• Hey Guest. Check out your NeoGAF Wrapped 2025 results here!

Watch Dogs 2 PC performance thread

-Mouse code is excellent. Feels like raw input, and feels like control latency has been prioritized in development. Even when the fps goes lower (I tested around without TF etc) the input still feels lag free and snappy. First class job here as well.
That's nice, too many games don't pay enough attention to that.

PCGH did a CPU driver overhead test with two Radeons vs. one Geforce, well the results are quite shocking in this game:

cpu3auh0.jpg

http://www.pcgameshardware.de/Watch-Dogs-2-Spiel-55550/Specials/Test-Review-Benchmark-1214553/

Notice that the rendering resolution is basically at 180p (720p + pixel-densitiy at 0,25).
This is purely a CPU throughput test or how good the driver distributes work to the GPU.
Nvidia is getting 50% more performance out of the same processor as AMD.
That's only shocking if you weren't paying any attention in the past 5 years. The results are very much in-line with what you would expect on AMD CPU-wise in a demanding DX11 game.
 
Probably powering the DedSec operation trying to shut down Vivendi :^)

DudeRandom84 has some side by side video comparisons of PC and Pro btw. Most apparent thing to jump out with YT compression is shadows, they look pretty crappy on Pro and "jump" quite irritatingly to adjust to ToD/lightning.

https://m.youtube.com/watch?v=sO1fMsgjeh4

Pro's AA looks better but PC's shadows are higher res. Pretty good showing for the Pro considering.
 
Pro's AA looks better but PC's shadows are higher res. Pretty good showing for the Pro considering.

hes using HTFS so yes shadows are monumentally better. that setting isnt playable unless youre at 1080p on any single gpu tho. aa on the pro does seem better than any of the non-txaa options
 
It's remarkable how close the PS4 Pro is to the PC version. As I was saying a few pages back, aliasing is less pronounced (even if we get a softer image as a result), and besides the shadow resolution, they're pretty much on par during gameplay.
 
It's remarkable how close the PS4 Pro is to the PC version. As I was saying a few pages back, aliasing is less pronounced (even if we get a softer image as a result), and besides the shadow resolution, they're pretty much on par during gameplay.

Close to what settings? What resolution is it really running at? What AO methode is it using? Is a pc with similar specs able to pull out the same performance? Also the good old 30 vs 60 fps thing.
Those are important things to consider when comparing ports/versions and just comparing YT videos isn't enough here. DF used to do that, but lately they are missing out a bit.

highly unlikely temporal filtering isnt being used to hit that 1080p on ps4

The same thought crossed my mind, but either DF missed it or its way better on ps4 then on pc.
 
Close to what settings? What resolution is it really running at? What AO methode is it using? Is a pc with similar specs able to pull out the same performance? Also the good old 30 vs 60 fps thing.
Those are important things to consider when comparing ports/versions and just comparing YT videos isn't enough here. DF used to do that, but lately they are missing out a bit.

Close to the settings in the video above.

Pc settings were at max settings 1080p with HFTS Shadows and MSAAX4. The Texture pack was also installed. I used two GTX 1080's in SLI and an Elgato HD60 to record.

Now, that's a $1200+ PC (not talking about SLI) against a $400 console. 60 FPS aside, it IS remarkable.
 
Close to what settings? What resolution is it really running at? What AO methode is it using? Is a pc with similar specs able to pull out the same performance? Also the good old 30 vs 60 fps thing.
Those are important things to consider when comparing ports/versions and just comparing YT videos isn't enough here. DF used to do that, but lately they are missing out a bit.



The same thought crossed my mind, but either DF missed it or its way better on ps4 then on pc.

i dunno, i watched the high quality DF video comparing ps4 pro to ps4 and xbone and the IQ is abysmal on OG ps4. shimmering, fuzziness and edge artifacts everywhere
 
60 FPS aside
You say "60 FPS aside", as if that alone didn't require at least twice the CPU and twice the GPU performance.

