRAMID, YOU GET THAT RIGHT?
YOU WOULD NEED TO ACTUALLY MOVE THE SUB SECTIONS OF THE PYRAMID TO MAKE A NEW SHAPE, RIGHT? LIKE MAYBE A PYRAMID WITH LINES OR SOME SHIT, RIGHT?
YOU DO UNDERSTAND THAT? OR IS THIS BOARD FILLED WITH LITERAL 30 IQ FERRETS?
HOW DID THIS IMAGE
EVER
GET ANY TRACTION
YOU DO NOT NEED TO UNDERSTAND 3D
YOU JUST NEED NOT TO BE A RETARD

Attached: 1553901658339.jpg (700x700, 68K)

fuck you and fuck everyone in this thread and fuck me fuck this board it is shittier day by day but this image always gets me every time fuck
do you understand that textures can be baked from high-poly meshes on to low poly meshes? and from insanely high poly meshes too thus resulting in all sorts of multi-purpose maps for lighting interactivity irrespective of how high poly the mesh itself is
do you understand that certain level of facial animation or animation in general is literally impossible at low poly
do you understand that your average mobile game mesh for some League of Legends mobile port shit game has 40000 polys at least?
BUT MOST OF ALL
Do you understand that multiplying polycount when the detail isn't there will not result in anything, because it is the same fucking SHAPE RIGHT? THE SAME FUCKING MESH, RIGHT?!
DO YOU GRASP THE FACT THAT IF YOU TAKE A MESH THAT HAS BEEN CREATED WITH A CERTAIN LEVEL OF DETAIL IN MIND AND JUST DOUBLE IT'S POLYCOUNT THAT WON'T CHANGE SHIT BECAUSE YOU ARE JUST INCREASE THE NUMBER OF POLYGONS THAT MAKE UP THE SAME FUCKING SHAPE WITHOUT USING THE EXTRA POLYGONS, YOU GET THAT RIGHT?!
IF YOU HAVE A FUCKING PYRAMID AND CUT EACH OF IT'S SIDES INTO SMALLER TRIANGLES THAT WILL NOT CHANGE THE SHAPE OF THE PYRAMID, YOU GET THAT RIGHT?
YOU WOULD NEED TO ACTUALLY MOVE THE SUB SECTIONS OF THE PYRAMID TO MAKE A NEW SHAPE, RIGHT? LIKE MAYBE A PYRAMID WITH LINES OR SOME SHIT, RIGHT?
YOU DO UNDERSTAND THAT? OR IS THIS BOARD FILLED WITH LITERAL 30 IQ FERRETS?
HOW DID THIS IMAGE
EVER
GET ANY TRACTION
YOU DO NOT NEED TO UNDERSTAND 3D
YOU JUST NEED NOT TO BE A RETARD

Attached: proof god doesnt exist.webm (720x1280, 2.93M)

he probably meant that engines that do something in an automatized fashion usually net a more unoptimized result, which generally is true

>t.

Attached: 1546948479160.jpg (480x480, 43K)

Graphics belong to the artist. Higher technical specs only open up more freedom to do what the artists want. Sometimes that's not a good thing, humans often do their best work with harsh limitations.

Attached: 43000polygons.jpg (600x600, 28K)

Not an argument.

When you use an engine like Unity, or something that will actually create chunks of programming blocks for you, typically they'll have their own flaws and baggage that come with it because the engine is trying to approximate what you want it to do. Sometimes this isn't an issue, but most of the time this is where you'll get really weird bugs. On top of which, you can't necessarily open one of these engines and tell it to skip certain steps it doesn't need to do, thus creating unoptimized code under the hood.

>Most of them are just that, rocks and grass, but we don't know how to fake that fast enough and efficiently enough.[...]
What you're saying does have some merit, but I still feel that most AAA games that come out these days still focus mostly on the graphics side as a push from somewhere on the totem pole. But I'll take what you say into consideration for now and process it for awhile.

>empty world

But, lets take Skyrim for example, what's wrong with using 2 triangles and a sprite, or 4 triangles and a sprite on the point where two the two planes intersect, for decorative shit that the player would barely interact with anyways except to either place on a shelf or toss around like a retard?

Attached: BMthink.png (207x218, 5K)

Wow this guy is mad

I am an art director at a gamedev, trust me, no one here wants to work more than we have to. Certain things are unachievable with a lower budget, that includes the AAA "look", but nothing is stopping the AAA companies from having good gameplay like you would want other than their own bad understanding of game design. The only thing that legitimately gets hampered by modern large budget development would be something like detailed varied roads of Morrowind that lead you to where you have to go since that would need even more unique assets and consideration for world design versus something as simple as a quest pointer...but we are already creating oversized worlds with increasingly bigger worlds every generation so that is almost a non-excuse.
A new generation of smart game designers has to rise up that does not compromise their gameplay decisions.

That's Beethoven, not Mozart.

