Current Generation Games Analysis Technical Discussion [2022] [XBSX|S, PS5, PC]
- Thread starter BRiT
- Start date
nailed it.
Silent_Buddha
Legend
I still don't get the point of ignoring motion in games when saying X game looks similar on this hardware versus that hardware. It's like...
One person - "Hey, how did you get your game to look that good, mine doesn't look that good?"
Their friend - "Stop moving when you are playing the game then it'll look the same."
One person - "Oh hey thanks, I thought something was wrong with my game."
Their friend - "You're welcome."
One person - "Ummmm, how do I play the game?"
Their friend - "Well, start moving."
One person - "But then my game doesn't look as good as yours."
Their friend - "That's not the point of graphics in games, graphics in games is all about standing still and not doing anything."
One person - "Oh, I've been doing it all wrong all this time by moving around in the game while I'm playing it."
Regards,
SB
One person - "Hey, how did you get your game to look that good, mine doesn't look that good?"
Their friend - "Stop moving when you are playing the game then it'll look the same."
One person - "Oh hey thanks, I thought something was wrong with my game."
Their friend - "You're welcome."
One person - "Ummmm, how do I play the game?"
Their friend - "Well, start moving."
One person - "But then my game doesn't look as good as yours."
Their friend - "That's not the point of graphics in games, graphics in games is all about standing still and not doing anything."
One person - "Oh, I've been doing it all wrong all this time by moving around in the game while I'm playing it."
Regards,
SB
Which is impressive given the poor quality of the pictures for comparison purposes and the fact that this is Twitter so most people were probably looking on their phones.
Dunning–Kruger effect - Wikipedia
I am right because i am right, now i am gonna prove you that i am right with this test! Even thou its obvious to everyone that i am wrong, it has only strengthen my confidence that i am right.
My friend used to say, if one person tells you "you have tail" ignore it, but if 2 or 3 persons tells you "you have tail" maybe its time to look behind and check.
Last edited:
It's a bit weird. There's a very relevant point to make that even if more people could accurately tell the difference in a side by side, whether they'd actually care in practice is a very different question. Especially for a console gamer who has no alternative option to even compare directly with personally. There's plenty of merit in 'good enough' on that front.
But clearly his agenda here is beyond this and anybody who knows him will understand that right away. He's simply deflecting in an effort to avoid admitting he was wrong about something.
But clearly his agenda here is beyond this and anybody who knows him will understand that right away. He's simply deflecting in an effort to avoid admitting he was wrong about something.
If his point were "resolution doesn't matter much, you get basically the same great looking game no matter what" I would be more sympathetic, but his point is "actually, it's good that I mislead viewers".
If you're not good at or interested in reviewing resolution differences why not just cut it from videos? I personally don't think a little softness matters much at all, I'd gladly watch a reviewer who praised low resolution games for saving that frame budget for rendering interesting effects.
Plenty of serious tech reviewers ignore or downplay certain components of an image, too. Digital foundry doesn't really care much about motion reconstruction artifacts like ghosting (at least not compared to me -- not trying to put words in their mouths), and will sooner criticize a game for not having reconstruction than they'll praise it for having a stable artifact free (low resolution) image. That's fine! It's fine to have editorial opinions about what "good graphics" means. NXG doesn't really believe resolution doesn't matter though, he believes the ps5 is magic.
If you're not good at or interested in reviewing resolution differences why not just cut it from videos? I personally don't think a little softness matters much at all, I'd gladly watch a reviewer who praised low resolution games for saving that frame budget for rendering interesting effects.
Plenty of serious tech reviewers ignore or downplay certain components of an image, too. Digital foundry doesn't really care much about motion reconstruction artifacts like ghosting (at least not compared to me -- not trying to put words in their mouths), and will sooner criticize a game for not having reconstruction than they'll praise it for having a stable artifact free (low resolution) image. That's fine! It's fine to have editorial opinions about what "good graphics" means. NXG doesn't really believe resolution doesn't matter though, he believes the ps5 is magic.
betternatethanlever
Newcomer
Lighting looks really good. I love the soft diffuse shadows the game is doing. Does anyone know what they're doing for these shadows? Seems to only apply to dynamic objects and characters. Other environments seem to use baked/static shadows. So it's definitely not ray traced indirect shadows.
