Digital Foundry: Days Gone Remastered DF Review - PS5/PS5 Pro - An Improvement Over PS4 Back Compat? It's Complicated

adamsapple

Or is it just one of Phil's balls in my throat?




Overlooked by many at launch - but still genuinely a superb game in our view - Days Gone Remastered is a fascinating PS5/Pro release. It contains many of the PC version's upgrades - sometimes improved - and adds new visual effects and graphics modes. But to what extent is image quality an improvement over the PS4 version running on PS5 at 60fps under back compat plus? We had lots of questions on this - and Oliver has answers.



00:00 Overview
00:48 Lighting and Settings Changes
06:15 PS5 Modes
10:15 PS5 Pro Modes
12:32 Observations and Conclusion
 
Last edited:
Bored Cabin Fever GIF


Patiently waiting for adamsapple adamsapple summary so I don't have to watch the video.
 
Last edited:
- Overhauls lighting, especially at night time.
- The original game was very bright at night, now it's darker.
- SSGI adapted from the PC release, subtle but impactful, mostly indoors.
- Shadow resolution has been improved, 2x in some places, but can have oddities like fences leaving pin sharp shadows at long distances.

- Game now has an FoV toggle, but now the default is lower than the original release.
- Cut-scene FoV is also lowered, they always appear slightly zoomed in now.
- Gameplay FoV set to 76 matches the original PS4 release.

- OG PS4 Pro version ran at CB 2160p at 60fps in back compat mode.
- New release's Performance mode runs at 1440p with no up-sampling. Can look soft and 'imprecise' compared to PS4 BC mode.
- Performance mode also sees Dynamic resolution scaling.
- Image stability is comparable but is softer / less sharper.

- PS4 BC version was mostly 60 but sees notable drops in large freaker hords and occsassional small drops otherwise.
- The new version sees less drops and VRR benefits where it was disengaged in the PS4 version.
- Load times are also faster. 42 vs 32 seconds at initial boot.

- Quality mode is native 4K. Excels at depicting fine foliage.
- But targets and runs at 30fps.
- 'Decreae in fluidity won't be worth it for most players'

- PS5 Pro:
- 3 modes. Performance, Quality, Enhanced (uses PSSR)
- Performance: Same as PS5 with the only difference is DRS 1800p instead of 1440p.
- In practice the difference is slim and old PS4 code still remains sharper.
- Enhanced: 60fps, 1584p and uses PSSR.
- Can expose low resolution edges during movement and blurs out foliage, plus can have distracting noise and shimmer.
- It is much sharper than the basic Performance mode, close to the OG CB mode.
- Quality mode runs at 5K 2880p and downscales to 4K
- Looks margianally cleaner than PS5 Quality, the differences are subtle. 'The two are mostly indistinguisable'.
- Locked at 30 fps but DF suspect more players will pick the 60fps modes.
 
Last edited:
I don't read many of these PS Pro comparisons, but it seems playing the old PS4 version on the base PS5 is usually the best way to play.
Pro adds weird artifacts that makes me think the Pro isn't worth getting.
 
So technically the PS4 version on PS5 is the superior version by the looks of it. What a fucking joke gaming has become
Its absolutely not.

Checkerboard is dog shit, and old PS4 code had bad pop in.

Game is running on ultra now, don't understand why that doesn't seem to get attention.
 
- Can expose low resolution edges during movement and blurs out foliage, plus can have distracting noise and shimmer.
They are talking about the enhanced mode? And they have provide a clip about such stuff? I searched in the video but honestly can't find it. The only weird thing that I noticed it's how much blurried is the native 1800p. It's absolutely bizzarre and not normal to be so much blurried compared 4k cbr. I'm quite confuse here to what going on.
Anyway it's quite curious now this guy of DF suddenly noticed so many flaws on the PSSR in motion versus the CBR but with FSR2 compared the PSSR, it preferred the first one because in still was better. How the subjective perspective can change sometimes.
 
Last edited:
So basically 2160 CBR looks basically the "same" as 2880p downsampled to 4k...