>do you understand that certain level of facial animation or animation in general is literally impossible at low poly

But why would I need super detailed facial animations when I'm staring at the character's ass or stuck in first person?

>mobile games
>needing to devote time to anything other than rendering you're shitty waifu

Texture resolution and good bump mapping are more important than polygon count.

t. doesnt even play games

>I am an art director at a gamedev
Honestly, I already figured that out just by the way you speak. Working in an office job myself, I could tell you had some experience working with a team to actually produce results.

Thank you for the info user. You've given me some food for thought on the subject.

then go play modded minecraft

You're 100% right. While the methodology of this image isn't perfect, it clearly demonstrates the basic concepts involved. When you are using triangles to approximate a smooth surface, on a limited-resolution monitor, you eventually reach an effective cap on fidelity. That's why we use baked textures and lighting and advanced shading shit, because they get you to that cap for cheaper.

>knows nothing
Normalmapping and tesselation means polycounts are almost irrelevant

We are using technologies like that in the industry all the time, sometimes for far away details there is literal DOOM shit going on with just flat textures that follow the camera, it's just that certain things can not be achieved that way.
Lamps or slightly larger objects you might get near to can not be faked that way for example which is a very large portion of assets already.
Even shit like snow, you think that who needs that? Just make a texture of snow and plop some footprints when the character walks over the snow, but then RDR 2 comes out with it's tesselated dynamic snow and now you are fucked, because that is part of the expectation now and it genuinely looks cool...so all of AAA gets locked into this thing where if we don't do like that or better that means that our game is somehow outdated.
Of course this is not entirely true and if the game is truly great it will push through, but investors have a lot more confidence in visuals convincing the players that visuals will be good since it's straight forward than the trailer convincing players that gameplay will be good, since that is largely subjective and hard to convey.

Minecraft with proper texturing and lighting does look amazing. Look up "ray traced minecraft".

It's mozart, beethoven was black.

Yeah, game design is in a terrible place today. It's all so derivative. A game designer no longer goes how do I design this game so it feels like I'm really shooting a russian in the face, instead they go how can I design this game so it's like Call of Duty. Game designers today are all raised on playing games and know little else so all they do is try to do games that are like the games they loved.

PS2 era was 2 for things close to the camera, 1 for everything else. This isn't hyperbole or shitposting. Models were really low poly back then. Textures and lighting effects were used to make things look better.

Trust me, most of the thoughts that are on Yea Forums exist in game dev circles around the world, and there are many smaller developers that show up at GDC every year specifically with hopes that they will be the guys that will manage to become AAA one day but also not compromise on that vision, CDProjekt Red is often seen as an example of how it can be done...even though CDPR did a lot of compromises, just not as much as one would usually take in the industry at that size. Plus they are Pollacks so everything is cheaper there.
Godspeed user, glad to help out.

Exactly user, I agree completely. Gamers that grew up playing games, paradoxically are the problem, instead of gamedevs of the old who had a goal and just wanted to achieve it without any preconceived notions moder game devs are just gluing puzzle pieces together from what they already know.

Based

googled it and found this

Attached: c59.png (500x551, 424K)

youtube.com/watch?v=WoSwEA7QR68

damn i just googled moar and now found this

Attached: minecraft 16K textures.png (400x391, 306K)

that's roblox not minecraft

i dont get it can you rephase it

Attached: 1411705792552.jpg (803x790, 91K)

Those models are 13 years apart, as are these.

Diminishing marginal returns are real.

Attached: 13 years.png (813x648, 603K)

>he thinks abs = super strong

If you work out and you have abs like that, you're starving.

they made him more realistic

kratos almost looked cartoony, specially his face, his looks were too idealized

Diminishing returns does NOT mean no difference. It means exactly what it means, that’s despite the substantial difference in the pieces that make up the image, the differences are not nearly as impactful as they once were . Essentially, the amount of pieces and detail has reached a point where the predecessor looks “good enough” and the successor difference aren’t as noticeable or as impressive as earlier iterations. Which is EXACTLY what your image highlights. Wow you spotted the differences. To your they’re substantial. For some reason. Even though they aren’t really and it’s barely noticeable and nobody else fucking cares. This, by the way, is called autism. You have autism. You are on the spectrum.

He's a dad, so now he's got the god equivalent of a dad belly.

this image is the spectrum of soul to soulless

>barely noticeable
literally kill yourself
>You have autism.
you have retard

crased and binged

It’s barely noticeable. Your autism on the other hand is highly noticeable as everybody in this thread has pointed out

Is a perfect example of how the returns were once significant. Now over the same period of time, they are not. And yes they are barely noticeable, that’s a fact. Just because they aren’t to you does not “diminish” this objective fact. I would suggest that you kill yourself except I know you might consider it so I will refrain. It is, after all, the only viable cure to autism. But don’t do it, user, we need you around for the banter.

clamped

at which point does the soul die

Attached: 1564837668769.jpg (1280x720, 71K)