Digital Foundry Article Technical Discussion [2018]

Status
Not open for further replies.
Spidey ought to have a heavily scratched out sketch of a spider tank in his room. -"But WHY?!"

:V
 
The eye tracking moving objects isn’t rethoric, that’s how our visual system works.
Frame rates and displays have grave limitations, so using band aids to alleviate those makes sense. The problem with motion blur is that without knowing what the viewer looks at, it will be percieved as wrong, causing irritation or even breaking immersion completely. As in giving the impression that your eye is malfunctioning rather than being a part of the limitations of the rendering tech.

In that sense it is similar to DOF effects, where not being able to focus wherever you want in the scene is bizarre. (Actually DOF in games is just wrong on many levels, not the least of which being that a normal eye focussed on a couple of meters actually perceives everything as sharp even without refocussing. Our FOV and pupil aperture takes care of that.)

People simply find this immersion breaking as it conflicts with how vision works in general.

There is a real problem that motion blur is trying to adress, but I’d contend that I’d rather see band-aids being used that doesn’t introduce jarring artifacts of their own. And of course, what we really should be doing is attack the problem at its source.
I understand all that rationale, I didn't call it rethoric because its wrong, I called it so because its incomplete. It ignores frame persistance. Thanks to that, often ignored detail, motion will always be perceived as wrong by your eyes, with or without motion blur, with or withouy eye tracking.
You see an object on screen, you ajust your eyes where you expect it to be on the next instant, but the next frame has not been displayed yet, and the current one is still onscreen, ok so the object is now static, oh wait, another couple miliseconds passe, it did move, ok lets ajust for where it will be in the next 8ms... Oh, 8ms later its still in the same place, it will take another 8 until it teleports to the next position (assuming 60fps)... This is the effect of high persistence monitors (all consumer level monitors really) and it feels like judder, and is unavoidable with or without MB, but MB does make it feel less jarring.
What devs try to do is avoid trying to simulate the world as seen through the eyes altogether, and instead emulate how it feels like if you are seeing the game through a movie camera, since we are used to the same things you mentioned (camera relative MB, non adjustable DOF, lens flare, etc) on that context, and mostly everybody can endure it perfectly well in that context.
 
I understand all that rationale, I didn't call it rethoric because its wrong, I called it so because its incomplete. It ignores frame persistance. Thanks to that, often ignored detail, motion will always be perceived as wrong by your eyes, with or without motion blur, with or withouy eye tracking.
You see an object on screen, you ajust your eyes where you expect it to be on the next instant, but the next frame has not been displayed yet, and the current one is still onscreen, ok so the object is now static, oh wait, another couple miliseconds passe, it did move, ok lets ajust for where it will be in the next 8ms... Oh, 8ms later its still in the same place, it will take another 8 until it teleports to the next position (assuming 60fps)... This is the effect of high persistence monitors (all consumer level monitors really) and it feels like judder, and is unavoidable with or without MB, but MB does make it feel less jarring.
What devs try to do is avoid trying to simulate the world as seen through the eyes altogether, and instead emulate how it feels like if you are seeing the game through a movie camera, since we are used to the same things you mentioned (camera relative MB, non adjustable DOF, lens flare, etc) on that context, and mostly everybody can endure it perfectly well in that context.
But movies aren’t games. (And movies aren’t shot with 1970s equipment any longer, so adding old film camera artifacts is just weird.)
I play a barbarian walking down the mountain side, sweat chilling my bare manly chest as I have just heroically defeated the dreaded snowwyrm. I gaze down the valley as the sun sets over the mountain range and my vision is filled with a lens flare from a lens that has six aperture blades, and apparently has six or so uncoated (pre WW2?) lens surfaces so it might be a tele tessar rather than a sonnar design...- wait a minute WTF??? Why are my eyes filled with lens flares, what the hell?

It’s immersion breaking.
 
You're not seeing this happen as it happens through your eyes. You're watching your GoPro footage you captured when you had these adventures many years ago and shared them with the world on YouTube.
Arrrgghh!
:)
Seriously though, gaming typically has identification at its core. YOU are the protagonist, regardless of perspective, even in the days of ASCII graphics when you were a "@" sneaking down a dungeons levels in mortal fear of "D":s.
You are not the audience watching a story as told by script and director. Movies are visual books in that regard. Games are fundamentally different, realistically rendered or not. And even realistically rendered, that which detracts from identification/immersion is problematic.
 
I'd argue not all games have you as the protagonist. In plenty of them I can't relate at all to the protagonist! But you're definitely right for a lot of games which have this identity crisis problem, where the game is centred on you in a role whereas the art is centred on you as a spectator or cameraman. The moment you go third-person though, the logic would be to have you as a cameraman or bystander. I guess the artistry actually needs recognition of its own language rather than just borrowing from movies and understanding what's needed to achieve player presence.
 