Uh-huh. That sounds super-convincing, funny they didn't rave about the IQ at the time...

Here's the absolute bottom-line, the only thing that the old version has going for it is sharpness. Everything else is worse, and yet its "complicated".
How does this argument hold any water?
 
Last edited:
So basically 2160 CBR looks basically the "same" as 2880p downsampled to 4k...

Uh-huh. That sounds super-convincing, funny they didn't rave about the IQ at the time...
If it was the case something is broken in the ps5 pro version. As the more jaggies artifacts noises statement in motion sound like quite bizzare. There is no way the CBR is less jaggy or noisy compared an AI upscaler quite the contrary. That's why I asked if someone can provide a clip or if it's showed in the video.
The only thing I can say personally the 1800p mode has definitely something of wrong, it's too much blurried and not seems normal to me to be so worse compared 4k cbr.
 
Last edited:
If it was the case something is broken in the ps5 pro version. As the more jaggies artifacts noises statement in motion it's quite bizzare. There is no way the CBR is less jaggy or noisy compared the AI upscaler That's why I asked if someone can provide a clip or if it's showed in the video.
The only thing I can say personally the 1800p mode has definitely something of wrong, it's too much blurried and not seems normal to me to be so worse compared 4k cbr.
CBR 4K should be as sharp as native 1800p... on a native 1800p screen. The problem is that you get a basic bilinear upscaling from 1800p to 4K that further spoils and blurs the image. This is why upscaling works better when done up to full 4K, even if you start really really low like 1080p. This is how they do with PSSR and DLSS, whatever the native resolution, they always upscale to 4K to avoid some ugly bilinear upscaling they way it was done on PS4 (most games on PS3/360 got native resolutions so they mostly avoided a crude upscaling done by the GPU).

The fact that they are doing that on PS5 is really amateurish. This not lazy developers, it's incompetent developers.

We can finally see that CBR when done well to 4K really rocks!
 
Last edited:
CBR 4K should be as sharp as native 1800p... on a native 1800p screen. The problem is that you get a basic bilinear upscaling from 1800p to 4K that further spoils and blurs the image. This is why upscaling works better when done up to full 4K, even if you start really really low like 1080p. This is how they do with PSSR and DLSS, whatever the native resolution, they always upscale to 4K to avoid some ugly bilinear upscaling they way it was done on PS4 (most games on PS3/360 got native resolutions so they mostly avoided a crude upscaling done by the GPU).

The fact that they are doing that on PS5 is really amateurish. This not lazy developers, it's incompetent developers.

We can finally see that CBR when done well to 4K really rocks!
So that's the bilinear upscaling who blurred so much, I get it. Anyway I really really doubt in motion CBR it's more precise over the PSSR. Seems extremely unlikely and tech wise almost impossible to achieve, outside they have screwed something big during the coding. What Oliver said there seems pure crazyness to me. Again there is a clip about such stuff? I stopped to give credit to the personal take from awhile, especially from DF.
 
Last edited:
yeah I have played the original checkerboard solution

is oliver blind? all the artifacts??

no way it looks better imo. or close
From my personal experience Oliver is often incredibly harsh when he talks about the PSSR flaws. I'm the first who said there are some flaws but no way CBR offers less noises, artifacts or jaggies over the PSSR, seems a real hyperbole. Outside maybe extremely bad coding. He should at least provide a deeper analysis about such stuff and why not try to contact the developers because it shouldn't happen and not ordinary.
 
Last edited:
It's really good in a few games that I remember. This is one, another is Sekiro.
Can be good but no way can be better of the PSSR especially in motion. RE4R at 868p with PSSR (120fps mode) put in the shame 4k CBR on the same ps5 pro. I'm not joking.
 
Last edited:
yeah I have played the original checkerboard solution

is oliver blind? all the artifacts??

no way it looks better imo. or close


The 'better' part they say multiple times is related to IQ. The CB version looks sharper than either PS5 versions Performance mode.
 
