• Hey, guest user. Hope you're enjoying NeoGAF! Have you considered registering for an account? Come join us and add your take to the daily discourse.

[Digital Foundry] The Last of Us Part 1/Part 2 - PS5 Pro Tech Review + PSSR vs PC DLSS/FSR 2 Face-Off



We've seen preview code for one of these games, but now we've got the full lowdown on both 'Parts' of The Last of Us, with image quality and performance testing, VRR base PS5 vs Pro unlocked frame-rate analysis, plus a look at how PSSR compares in TLOUp1 on Pro against the PC version running FSR 2 and Nvidia DLSS.

Moderating GIF
 
Last edited:

Kerotan

Member
So if you have VRR 120hz you should play the performance Pro mode? I played a lot of different TLOUP2 chapters and TLOUP1 prologue and it's mind blowing good. I just can't remember the setting.
 

Radical_3d

Member
I started playing Part I today as part of my catch up with the backlog. Sucks not to be able to know the best experience between fidelity vs Pro is, one Digital Foundry video later. I guess I’ll keep with the pro…
 
Last edited:

Zathalus

Member
Its against the narrative.. they need the 5080 to release first! They don't want to embarrass RTX 4090/80
The video of this thread is very positive. The previous video from DF was about Stellar Blade was extremely positive as well. The direct tomorrow covers Callisto Protocol and is positive and praises the implementation. Pre-release they released half a dozen video that were praising the Pro.

Stop having this bizzare prosecution complex on behalf of the Pro. It’s silly.

Didnt't stop them with alan woke, and it's not even mediocre
Too true. It’s not mediocre at all and is one of the better games of last year. Glad we agree.
 
Last edited:

Gaiff

SBI’s Resident Gaslighter
Didnt't stop them with alan woke, and it's not even mediocre
Alan Wake is much more recent and does a lot more visually than Callisto Protcol.
The video of this thread is very positive. The previous video from DF was about Stellar Blade was extremely positive as well. The direct tomorrow covers Callisto Protocol and is positive and praises the implementation. Pre-release they released half a dozen video that were praising the Pro.
See how the goalposts keep moving? At first, they were claiming DF was avoiding comparisons between DLSS and PSSR. When DF did them and concluded that DLSS was still the best solution, it moved to DF purposefully covering bad games to undermine the Pro. When DF covered 2 good games in a row, including a lot of other good ones such as Spider-Man 2 and F1 24, it moved to DF not wanting to cover Callisto Protocol.

The bottom line is, those guys aren’t interested in how the Pro fares or in DF’s coverage. They just want console war ammo to battle against PC gaming. Unless the Pro clearly wins, they won’t be happy and they think they’ve found an obvious winner with Callisto Protocol (hint, it’s not what they think it is).
 

PandaOk

Member
Black woman in a prominent role.
Season 5 Nbc GIF by The Office

Checks out sadly.
cClWqaT.jpeg

Add the graphical noise and glitches, this is not good enough.
Imo in TLOU II PSSR makes the game look worse. Really not a fan of a good amount of these implementations. What’s the use of sharper if it’s frequently exposing graphical errors and worse stability, because honestly that just makes everything look more gamey and even lower res (that’s what we normally associate things like pixel crawl with).
 
Last edited:
DLSS / DLAA in "the last of us remake" looks amazing on the static image (like a downsampled image), however I saw ghosting during motion on the darker colors on certain occasions (maybe different DLSS preset would fix this ghosting issue, but I only tested default "E" preset).

Oliver says that DLSS is more blurry during motion than PSSR, but it depends on whether you play with a gamepad or a mouse. When I played on K+M, there was absolutely no motion blur with DLSS because my mouse movements were too fast and DLSS was not increasing AA to improve temporal stability. However, if I moved the camera very slowly on the gamepad, DLSS would apply AA and I could definitely see a slight reduction in sharpness. TAA did not have this problem, so I played at TAA native.
 
Last edited:

Bojji

Member
LOU-1.gif





LOU-2.gif





LOU-3.gif





LOU-4.gif





LOU-5.gif





LOU-6.gif



It's interesting to see how DLSS is sharper when stationary, while PSSR has the clarity in motion.










LOU-7.gif



It's also interesting to see how people perceive over sharpened image with lesser fine detail as something better including the DF's own staff :messenger_tears_of_joy: :messenger_tears_of_joy: :messenger_tears_of_joy: :messenger_tears_of_joy:

They used minimal sharpening with both fsr and dlss in this comparison. You can always increase it in game.
 

SlimySnake

Flashless at the Golden Globes
When I played the game, i thought Fidelity mode looked sharper but honestly PS5 Pro mode was 99% there. the thing is fidelity mode is already close to 50 fps and might even hit 60 fps at times in smaller areas so i didnt see a big reason to switch to pssr, but honestly, this is the best pssr implementation out there next to demon souls so ignore all the nitpicking.

PSSR having a 3.5 ms cost is in line with my findings with tlou2. this whole 2ms nonsense was just the best case scenario. even dlss is not 2ms. maybe in some games, but these AI upscaling solutions definitely have a cost and a lot of it is probably due to the fact that some of the effects still have to run at 4k.

In my testing, the difference between the unlocked ps5 mode and pro mode was around 25%. definitely not a cpu limit because it can easily hit 120 fps in some of the smaller areas and most of tests were basically in an empty field or a corridor. DF's findings line up with what we've seen in other games with 30-35% more performance. again, 45% in the leaks was the best case scenario.

i honestly dont know how you can release a console thats just 30-35% more powerful. especially for $700 but maybe they looked at nvidia pricing and said to themselves, why the fuck not.
 

mckmas8808

Mckmaster uses MasterCard to buy Slave drives
When I played the game, i thought Fidelity mode looked sharper but honestly PS5 Pro mode was 99% there. the thing is fidelity mode is already close to 50 fps and might even hit 60 fps at times in smaller areas so i didnt see a big reason to switch to pssr, but honestly, this is the best pssr implementation out there next to demon souls so ignore all the nitpicking.

PSSR having a 3.5 ms cost is in line with my findings with tlou2. this whole 2ms nonsense was just the best case scenario. even dlss is not 2ms. maybe in some games, but these AI upscaling solutions definitely have a cost and a lot of it is probably due to the fact that some of the effects still have to run at 4k.

In my testing, the difference between the unlocked ps5 mode and pro mode was around 25%. definitely not a cpu limit because it can easily hit 120 fps in some of the smaller areas and most of tests were basically in an empty field or a corridor. DF's findings line up with what we've seen in other games with 30-35% more performance. again, 45% in the leaks was the best case scenario.

i honestly dont know how you can release a console thats just 30-35% more powerful. especially for $700 but maybe they looked at nvidia pricing and said to themselves, why the fuck not.

Because the 35% is frame rate only. You're not including what the image looks like.
 

sachos

Member
Its interesting how the Pro mode looks better than original Fidelity in Part 2 but for Part 1 is a bit more debatable. I still prefer it over the Part 1 Fidelity, it looks too over sharpened imo.
I really hope Sony can improve on that noise pattern for PSSR, even when upscaling clean non RT image from 1440p it is still there (although much less so).
 
Top Bottom