• 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.

God of War Ragnarök PC - DF Tech Review - PC vs PS5 - Optimised Settings

rodrigolfp

Haptic Gamepads 4 Life
Yeah there's a HFR mode, normally between 80-90.
MAIN.00_20_09_41.Still026.jpg
Then I will correct my mistake.
 

hinch7

Member
TLDW for folks who can't be bothered go through the video, its a quick and dirty PC port but runs reasonably well so long as you don't use Zen 2 processor.

Quite a few visual/performance bugs and looks like they just slapped on DLSS without much thought. At least they kinda band-aid patched the tesselation issue.
 
Last edited:
There are 4 modes on PS5.

Performance Mode: 60fps cap, resolution varies between 1440p-2160p with an average of 1872p. Settings unknown besides that shadows are between Low and Medium.

Quality Mode: 30fps cap. Resolution locked to 4K. Settings comparable to PC's Ultra.

If you have a 120Hz screen. You get two additional modes. A 40Hz mode, and a high frame rate mode.

High Frame Rate Mode: Unlocked fps that varies between 70-90fps. Resolution stays at 1440p. Same settings as Performance Mode.

40Hz Quality Mode: 40fps cap. Resolution varies between 1800p to 4K. Similar settings to Quality Mode.



You forgot the most important one

Quality + HFR unlocked (VRR)

This will offer the best IQ on PS5 Pro


It's like the 40fps quality mode but uncapped

Just with Boost mode it will be amazing to see

I think Santa Monica will patch the game anyway to use PSSR
 
Last edited:

Gaiff

SBI’s Resident Gaslighter
You forgot the most important one

Quality + HFR unlocked (VRR)

This will offer the best IQ on PS5 Pro


It's like the 40fps mode but uncapped

Just with Boost mode it will be amazing to watch
There is a 5th mode? That must have been added later.
 

Zathalus

Member
@7:00 He compares PS4 using lowest settings vs PC using max settings to make the case that the PS5 SSD i/o isn't contributing to the faster asset loading.

Anyone else see the problem with that logic?

Also hilarious that he deviated from his traditional practice of showing PC optimized vs PS5 Performance mode settings and head to head framerate comparison.
You see the same behaviour turning the settings down all the way. Alex probably should have shown that as well, but to be honest the difference in quality isn’t this vast gulf that a PS4 has no problems with LOD loading and a 4090 does.

I recall the Ratchet & Clank comparison with the longer loading (or not loading at all) texture issue which was apparently proof of the SSD I/O superiority and was then promptly patched to be fixed the very next week. Or the slightly longer portal loading scenes which were found out to be artificially capped so the engine could perform JIT shader compilation.
 

DanielG165

Member
Any gaffers want to give impressions on this game? Not the perf so much, but the actual game itself. Me, I like 2016 well enough, but am not over the moon about the new direction like some people. I liked it well enough to buy a sequel, I've just heard some mixed reviews, especially from people who felt the same as me about 2016.
I find 2018 to be the better, tighter game. Ragnarok felt like a slog by comparison.
 

RoboFu

One of the green rats
I didn't watch but does he even test amd cards? Or is he still jacking off to his 4090 even when it has issues? 🤣
 

Gaiff

SBI’s Resident Gaslighter
he's is unprofessional, if you make tech and test, test it on every hardware, not only one.
Video was very lackluster. Here are optimized settings and here is a 5-second clip of them running on a 4060. Couldn't he show a full sequence and how they scale on low-end, mid-range, and high-end hardware? What's the general performance like? I get it that he isn't the biggest PlayStation fan, but he should have just let Oliver do that one. We learned very little from this video and as you've pointed out, he had a few misses, attributing to bugs what is a problem with the interaction between DLSS and the engine.
 

hinch7

Member
he's is unprofessional, if you make tech and test, test it on every hardware, not only one.
That would take an unsurmountable amount of time. And most who are into gaming on PC that play up-to-date titles; will have an Nvidia GPU. And will have access to RTX features as well as AMD ones.
 
Last edited:

Senua

Gold Member
I know everyone here shits on this game but I'm still gonna give it a go. Hopefully the DLSS issue will be solved soon
 

Bojji

Member
Sounds like a horrible port, anytime there are missing effects and features you failed your mission.

Yep it looks like a half working mess. Every port that doesn't have nixxes involved is shit for a few months.

Nixxes have their problems (VRAM management!) but they are much better.
 

squidilix

Member
Zen 2 CPUs (or at least the 3600) run terribly in this game. The 5600X, which is usually around 30-40% faster than the 3600, performs up to 99% better in Ragnarok. The 7800X3D, which is generally 100% faster, performs up to 247% faster in Ragnarok. Performance metrics reveal that the 3600 is poorly utilized as many of its threads have low usage. Could be a bug
This is the most interesting thing when you know the I/O complex on PS5 have this own processor for decompression, And the 3D Audio too with the Tempest Engine. So base CPU on PS5 are less struggle.
Of course, this is working if devs really want to push the PS5, mainly not happen with all shit port multisupport.
 
Last edited:

TGO

Hype Train conductor. Works harder than it steams.
Pointless comparison, the Pro update will probably make this obsolete.
 

RoboFu

One of the green rats
That’s not really what DF does. If you want a wide variety of benchmarks across multiple GPUs Hardware Unboxed is probably what you’re after.
it should be if they are talking about missing effects that is only missing on one brand of card and not the other. :messenger_dizzy:
 

hinch7

Member
PS5 here: 80fps.
Probably poor optimisation and CPU utilisation on the port. Alex did say said there is something off with the 3600 results. The 3600 shouldn't be that far behind in gaming (not way over 2x) and he gave examples in other games.
 