Danger Zone 2 is the next step toward the Burnout successor we've been waiting for
Crash-testing the new game from Criterion's founders.

https://www.eurogamer.net/articles/...o-the-burnout-successor-weve-been-waiting-for

The spiritual successor to Burnout we've been waiting for is coming - but perhaps not quite in the way you might have imagined. Danger Zone 2, a current-gen revamp of Burnout's crash mode, arrives on July 13th on PlayStation 4, Xbox One and PC, courtesy of Three Fields Entertainment. Meanwhile, further on down the road, Dangerous Driving is the circuit-based Burnout successor that follows it, arriving as soon as this winter - and developed by the same close-knit team of just seven people.


Also present is UE4's recently introduced dynamic resolution scaler, aimed at trading pixels for performance where needed. However, on Xbox One X at least, optimisation was based on a native 4K output, with DRS only added at the end of development. Physics are entirely CPU-driven, so in common with Dangerous Golf and the first Danger Zone, this is a 30Hz experience on all consoles bar Xbox One X, which has an optional unlocked performance mode that while not perfectly consistent, effectively trades physical resolution for temporal, offering a smoother ride and an experience that is closer to the feel of the original Burnout.
 
https://www.eurogamer.net/articles/...host-of-tsushima-gameplay-debut-tech-analysis

Ghost of Tsushima's E3 debut analysed: a stunning late-gen showcase
Sucker Punch bookends the PlayStation 4 era in style.

It may have been a Sony E3 media briefing short on new announcements, but there was no shortage of spectacle. Alongside The Last of Us Part 2, the gameplay debut for Ghost of Tsushima took us to the war-torn grasslands of 13th century Japan, depicted via a masterclass in real-time rendering, animation and physics simulation. Looking at its stunning opening vista shot, you'd be forgiven for thinking this is running on future Sony hardware - a prototype PlayStation 5, perhaps. At first glance, the environment animation, particle effects and lighting look a generation ahead - and certainly it's a big stylistic jump from the studio's previous work on InFamous First Light. But as the end credit tile reveals, the demo runs on hardware you may already own. It's a PS4 Pro, and so, the real surprise is the technical ingenuity going on in order to achieve such great results on existing console technology.
 
Amazing graphics, intresting to know how it will look on base ps4. Almost bought a ps4 slim last week, days of play edition (blue colour) that i thought looks cool. Then again the Pro is just $50 more with one less controller, but hey its for the SP games anyway. Going to settle for a Pro, 1080p tv but some games do downsampling.
As a pc gamer ps4 is the first after ps2 thats this attractive :)
 
My launch PS4 was dying on its arse, so I traded it in for a Pro, in spite of only having a 1080p TV. Worth it IMO: downsampling has gotten rid of shimmer, and some games have a 1080p mode with other enhancements.

God of War was just perfect at 1080p60ish :love:
 
Im getting one for the exclusives but some of the cross plat games like wolfenstein 2 run well on the ps4/pro too, wolfenstein doesnt run that well on older gpus thx to older support in drivers. I could upgrade the tv room pc with just a gpu (gtx 970 or better) but that wont let me play those Sony exclusives, with a pro it comes close to a 970 in multiplats.
 
I see Uncharted - Ancient Japan edition.

Edit: I was most impressed with the quality of the dynamic fire in the leaves. It looks authentic and behaves realistically. That and the leaf shading in the tree.
You know what, an Uncharted game set in the ancient time wouldn't be such a bad idea tho.
Yeah they're throwing a lot of techs in this demo, SSS, volumetric cloud with dynamic simulation, volumetric lighting, the god damn field tech, 300k of leaves blowing with interaction, dynamic TOD and weather, high quality Bokeh DOF and the pristine character models. Game is a stunner.
 
https://www.eurogamer.net/articles/...host-of-tsushima-gameplay-debut-tech-analysis

Ghost of Tsushima's E3 debut analysed: a stunning late-gen showcase
Sucker Punch bookends the PlayStation 4 era in style.

It may have been a Sony E3 media briefing short on new announcements, but there was no shortage of spectacle. Alongside The Last of Us Part 2, the gameplay debut for Ghost of Tsushima took us to the war-torn grasslands of 13th century Japan, depicted via a masterclass in real-time rendering, animation and physics simulation. Looking at its stunning opening vista shot, you'd be forgiven for thinking this is running on future Sony hardware - a prototype PlayStation 5, perhaps. At first glance, the environment animation, particle effects and lighting look a generation ahead - and certainly it's a big stylistic jump from the studio's previous work on InFamous First Light. But as the end credit tile reveals, the demo runs on hardware you may already own. It's a PS4 Pro, and so, the real surprise is the technical ingenuity going on in order to achieve such great results on existing console technology.
The game looks good, but no :LOL:
 
Ghost analysis
DF going gaga about how next gen it looks and got 1800p CBR from the video. I mean, the Witcher 3 footage in there literally looked like a joke next to it.

The game physics is on a whole other level. I haven't been this impressed with game physics since the first time I laid my hands on the TAC Gun (especially the modified version) in Crysis. :LOL:
 
The game looks good, but no :LOL:
Well the thing is that the game looks significantly better than everything else bar TLOU2, Death Stranding or Cyber Punk? That it doesn't belong in this gen even though true next gen could be yet again significantly better:).
 
Status
Not open for further replies.
Back
Top