Image Quality and Framebuffer Speculations for WIP/alpha/beta/E3 games *Read the first post*

Im no expert in pixel counting, but that looks lower than 900p, has that been confirmed?
Someone should be able to work it out from that rope on the top shot
I see from arwin's site the previous game was 1600x900
I had a look at the previous game and it does look higher resolution

\edit: looked a bit more at some screenshots, and it looks like the texture resolution is really really low and thats causing it to look so bad
 
Last edited by a moderator:
Im no expert in pixel counting, but that looks lower than 900p, has that been confirmed?
Someone should be able to work it out from that rope on the top shot
I see from arwin's site the previous game was 1600x900
I had a look at the previous game and it does look higher resolution

\edit: looked a bit more at some screenshots, and it looks like the texture resolution is really really low and thats causing it to look so bad

The problem doesn't come from the resolution of textures which aren't worse, IMO, than similar open world games lilke AC4 or Shadow of Mordor. The problem comes from the cheap heavy post AA + cheap vaseline effect + cheap hardware upscaling.

A cheap, i mean very cheap, I meant really badly cheap solution has never worked very well in any projects whatsoever.

In order to remove 1% of image problems (the jaggies), they effectively destroyed by collateral damages 99% of the rest of the image... And we can still see the jaggies! :oops:

I somehow doubt there is still a morphological or a temporal component in this post AA. Some long edges are not anymore properly anti aliased and normally any temporal component will give a supersampled look to the game like with KZ, AC4, Infamous, Driveclub or even Quantum break.

Here this 900p game (verified at least vertically) doesn't look supersampled at all. Quite the contrary. It looks similar (in clarity) to a great looking 720p game like UC3.

EDIT: Obviously no morphological AA component here:
gUgb.png
 
Last edited by a moderator:
Vaseline? :???:
:) You are giving Cristiano Ronaldo's secrets away.

The other day in the Barça-Real Madrid match there was an American (you know americans still aren't that good at football --soccer) commentating the match on TV (stream) and he went like this when talking about Cristiano Ronaldo:

"It is as prickly as a porcupine's backside." "He's as slippery as an eel covered in Vaseline." :LOL::mrgreen::LOL::LOL:
 
Last edited by a moderator:
The problem doesn't come from the resolution of textures which aren't worse, IMO, than similar open world games lilke AC4 or Shadow of Mordor. The problem comes from the cheap heavy post AA + cheap vaseline effect + cheap hardware upscaling.

A cheap, i mean very cheap, I meant really badly cheap solution has never worked very well in any projects whatsoever.

In order to remove 1% of image problems (the jaggies), they effectively destroyed by collateral damages 99% of the rest of the image... And we can still see the jaggies! :oops:

I somehow doubt there is still a morphological or a temporal component in this post AA. Some long edges are not anymore properly anti aliased and normally any temporal component will give a supersampled look to the game like with KZ, AC4, Infamous, Driveclub or even Quantum break.

Here this 900p game (verified at least vertically) doesn't look supersampled at all. Quite the contrary. It looks similar (in clarity) to a great looking 720p game like UC3.
Now that you mention it... what if the PS4 is just meant for 1080p and it doesn't use an upscaler other than some built-in very basic solution? I mean, that might explain why Sony are pushing 1080p so much.

It's also known that Assassin's Creed developers will just go with what they got instead of creating an upscaling solution themselves --remember ACIV initial controversy with upscaling on Xbox One.

If both games are 900p, I wonder how the Xbox One version will look like -FX aside, where the PS4 should have the edge- 'cos the hardware upscaler has been programmed again and improved.
 
Now that you mention it... what if the PS4 is just meant for 1080p and it doesn't use an upscaler other than some built-in very basic solution? I mean, that might explain why Sony are pushing 1080p so much.

It's also known that Assassin's Creed developers will just go with what they got instead of creating an upscaling solution themselves --remember ACIV initial controversy with upscaling on Xbox One.

If both games are 900p, I wonder how the Xbox One version will look like -FX aside, where the PS4 should have the edge- 'cos the hardware upscaler has been programmed again and improved.

No. There is no real proof that the hardware upscaler has being improved in any way. When some XB1 games use a custom software upscaling solution like Ryse and probably Quantum break, then yes, the upscaling is definitely better than PS4 hardware solution. But this custom software upscaler eats ressources and is only used in a few exclusive games.

If there was a high quality free hardware upscaler in XB1 then it would be used by all sub-1080p multiplats. It's not the case and all those XB1 multiplats use a cheap upscaling solution very similar to PS4 upscaler.

And the XB1 hardware upscaling solution (after the patch which got rid of the sharpen effect) is virtually identical in the PS4 upscaler quality from what I could tell when I compared both upscaling on several games. I assume that when they got rid of the nasty sharpen effect they simply bypassed the hardware upscaler by using the AMD hardware upscaler instead, which is exactly the same for all this family of AMD GPUs.
 
Now that you mention it... what if the PS4 is just meant for 1080p and it doesn't use an upscaler other than some built-in very basic solution? I mean, that might explain why Sony are pushing 1080p so much.

It's also known that Assassin's Creed developers will just go with what they got instead of creating an upscaling solution themselves --remember ACIV initial controversy with upscaling on Xbox One.

If both games are 900p, I wonder how the Xbox One version will look like -FX aside, where the PS4 should have the edge- 'cos the hardware upscaler has been programmed again and improved.

