Current Generation Games Analysis Technical Discussion [2020-2021] [XBSX|S, PS5, PC]
- Thread starter Shortbread
- Start date
- Status
Here's another comparison -- from your screenshots, ps5 in the left is missing shading on the background tree, more pixelated clouds, blue artifact..... could this be "primitive shaders" making it look worse???? (Or maybe the shots aren't 100% identical, and this is grasping at straws!)
Yeah this is completely image reconstruction artifacts.
Johnny Awesome
Veteran
It was also pointed out that LOD can be the reason sometimes for these differences, if you happen to capture the screen at an LOD boundary on one console, but not the other for instance.
Silent_Buddha
Legend
Yup, cherry picking of frames in order to try to make one platform artificially seem better than the other. Too much of that happening in this thread.
Regards,
SB
This seems to be a bit of a trend across a bunch of recent titles and I'm wondering why. This has to be something that is wholly in the control of the developer, you can be as optimistic or conservative as you like regarding your prediction for your target resolution fitting in your frame time.
So weird.
Starting to wonder if it's a mandate from Microsoft? Why else would all 3P games be following the same trend here?
But for this game in particular, yeah it does look like there are some differences besides just more frame drops on Series X.
[PS5]
[PC/SERIES X]
Apparently this is somewhat a regular thing between the two versions of the game; also cloth physics supposedly being worst on PS5. Wondering if others can confirm this. And I peeped some of the screenshots above; is it a standard case that the textures are higher-res on PS5? Or is it a case of the textures being slower to load on Series X? Because IIRC that was something of an issue on the Series X version of Control WRT some of the ray-traced settings.
The Mesh/Primitive Shader talk was really enlightening to read through, quite a lot of it giving me a better understanding of the real differences between them. Will definitely help going forward on that topic. One thing tho is some keep thinking Series X doesn't have Primitive Shaders, but I'm pretty sure Microsoft confirmed it does in their Hot Chips presentation:
They list the Mesh Shaders as part of their Geometry Engine (this is as it is on PC) but also mention distributed Primitives. I just take that to refer to Primitive Shaders but if I'm wrong someone can correct me. Like a few others were saying, though, very hard to picture Microsoft would remove Primitive Shaders altogether since it's a part of RDNA 2 at its core.
If devs aren't even using Mesh Shaders yet, assuming games on PS5 are just passively using its Primitive Shaders, would those games not be almost impossible to run on Series if they had no Primitive Shaders whatsoever and yet devs also aren't leveraging the Mesh Shaders yet?
You know it's funny looking at these up-close sometimes because in this context I almost think I'm looking at Saturn or PS1 games xD. I know these are cropped and extremely zoomed-in shots way beyond original internal or external resolutions but I can't help it
The dramatic lighting differences are interesting -- makes me think the two builds are in a kinda different place (it is a 'demo' afterall) and a lot could be subject to change on both sides. Not sure I trust isolated internet screenshots though, maybe the ps5 ones were taken before they load in a second later or something.
I wonder where the red orange light stripes on the right side are in the XSX version. They seem to be completely missing.
@cwjs's theory on the two builds possibly being in different states might be on a good path. Even though this game comes out very soon, these builds are probably months old. If they're leveraging the last-gen versions to build the next-gen (current-gen) ones, there's probably more recompile work needed on PS5's side especially with upgrades involved, even though it's a native port (I'm thinking these are native in terms of being recompiled for the new platforms, not developed from scratch on them).
those look like particle effects to me, like someone is doing a welding job and stuff is sort of flying down.
Should be easy enough to test by comparing the quality of textures close to the camera. If they look identical then the assets would be the same. Since we know the difference can be seen at distance, if the asset quality is the same up close and different at distance, then it's likely filtering or texture LOD bias.
Nope the comparisons have being done by VGTech. He is always doing identical comparisons, he is the best at this. But you maybe right by saying it could be a difference in reconstruction method.
Reconstruction seems to be done using 2 different methods here. It's much sharper on PS5 or they did apply a blur filter on XSX, but would be the point?. It could explain indeed the seamingly low resolution textures on XSX.
It reminds me Resident evil 2 that also looked higher res on Pro even if they game was supposedly running at higher res on One X.
Here is a comparison I have done using VGTech images (the textures are close to the camera). Even the non-oblique textures look blurry so I don't think it's AF.
The size given was for the original release for last gen. Ie right from the start the PlayStation code base for whatever reason has has a smaller install.
Because of this I am not sure we can look at the next gen size and say because it's smaller they used oodle or something over the Xbox release outside of whatever was already in place.
Are we discussing different things or different angles?
That's zoomed in and far from the camera, though. If you have differences in filtering or LOD bias that's where it will show up. Near the camera you should see the full quality of the texture map to determine if they are different. If they are not, then it's a filtering of LOD difference that's making the textures look worse at distance. Also, as others have mentioned, a blur filter or other forms of AA during reconstruction can also destroy texture quality.Here is a comparison I have done using VGTech images (the textures are close to the camera). Even the non-oblique textures look blurry so I don't think it's AF.
I think it could come from very different reconstruction methods. Because it's not only textures, as seen in a previous post, this blur is also seen on geometry.
Was this dig necessary? VG Tech is pretty good with their analysis, even John over at DF recognized their work as being authentic and fair.
I believe those are all oblique. The ground is definitely always going to be oblique (unless you look straight down). The wood is aliased if you zoom in meaning oblique. (it's also pointing away from you)Nope the comparisons have being done by VGTech. He is always doing identical comparisons, he is the best at this. But you maybe right by saying it could be a difference in reconstruction method.
Reconstruction seems to be done using 2 different methods here. It's much sharper on PS5 or they did apply a blur filter on XSX, but would be the point?. It could explain indeed the seamingly low resolution textures on XSX.
It reminds me Resident evil 2 that also looked higher res on Pro even if they game was supposedly running at higher res on One X.
Here is a comparison I have done using VGTech images (the textures are close to the camera). Even the non-oblique textures look blurry so I don't think it's AF.
These are often signs of lower quality texture filtering. Some engines get hit harder with texture filtering than others. Not sure why, maybe someone else can pipe in, but I've seen AF not matter at all, to mattering a ton. This just might be one of those situations.
Last edited:
- Status
Similar threads
- Locked