Add other incredibly demanding (for good reason) settings like HTFS on top of that and I don't find the performance on any platform particularly remarkable -- everything runs like it should.
 
Close to the settings in the video above.



Now, that's a $1200+ PC (not talking about SLI) against a $400 console. 60 FPS aside, it IS remarkable.

So we are comparing a native 1080p +4x MSAA capture vs a downsampled temporal 1440p + fxaa or smaa solution and judging both based on YT compression. Okay.

Also comparing hfts shadow performance vs high/medium settings is just unreasonable.
 
So we are comparing a native 1080p +4x MSAA capture vs a downsampled temporal 1440p + fxaa or smaa solution and judging both based on YT compression. Okay.

Also comparing hfts shadow performance vs high/medium settings is just unreasonable.

the settings hes using for the pc version are often < 20 fps in the open world on a 1080. not even a realistic comparison to begin with
 
You say "60 FPS aside", as if that alone didn't require at least twice the CPU and twice the GPU performance.

Add other incredibly demanding (for good reason) settings like HTFS on top of that and I don't find the performance on any platform particularly remarkable -- everything runs like it should.

I have a PC as well, and I love it, don't get me wrong.

Well, my main point is about how it looks, not about how it plays. I think I'm safe to say that we have two main "groups" here, people who value IQ and people who value both - IQ and higher framerates.

All I'm saying is that the PS4 Pro delivers comparable IQ for a very good price point. I think that's quite easy to see.

So we are comparing a native 1080p +4x MSAA capture vs a downsampled temporal 1440p + fxaa or smaa solution and judging both based on YT compression. Okay.

Also comparing hfts shadow performance vs high/medium settings is just unreasonable.

I'm just expressing my opinion. I had the chance to try both versions of the game, the YT comparison is on point, IMO. And believe me, you quickly forget about HTFS when the framerate tanks - the option is there sure, but when it becomes unusable even when trying to run the game @ 60 FPS even on the fastest single GPU on the market, there's no point in using it.
 
I have a PC as well, and I love it, don't get me wrong.

Well, my main point is about how it looks, not about how it plays. I think I'm safe to say that we have two main "groups" here, people who value IQ and people who value both - IQ and higher framerates.

All I'm saying is that the PS4 Pro delivers comparable IQ for a very good price point. I think that's quite easy to see.



I'm just expressing my opinion. I had the chance to try both versions of the game, the YT comparison is on point, IMO.

Well YT is never on point when it comes to IQ.
 
DF says it runs at "up to 1800p checkerboarding" on ps4pro. no other info is given on resolution. probably have to wait for faceoff for further clarity.
 
-Mouse code is excellent. Feels like raw input, and feels like control latency has been prioritized in development. Even when the fps goes lower (I tested around without TF etc) the input still feels lag free and snappy. First class job here as well.
-

Not just the mouse code, the entire mouse and keyboard controls and configuration is excellent. The first game was awful, with terrible mouse input and aim assist that wouldn't go away even using a mouse that made it feel even worse. The M/KB controls on WD2 is like a complete revelation in comparison. They have keybinds for literally everything you could think of. They let you rebind everything. For things like Walk and sprint they let you choose between Hold and Toggle (Looking at you, Deux Ex). They even have a new binding for 'traffic speed' when using a vehicle so you aren't constantly maxing out the engine when tapping 'W', which I think every open world game with driving should have now.

It's like they actually had someone play the game with m/kb for an extended time and put in the effort to make it good and configurable to what people would want. Which is more than I can say for most PC ports of multiplat games where it feels like they were only tested on a gamepad.
 
found the issue for my stutters, with ultra textures the frametime jumps really high like 50-120. thats with a 8gb 390 <.<
 
Just going to jump back in and make sure people who are struggling with performance TURN OFF SCREEN SPACE REFLECTIONS.

50% performance hit on my GTX 1080 at 1440p. This is significantly more than the Nvidia tweak guide would have suggested.

I can keep everything else at Ultra now which is nice and the game runs super well.
 