It looks very convincing for the most part especially when every other game, even AAA games, look so flat in indirect lighting with static AO blobs underneath everything. Metro and Control are the only games I've played that use convincing RT indirect shadows.
Capsule Shadows I presume:
https://docs.unrealengine.com/4.26/en-US/BuildingWorlds/LightingAndShadows/CapsuleShadows/Overview/
Originally added for Paragon in those areas of the screen in shadow for all the minions and characters
https://docs.unrealengine.com/4.26/en-US/BuildingWorlds/LightingAndShadows/CapsuleShadows/Overview/
Originally added for Paragon in those areas of the screen in shadow for all the minions and characters
davis.anthony
Veteran
DX11 is the default API. I doubt the game was optimized for DX12 at all, so there's no point in comparing them.
davis.anthony
Veteran
Last edited by a moderator:
betternatethanlever
Newcomer
Interesting. I assume Epic has patented this as it's the only reason I can think of it not being used in other games. Last of Us on PS3 had these so it can't be that computationally expensive. It looks a 1000x better than any AO implementation, at least for characters and dynamic objects. There's very few things that irk me more than when a game has no indirect shadows in obviously indirect GI lighting, instead they either have an ugly black halo surrounding the character or blob at the bottom due to the AO that looks totally out of place.
Flappy Pannus
Veteran
You have to force a command line parameter to launch in DX12. It's not exposed in the UI, it doesn't default to it, no one will launch it in DX12 without specifically googling how to do it.
Yes, the performance is brutal in DX12, but it's clearly not intended to be run in that mode. From what I can tell it's doing ray traced reflections, but they may not even be RTX accelerated with how abysmally slow it is. Maybe something that is coming in a future patch, but currently I don't see the point in spending that much critique on a mode that the developers clearly don't intend the public to be using (your cat's body reflection is completely static for example when moving, it's clearly broken).
When/if they actually make this mode visible and indicate it's an actual option, sure.
Speaking of the PC version, the dev has acknowledged the shader compilation stutters (reminder this is not just a DX12 issue!), and a patch is planned. From ArsTechnica's Review:
From my experience with the game the shader compilation stutters are actually not that heavy, most are in the first 30 minutes. At least on my system, the performance issues seem to stem more from (once again) a wonky vsync implementation, which thankfully being DX11 at least you can fix. There are moments in the game where I was dropping below 60fps with 80% GPU utilization, forcing Fast Sync clear this up - of course if you're running a VRR display you won't have to deal with this. Also some basic UE traversal stutters, and a poor optimization of some alpha effects it seems - there is a small alpha smoke effect for example that occurs from your robot companion when you overload them after firing a weapon for an extended period that can take up an additional ~40% GPU usage when it gets close to the camera which is bizarre.
Otherwise though, playing at 90% 4K scaling with slightly higher-than-PS5 settings (going from the Analista video) gets me 60+ fps in most areas of the game on my 3060. So perhaps not a stellar port (no DLSS, booo) but not sure how in the heck that SkillUp review got those abysmal performance figures on a 2080TI (50fps at 1440p!) - judging by my results, 4k maxxed should be giving you 70-80fps on that gpu.
Last edited:
betternatethanlever
Newcomer
Is there a reason why more games don't use capsule shadows? It looks so convincing without all the ugly artifacts of SSAO.
I always wonder how that's possible with games that are using UE4. How is it possible to not integrate DLSS into your game?
Do the developers not know about it? That's the only reason I can think of. Literally, every layman can integrate DLSS in less than 5 minutes with UE, it's as simple as downloading DLSS, move folder to UE4 folder and then click enable. That's it.
Similar threads
- Locked