UC4: Best looking gameplay? *SPOILS*

[Jupiter]

My 10s of current games, points not set in stone.

Anti-aliasing: - difficult, probably Uncharted 4
Physics -
Geometry: -
Tessellation : Crysis 2, nothing comes close
Normal/POM: -
Textures: Star Citizen Alpha by far (static objects, weapon, clothing etc. tech still WIP)
Shadows/Dynamic light sources: - (Star Citizen will get tiled lighting for more light sources)
GI: Alien Isolation
AO: Quantum Break
DOF: Ryse, Crysis 3
Animation: Uncharted 4
PBR: SW Battlefront, The Order 1886
Foilage Rendering: Far Cry Primal
Hair Shading: Paragon
Skin Shading: Uncharted 4
Reflections: -
Volume Smoke/FX: Star Wars Battlefront/ Star Citizen Alpha/ Batman, nVidia effects

 

[Jawed]

I think the idea is that each technical aspect should be treated separately and be supported by screenshots that show "state of the art" and "worst possible". Nothing should score more than 5. That way, the scores have room to scale upwards.

A wiki might be easier. To retain B3D's forum style arrangement a new sub-forum for real time game graphics technology scoring is required. Each technology would then have its own thread. OP can then link to the post that contains the new "state of the art" as that changes over time.

 

[VFX_Veteran]

There are too many variables and too many different games targeting different genres/techniques/art-tech direction to make sense out of this. How can you compare Star Citizen to The Division for example, both have strong points and negatives but there's not much common between the two because of the different style, gameplay, scope etc. I'm at least interested in this but i don't know if the outcome of the thread will be actually useful.

Don't be a Debby-downer. It should be an article and start with UC4 (the most popular HOT game today). Having contributions from the graphics programmers in these boards and us all coming up with an overall GRAPHICS review score could potentially make Beyond3d the de facto standard for the internet to get proper unbiased reviews. Who knows what could become of this?

 

[Clukos]

The problem with this is that for example in Star Citizen you have immense scope, which in turn means that you have some things that don't look right because there's so much you can do with a given budget and time. Do we take the absolute best case scenarios from Star Citizen and compare it to other games? Do we take a mixture of good and bad? Who's there to mediate how many good or bad examples are being posted? Not everyone has access to any given game at any time, only some people will be posting examples.

 

[VFX_Veteran]

My 10s of current games, points not set in stone.

Anti-aliasing: - difficult, probably Uncharted 4
Physics -
Geometry: -
Tessellation : Crysis 2, nothing comes close
Normal/POM: -
Textures: Star Citizen Alpha by far (static objects, weapon, clothing etc. tech still WIP)
Shadows/Dynamic light sources: - (Star Citizen will get tiled lighting for more light sources)
GI: Alien Isolation
AO: Quantum Break
DOF: Ryse, Crysis 3
Animation: Uncharted 4
PBR: SW Battlefront, The Order 1886
Foilage Rendering: Far Cry Primal
Hair Shading: Paragon
Skin Shading: Uncharted 4
Reflections: -
Volume Smoke/FX: Star Wars Battlefront/ Star Citizen Alpha/ Batman, nVidia effects

No games not yet released can be compared against games that are released. We gotta keep this as unbiased as possible.

 

[VFX_Veteran]

The problem with this is that for example in Star Citizen you have immense scope, which in turn means that you have some things that don't look right because there's so much you can do with a given budget and time. Do we take the absolute best case scenarios from Star Citizen and compare it to other games? Do we take a mixture of good and bad? Who's there to mediate how many good or bad examples are being posted?

SC can't be included at all. It's not a shipping title.

 

[Clukos]

SC can't be included at all. It's not a shipping title.

I brought up SC as an example, any game can have good and bad scenarios (because no technique is perfect). Can we really trust everyone to be posting fair examples of "here this technique looks good", "here it looks bad"? Not sure... I don't even trust myself in doing that.

 

[Shifty Geezer]

There are too many variables and too many different games targeting different genres/techniques/art-tech direction to make sense out of this.

That's exactly what this addresses! At the moment we have people saying, "Game X i the best because it has teh Awesom lighting," and others saying, "bu the lighting is static and Game Y is better as not as pretty but real time ToD cycle."

This way, each game is scored on its strengths and weaknesses. The priorities are then open for people to interpret personally. So we'll see which games have the best TOD implementations, and lighting implementations, and can decide ourselves which we prefer.

How can you compare Star Citizen to The Division for example, both have strong points and negatives but there's not much common between the two because of the different style, gameplay, scope etc.

It'll be a pure technical breakdown. Things like aesthetics, a huge part of a game, will be ignored. There'll also be issues of style (eg. cel shading compared to realism). But we gotta start somewhere.

I'm at least interested in this but i don't know if the outcome of the thread will be actually useful.

What's the useful outcome any of the discussion here?! Certainly the eternal versus discussions aren't useful. In that respect, this idea is useful to give more depth to future graphics discussions and make meaningful comparisons and following state-of-the-art realtime graphics developments less haphazard.