This game is now locking up my Windows Explorer after exiting... I can't open any new folders, thumbnails won't load in already open folders, task manager will not come up..

Something is seriously messed up with this title. I've never had a game force me to reboot my PC to make it behave again.

This also happened to me one time as well, I think I forgot to mention it, had to to do a hard reboot. At least now I know for sure what was at fault.



EDIT: As I also expected, the issue with the XBO S controller not working through BT is also tied to this shit anti-cheat system.

-eac_launcher

This will completely disable MP but apparently the controller will work as will RTSS and everything else affected by this garbage. I'd be willing to wager it's the cause of the freezing and BSODs too.
 
After all the division cheating it seems that ubi soft overreacted a bit in wd2. It's not like this is a highly competitive game in the first place (I actually like the invasions). But there are leader boards and some 'questionable' people will do anything to see their name on there...
 
Close to the settings in the video above.



Now, that's a $1200+ PC (not talking about SLI) against a $400 console. 60 FPS aside, it IS remarkable.

Well that's max settings performance with much higher image quality, you can always run graphical settings equivalent to the PS4 Pro version and get significantly higher levels of performance.

So we are comparing a native 1080p +4x MSAA capture vs a downsampled temporal 1440p + fxaa or smaa solution and judging both based on YT compression. Okay.

Also comparing hfts shadow performance vs high/medium settings is just unreasonable.

Digital Foundry are probably going to take a look at the game on PC, it will be interesting to see how it compares.

This also happened to me one time as well, I think I forgot to mention it, had to to do a hard reboot. At least now I know for sure what was at fault.



EDIT: As I also expected, the issue with the XBO S controller not working through BT is also tied to this shit anti-cheat system.

-eac_launcher

This will completely disable MP but apparently the controller will work as will RTSS and everything else affected by this garbage. I'd be willing to wager it's the cause of the freezing and BSODs too.

That's a real shame, hopefully these problems get worked out.
 
Just going to jump back in and make sure people who are struggling with performance TURN OFF SCREEN SPACE REFLECTIONS.

50% performance hit on my GTX 1080 at 1440p. This is significantly more than the Nvidia tweak guide would have suggested.

I can keep everything else at Ultra now which is nice and the game runs super well.

i have it on. i thought it meant all reflections but it seems it's just ground reflections and there is a separate setting for other reflections.

http://www.geforce.com/whats-new/gu...e-guide#watch-dogs-2-screen-space-reflections
 
You say "60 FPS aside", as if that alone didn't require at least twice the CPU and twice the GPU performance.

Add other incredibly demanding (for good reason) settings like HTFS on top of that and I don't find the performance on any platform particularly remarkable -- everything runs like it should.

Yeah seriously. PC gaming is not a box with set values. It's a hardware mix ranging from much weaker than PS4 to much more powerful. Comparable hardware nets comparable results unless there is something heinous working against that law. Console makers are not currently taking a loss on hardware. They use currently available technology like anyone building a PC. Their's no magic or help from aliens going on here.

That said, Watch Dogs 2 looks solid on PS4 Pro. It's a testament to what can be done on cheap hardware, console or PC. But let's not compare it with hardware that's far more capable. It's not fair to the Pro and it creates a silly dialogue.

One of the first things I heard screaming in my head after playing the PS4 Pro version extensively and coming to my PC was how much better the IQ was. And then go down the line for improvements in other areas.

Something I noticed on PS4 Pro is that the resolution is no where near the 1800p quality often mentioned in places like DF. That might be the output resolution but in action the resolution looked like 900p-1080p with some sort of soft AA. It was a good end result but I've been much more impressed with the IQ of other PS4 Pro games.
 

I feel that this is a mistake on DF's part where they are saying that "Checkerboarding up to full 4K is more demanding and requires half the basic resolution - a 1920x2160 buffer -". Notice that Cerny never says that and instead he explicitly says "we can create images of a higher quality than if our 4m colour samples were arranged in a rectangular grid...".

"A 1920x2160 buffer" is a rectangular grid, something which Cerny clearly states as missing in checkerboarding, and seems like something which DF has added from themselves to "explain" what "half the basic resolution" mean.