I see we can now add the word "vintage" to the growing list of Digital Foundry vocabulary that desperately needs its own remastering. I'd probably pay $10 to never have to hear "besoke", "interesting" and "in the here and now" ever again.
 
sharper =/ better

CBR is old and has a lot of artifacts

Yes, it's very strictly in relation to sharpness / IQ, not visual effects or fidelity or pop-in.

There are upgrades but maybe they're not as transformative as some of the other 'remasters', going by the video, IMO. I'm not gonna put the $10 for this, the original seems good enough as is.
 
sharper =/ better

CBR is old and has a lot of artifacts

well implemented CBR looks better than nearly any "modern" reconstruction method.
a reminder that Dark Souls remastered's CBR was so convincing that both Digital Foundry and VG Tech didn't notice it is being used at all. the game looked like a slightly soft native 1800p in motion.

the advantage of DLSS or PSSR is not really quality, but quality in relation to the upscale factor. CBR can only ever really create the illusion of a 2x resolution boost, while the more modern ones like DLSS can go up to 4x and still look decent.

so the advantage isn't quality, it's performance.
 
Last edited:
well implemented CBR looks better than nearly any "modern" reconstruction method.
a reminder that Dark Souls remastered's CBR was so convincing that both Digital Foundry and VG Tech didn't notice it is being used at all. the game looked like a slightly soft native 1800p in motion.

the advantage of DLSS or PSSR is not really quality, but quality in relation to the upscale factor. CBR can only ever really create the illusion of a 2x resolution boost, while the more modern ones can go up to 4x and still look decent.

so the advantage isn't quality, it's performance.
how is that determined though?

I've seen a lot of artifacts on CBR implementations

Ghost of Tsushima is pretty sharp but has obvious lines in the grass, GT Sport was a mid implementation that had a lot of sharper than normal edges, all the resident evil games have obvious cb artifacts.

Days Gone has artifacts on the grass in movement and also on the trees when very far from the player. I'm just not a fan of it
 
how is that determined though?

I've seen a lot of artifacts on CBR implementations

Ghost of Tsushima is pretty sharp but has obvious lines in the grass, GT Sport was a mid implementation that had a lot of sharper than normal edges, all the resident evil games have obvious cb artifacts.

Days Gone has artifacts on the grass in movement and also on the trees when very far from the player. I'm just not a fan of it

all of this can also happen with DLSS or PSSR. just like other issues that CBR typically doesn't have.
you see in the video how the PSSR version blurs moving grass for example, an issue not found with CBR.

the resident evil games also look really clean considering it's not a native image. especially if you compare it with the same games running with FSR2, which looks awful.
so clearly CBR is superior to FSR2/3 at the very least.
 
Last edited:
The 'better' part they say multiple times is related to IQ. The CB version looks sharper than either PS5 versions Performance mode.
So I checked the ps4 pro version because I was really curious. The only right part was about the sharpness: game is sharper on ps4 pro BUT just in still and in the close distance details. The rest is uncorrected and artifacts in motions are more prominent with CBR especially if you look at the long distance, jaggies included. I can guarantee it. Oliver proved again to be absolutely shallow when he talks about PSSR, as expected it's more his personal stance about such stuff than a tech verdict.
 
Last edited:
all of this can also happen with DLSS or PSSR. just like other issues that CBR typically doesn't have.
you see in the video how the PSSR version blurs moving grass for example, an issue not found with CBR.

the resident evil games also look really clean considering it's not a native image. especially if you compare it with the same games running with FSR2, which looks awful.
so clearly CBR is superior to FSR2/3 at the very least.
Sorry but it's not. 1440p to 4k with FSR2/3 is a league ahead to 4k CBR. Not saying CBR can't be good but you can't have a better reconstruction with a simpler/dated upscaler method.
 
Last edited:
Sorry but it's not. 1440p to 4k with FSR2/3 is a league ahead to 4k CBR. Not saying CBR can't be good but you can't have a better reconstruction with a simpler/dated upscaler method.

Ok but in this game's case it's not using any up-sampling in the bare Performance mode, the pics a few posts above show the difference in clarity in near / mid field.
 