 

[Shifty Geezer]

I think the idea is that each technical aspect should be treated separately and be supported by screenshots that show "state of the art" and "worst possible". Nothing should score more than 5. That way, the scores have room to scale upwards.

Definitely screenshots are required. I think I'd scale 10 as an example of best in class now, which can be reevaluated when a new title comes out and the scale renormalised. It's not like the numbers will stick and be recorded, in my mind. More a case of, "given this technique is at its best in Game X, Game Y achieves a reasonable result, I'll give it a 7."

I don't think a more objective score is possible.

 

[Shifty Geezer]

I get that.. but is that number an exaggeration? Are their games going up 32x from 1080p?

Yes, just a rhetorical example. I'm not sure where resolution comes into this discussion for UC, and it should be compared to IQ of console games. Its IQ is exceptional, but prone to some artefacts.

 

[milk]

Don't be a Debby-downer. It should be an article and start with UC4 (the most popular HOT game today). Having contributions from the graphics programmers in these boards and us all coming up with an overall GRAPHICS review score could potentially make Beyond3d the de facto standard for the internet to get proper unbiased reviews. Who knows what could become of this?

Let's calm ourselves down a little bit shall we?

 

[VFX_Veteran]

So who is voting for artcle and who is voting for thread?

If most of us want a thread, I'll make the article on my own website and base it off the thread. I really think Beyond3d having an article though will be epic! Just like I go to eurogamer.net for DF articles, or NX Gamer's videos (he's such a fanboy..LOL), people would come to Beyond3d for the graphics tech reviews..

 

[Jawed]

I think it's worth pointing out that such a discussion of all of these technologies in contemporary real time game graphics is one that development studios, themselves, must undergo. "Which technologies are we going to include?" "What artistic controls do we want to offer?" "How can we excite the graphics nerds?" etc.

 

[Jupiter]

Crysis 2 Tesselation - rocks, soil (texture mod is used)
Tesselation (2) by X-RAY-89, auf Flickr

Tesselation (3) by X-RAY-89, auf Flickr

Tesselation (4) by X-RAY-89, auf Flickr

Tesselation (1) by X-RAY-89, auf Flickr

 

[Clukos]

I think it's worth pointing out that such a discussion of all of these technologies in contemporary real time game graphics is one that development studios, themselves, must undergo. "Which technologies are we going to include?" "What artistic controls do we want to offer?" "How can we excite the graphics nerds?" etc.

But yeah, that is true. And the weighting system of what is important changes from game to game according to the needs of the original vision. Some games don't need TOD, some don't even have water. It's very hard to actually reach a consensus with this discussion about every game, i'm sure it's hard enough already doing that for one game.

 

[VFX_Veteran]

But yeah, that is true. And the weighting system of what is important changes from game to game according to the needs of the original vision. Some games don't need TOD, some don't even have water. It's very hard to actually reach a consensus with this discussion about every game, i'm sure it's hard enough already doing that for one game.

You are correct. Which is why they won't be scored on those. TOD should be roped into lighting/shadowing system. Water, agree..

For a game that doesn't have or trying to showcase these particular features, they can be labeled as N/A. And not factored into the overall score. I'm going to start my article but I really think Shifty is a much-much better writer than me. Shifty - talk to me.

-M

 

[Jawed]

So, it's pretty simple: if a game doesn't do something, it doesn't appear in the discussion.

I think a discussion about which game has the best graphics is something else: because games don't attempt everything, generally. I don't think there's much value in scoring games against each other, using the bulk of their scores. The best you could do is to say something like "this game attempts, A, B, D, R, Q and Z and fails miserably at B but the rest is near state of the art."

 

[VFX_Veteran]

Crysis 2 Tesselation - rocks, soil (texture mod is used)
Tesselation (2) by X-RAY-89, auf Flickr

Tesselation (3) by X-RAY-89, auf Flickr

Tesselation (4) by X-RAY-89, auf Flickr

Tesselation (1) by X-RAY-89, auf Flickr

Jupiter,

It might have lots of tessellation but it's not used to good effect. It looks like it's nothing but a noise heightfield which the rock has been displaced into (i.e. I don't think it's using the same texture as the rock itself). Truth be told, I haven't really seen a good tessellation use in a game yet. RoTR has fairly good tessellation, but it fails at footprints.. LOL!

 

[Clukos]

Yeah i don't think numbering makes sense in this case, you'd need only one person doing the math because else it could really get subjective, inconsistent and in the end not informative at all. Don't get me wrong, i like the concept in theory, i just don't think it can actually work in practice. What Jawed says is much easier to accomplish.

 

[VFX_Veteran]

So, it's pretty simple: if a game doesn't do something, it doesn't appear in the discussion.

I think a discussion about which game has the best graphics is something else: because games don't attempt everything, generally. I don't think there's much value in scoring games against each other, using the bulk of their scores. The best you could do is to say something like "this game attempts, A, B, D, R, Q and Z and fails miserably at B but the rest is near state of the art."

Well, if you think of it like a regular review score, then you can get a sense for what game tries to hit the most marks with quality results -- hence, GOTY candidates. We are just going deeper into the graphics category than ever done before.