The quality of the upscaler won't matter much unless the game has really nice AA anyway. Heck a simple bilinear upscale might be better if the AA is poor quality.
 
Image Quality and Framebuffer Speculations for Unreleased Games *Read the fir...

The previous AC? Nope, a quick google says a day 1 patch made AC IV 1080 on PS4...was 900 on X1


And so does techingames.net - version 1.02 on PS4 updated to 1920x1080.

AC IV was a cross platform game though, so definitely less demanding. But at 30fps I feel the PS4 should still be able to manage 1080p ...
 
No. There is no real proof that the hardware upscaler has being improved in any way. When some XB1 games use a custom software upscaling solution like Ryse and probably Quantum break, then yes, the upscaling is definitely better than PS4 hardware solution. But this custom software upscaler eats ressources and is only used in a few exclusive games.

If there was a high quality free hardware upscaler in XB1 then it would be used by all sub-1080p multiplats. It's not the case and all those XB1 multiplats use a cheap upscaling solution very similar to PS4 upscaler.

And the XB1 hardware upscaling solution (after the patch which got rid of the sharpen effect) is virtually identical in the PS4 upscaler quality from what I could tell when I compared both upscaling on several games. I assume that when they got rid of the nasty sharpen effect they simply bypassed the hardware upscaler by using the AMD hardware upscaler instead, which is exactly the same for all this family of AMD GPUs.

SSOD uses software upscaling according to the latest post in the Xbox hardware scaler thread. The guy from insomniac goes on to say that the hardware scaler had been improved thru an Sdk update. (Sharpen filter removed as well as a fix for crushed blacks)
I am not sure what the difference between both systems on chip scalers is if there is any difference, but I seriously doubt that they just bypassed the custom one in the X1.
Also why would there be another "standard AMD scaler" on the gpu? These are custom chips after all. Why would they put 2 scalers on die as opposed to just replacing the standard one with the revised custom version. That doesnt make alot of sense.
 
And so does techingames.net - version 1.02 on PS4 updated to 1920x1080.
Sorry when I quickly looked there I saw
ps4 1600x900 xbone 1600x900 and ignored the 3rd (I must of assumed that was PC)
If they update the game on the first day from 1600x900 to 1080p I'ld stick the 1080p number down cause thats currently the game's resolution
 
Sorry when I quickly looked there I saw
ps4 1600x900 xbone 1600x900 and ignored the 3rd (I must of assumed that was PC)
If they update the game on the first day from 1600x900 to 1080p I'ld stick the 1080p number down cause thats currently the game's resolution

Yeah, I am thinking about making a view that merges the latest data from different versions, perhaps showing the version number in there somewhere, and perhaps making that a proper matched table with empty values where none is given, but still lined out to exactly the same size etc. I've changed the sorting though so all version information is shown per platform for now.
 
No. There is no real proof that the hardware upscaler has being improved in any way. When some XB1 games use a custom software upscaling solution like Ryse and probably Quantum break, then yes, the upscaling is definitely better than PS4 hardware solution. But this custom software upscaler eats ressources and is only used in a few exclusive games.

If there was a high quality free hardware upscaler in XB1 then it would be used by all sub-1080p multiplats. It's not the case and all those XB1 multiplats use a cheap upscaling solution very similar to PS4 upscaler.

And the XB1 hardware upscaling solution (after the patch which got rid of the sharpen effect) is virtually identical in the PS4 upscaler quality from what I could tell when I compared both upscaling on several games. I assume that when they got rid of the nasty sharpen effect they simply bypassed the hardware upscaler by using the AMD hardware upscaler instead, which is exactly the same for all this family of AMD GPUs.
afaik, the Xbox One has been built with upscaling in mind from the get go, the engineers admitted that themselves -see Richard's interview on Eurogamer-. :smile2: (they knew the hardware isn't overly powerful)

Isn't the scaler programmable? I purchased AC IV when it came out and looked extremely sharp on my TV to the point of having the leafs of the trees flashing some kind of *aggressive* white dots -so difficult to explain-.

Now you play the game and without it being patched it looks more natural, Ubisoft did nothing at all. Same for other titles (Killer Instinct).
 
afaik, the Xbox One has been built with upscaling in mind from the get go, the engineers admitted that themselves -see Richard's interview on Eurogamer-. :smile2: (they knew the hardware isn't overly powerful)

Isn't the scaler programmable? I purchased AC IV when it came out and looked extremely sharp on my TV to the point of having the leafs of the trees flashing some kind of *aggressive* white dots -so difficult to explain-.

Now you play the game and without it being patched it looks more natural, Ubisoft did nothing at all. Same for other titles (Killer Instinct).

What you're describing is precisely the nasty sharpening filter in effect that has being later removed in all game with a XB1 patch.

How it has being "removed" is the real interesting question.
 
What you're describing is precisely the nasty sharpening filter in effect that has being later removed in all game with a XB1 patch.

How it has being "removed" is the real interesting question.
Yes, I've read it in this Digital Foundry article, some time ago.

http://www.eurogamer.net/articles/digitalfoundry-2014-has-microsoft-fixed-the-xbox-one-scaler

The question is how they did it, but it all points out to the fact that the hardware upscaler is programmable? So you create an algorithm, send it to the upscaler and it interprets/apply it.
 
Back
Top