Ok but in this game's case it's not using any up-sampling in the bare Performance mode, the pics a few posts above show the difference in clarity in near / mid field.
Yes I noticed immediately performance mode to be strangely blurried even on the ps5 pro. I barely touch it (I play it with the enhanced mode ) and I noticed the blurriness thanks to the DF videos, to be honest. Still the update it's very rough and with too many weird bug/glitches here and there. I'm expect a further patch because this release actually appears still unpolished.
 
Last edited:
Yeag OG DG on PS5 looks best and you are able to use GOAT controller DS4, however the remaster has Horde Assault and FOV settings.

Interesting that they lowered default FOV. I noticed that the middle setting 70 looked too zoomed in. But going from 76 to 85-90 is a good setting which makes it easier when fighting large hordes. For ambush camps and indoors I prefer 70-76
 
Yeag OG DG on PS5 looks best and you are able to use GOAT controller DS4, however the remaster has Horde Assault and FOV settings.

Interesting that they lowered default FOV. I noticed that the middle setting 70 looked too zoomed in. But going from 76 to 85-90 is a good setting which makes it easier when fighting large hordes. For ambush camps and indoors I prefer 70-76
LOD is notable improved and not just in the long distance, maybe it's that the reason. Speaking which I can't believe the DF guy said the enhanced mode has more artifacts than the ps4 pro. It's like to say prebaked lighting are more precise than raytracing. DF reliability is really getting worse the last years.
 
Last edited:
The only way I know what differences beyond the game looking darker there are between PS4 Days Gone running on my PS5 with the 60 fps patch and the native PS5 Days Gone Remaster is by watching YouTube comparison videos such as the Digital Foundry one. Otherwise, the game looks exactly as I remember it, which is still great but still underwhelming considering the 'remastered' label.

It does beg the question of what was the point of the remaster. Couldn't the PS4 game have just been updated to include the new lighting for PS5/PS5 Pro via a patch? It's not like the game really benefits much from running on PS5 in my opinion as the asset and texture quality remain the same. Also, the game looks blurry from the comparison video on PS5 versus the base PS4 version running on PS5 which really should never happen. Checkerboard rendering could be surprisingly good, arguably superior to the results we get from simple upscaling from a lower resolution using anything other than PSSR (which has its own issues anyway in the form of visual noise).
 
Game that didn't need to exist turns out to be wholly unnecessary

And yet, you feel the need to post about it. /facepalm

No joke, the endless bitching about this release is straight-up demented and sad.

Like, what exactly is the problem?

Is Days Gone a really good game? Yes, it is, and that's a pretty much solid consensus view 6 years on.

So the "Remaster" is PS5 only designation and is just a $10 patch on PC. It still brings the PS5 version into visual feature parity with the PC port, which is also pretty well regarded as looking and running good without making excessive hardware demands.

On top of that, does it add new content. Yes it does. People have been bitching for years that they can't experience horde combat until the late stages of the game, so you'd think giving people a full-on arcade sub-mode to do just would be welcomed... but oh no!

Like, literally the only possible complaint is that buying it all-in-one digitally at $40 is a bit expensive. Which it really isn't considering its a pretty long game with a ton of extras and replayability that still looks good by 2025 standards.

And even accepting that, that's a minority issue when millions have already bought the game.

I've never heard so much retarded pearl-clutching over a thing that by all objective measures is really good in my life.
 
It does beg the question of what was the point of the remaster. Couldn't the PS4 game have just been updated to include the new lighting for PS5/PS5 Pro via a patch? It's not like the game really benefits much from running on PS5 in my opinion as the asset and texture quality remain the same. Also, the game looks blurry from the comparison video on PS5 versus the base PS4 version running on PS5 which really should never happen. Checkerboard rendering could be surprisingly good, arguably superior to the results we get from simple upscaling from a lower resolution using anything other than PSSR (which has its own issues anyway in the form of visual noise).
If you patch the PS4 game you run the risk of PS4 players running the new version and experiencing worse perf.
 
Top Bottom