Last edited:

Stuart360

Member
Yeah the tessalation setting is bugged. I was getting 60fps on my 1080ti, but as soon as i entered that portal thing where you run along the tree, fps was like 15fps lol. Just in that portal area.
I messed around and tested each graphic setting and found it was the tessalation setting. Putting that down to medium and it was back to 60fps.
Pretty werid really.

I do have to say i like the game a lot more than the orig GOW. I didnt even ffinish that one but i'm already 18 hours into Ragnarok.
 

Bojji

Member
This is the most interesting thing when you know the I/O complex on PS5 have this own processor for decompression, And the 3D Audio too with the Tempest Engine. So base CPU on PS5 are less struggle.
Of course, this is working if devs really want to push the PS5, mainly not happen with all shit port multisupport.

This game runs up to 60fps on a fucking jaguar so it's not really at all CPU limited on PS5. Something is just fucked with the PC code:

9x1cvwD.jpeg




The CPU doesn’t matter



Fucked up comparison, performance of 3600 is completely abnormal.

Normally 3600 is less than half of 7800x3d so it should be 80-90fps.
 

Xyphie

Member
not PC, it's again NV DLSS

I can't seem to replicate this, getting fog more like the FSR picture in the intro sequence (I assume it's the same as there's credits on screen?). Is it the DLSS FG (can't test that because Ampere) causing it or just the upscaling? I replaced whatever the shipping version of DLSS in the game was with 3.7.20 before even launching it though, so maybe that helps if someone else is having that issue.

g1gogGo.jpeg

NRN34fs.jpeg
 

Skifi28

Member
I know everyone here shits on this game but I'm still gonna give it a go. Hopefully the DLSS issue will be solved soon
GAF tends to overreact. There's probably a bit too much Atreus and a few sections are very slow and kill the pacing, but the overall game is quite good with improvements to combat, enemy variety and bosses, exploration etc.
 

squidilix

Member
This game runs up to 60fps on a fucking jaguar so it's not really at all CPU limited on PS5

I dont say is CPU limited on PS5, I just say they using the PS5 CPU with goodsense and leave some task (I/O loading - Audio) with the rest of hardware. And the game can reach 90fps en PS5.

Again, PS4 version have not the same assets than the PS5 version, and the PS4 Pro build running on 1080p on Performance Mode and rarely achieved 60fps (more 45fps).

People expected what? Running this game with the potatoes 750ti / i3 4th gen?
 

Stuart360

Member
Its not even that cpu intensive. On my 3700x, cpu usage is around 15-25%, and the game only uses a few cores, with most cores/threads being inactive.
Its almost like the game only uses 5 cores as thats what the game was designed around with the PS4.
 

Gaiff

SBI’s Resident Gaslighter
The game has obvious issues with the thread usage on the 3600.

K3gO92N.png


You have a bunch that are practically idle. Average usage across the 12 threads on the 3600 is 35%. It's 58% on the 5600X.
 

ChiefDada

Gold Member
This would be the perfect mode to compare PS5 against PC as it's static 1440p in that mode.

Absolutely. Alex typically hides behind DRS when comparing PS5 vs PC. But now we have a game with fixed resolution and uncapped framerate mode, a la TLOU Pt.1 and I guess he doesn't want to go through that trauma again lol. We all know why he went with this peculiar comparison method.
 

Zathalus

Member
Is this true?

LIVDvZM.jpeg
Not really, the port devs oversold it in the interview.

Absolutely. Alex typically hides behind DRS when comparing PS5 vs PC. But now we have a game with fixed resolution and uncapped framerate mode, a la TLOU Pt.1 and I guess he doesn't want to go through that trauma again lol. We all know why he went with this peculiar comparison method.
Would it even be a valid comparison with all the performance bugs and missing visual parity? Plenty of games out there, some even Sony ports, that have far less issues, and have been compared to the PS5.
 

Nex240

Neo Member
@7:00 He compares PS4 using lowest settings vs PC using max settings to make the case that the PS5 SSD i/o isn't contributing to the faster asset loading.

Anyone else see the problem with that logic?

Also hilarious that he deviated from his traditional practice of showing PC optimized vs PS5 Performance mode settings and head to head framerate comparison.
Because the textures on PS4/PS5 are pretty similar, it wouldn't make a difference really.
 

Nex240

Neo Member
I can't seem to replicate this, getting fog more like the FSR picture in the intro sequence (I assume it's the same as there's credits on screen?). Is it the DLSS FG (can't test that because Ampere) causing it or just the upscaling? I replaced whatever the shipping version of DLSS in the game was with 3.7.20 before even launching it though, so maybe that helps if someone else is having that issue.

g1gogGo.jpeg

NRN34fs.jpeg
Did they just break the fog effect in a recent patch?
 

Gaiff

SBI’s Resident Gaslighter
Would it even be a valid comparison with all the performance bugs and missing visual parity? Plenty of games out there, some even Sony ports, that have far less issues, and have been compared to the PS5.
The biggest issue are the shadows. 70fps on Medium. 86fps on Low. PS5 runs between Medium and Low. Impossible to match. We can probably fill in the gaps, but this could skew the performance. Obviously, setting it to Low on PC wouldn't make sense, but Medium could also be a lot more demanding.
 
Last edited:

Xyphie

Member
try look also after cave, unless it's 4xxx series bug

I've played until
the Thor boss fight
and I can't seem to replicate it? Tried switching between TAAU/XeSS/FSR/DLSS and I can't really tell. Certainly nothing like your screenshot which is a night and day difference with no fog whatsoever thus far.

Did they just break the fog effect in a recent patch?

It's whatever the latest patch is. Patch 3 per Steam.
 
Last edited:
Top Bottom