In practice it's most likely a "checkerboard" of quads rendered in a full 4K buffer with each even/odd omitted meaning that it's not a rectangular grid as you have holes in it which you have to fill either by extrapolating spatial data or accumulating temporal data adjusted by motion vectors. I don't see how else a "checkerboard" term is even applicable here.

Shcking? that they perform almost the same? It's pretty grounded and known

Yeah, but some people still think that GW are somehow unfair to Radeons for some mysterious reason. Notice that HBAO+ is actually faster on 480, a repeat of the same result from last year where it was faster on Fury cards.

PCGH did a CPU driver overhead test with two Radeons vs. one Geforce, well the results are quite shocking in this game:

cpu3auh0.jpg

http://www.pcgameshardware.de/Watch-Dogs-2-Spiel-55550/Specials/Test-Review-Benchmark-1214553/

Notice that the rendering resolution is basically at 180p (720p + pixel-densitiy at 0,25).
This is purely a CPU throughput test or how good the driver distributes work to the GPU.
Nvidia is getting 50% more performance out of the same processor as AMD.

Chances are that this won't be anywhere near this pronounced in an actual playable mode like 1080p/High.
 
I played with the settings yesterday and was able to achieve a very stable 30FPS on 4K high/ultra settings with the hi-res pack.

It's beautiful...

(i5 2500K 8Gb RAM with GTX 1070)
 
I've had three crashes so far all in different situations.

1. During the loading screen when going into a side mission.
2. When I was just walking around while doing a co-op mission with someone.
3. The instant I fired my gun about a minute after starting the game.

Everything will freeze, then the audio will come back for half a second, and then the game closes. The event viewer shows that the display driver stopped responding each time but I got no pop up stating that the driver had recovered like I've seen in past driver crashes.
 
This also happened to me one time as well, I think I forgot to mention it, had to to do a hard reboot. At least now I know for sure what was at fault.



EDIT: As I also expected, the issue with the XBO S controller not working through BT is also tied to this shit anti-cheat system.

-eac_launcher

This will completely disable MP but apparently the controller will work as will RTSS and everything else affected by this garbage. I'd be willing to wager it's the cause of the freezing and BSODs too.

So you can deactivate it via launch argument is what you're saying?
 
I've had three crashes so far all in different situations.

1. During the loading screen when going into a side mission.
2. When I was just walking around while doing a co-op mission with someone.
3. The instant I fired my gun about a minute after starting the game.

Everything will freeze, then the audio will come back for half a second, and then the game closes. The event viewer shows that the display driver stopped responding each time but I got no pop up stating that the driver had recovered like I've seen in past driver crashes.

If you don';t use shadowplay, try uninstalling geforce experience. That software used to make my computer crash all the time. Now I just install the driver and physX only. You can use FRAPS to monitor your framerate.
 
yeah im 90% sure my 3570k is the bottleneck here. Even changing the resolution can't net me 60fps in the busier areas of san fransico.

Shame. Best keep my eyes out for the 7700k in 2017.
 
It's remarkable how close the PS4 Pro is to the PC version. As I was saying a few pages back, aliasing is less pronounced (even if we get a softer image as a result), and besides the shadow resolution, they're pretty much on par during gameplay.

With the exception of framerate. Which of course is the big boost a pc can give over the console versions.
 
Yeah this is how it looks for me too. Atrocious.

Edit: Fixed it!!

It was due to having MFAA forced in Nvidia driver settings.

After a quick test before heading to work, you seem to be correct sir. I thank you for your help. I had MFAA on globally. This would be the first game that hasnt played nice with that option be it implemented or not.
Weird. ¯_(&#12484;)_/¯

I finally had to disable MFAA globally too, had a few games where it looked bad :
- Rainbow Six Siege (with temporal filtering)
- Steep (with temporal filtering)
- Forza Horizon 3 (but it looks and works great in Forza 6 Apex)

Now I set it game per game, which is kinda rare as a few games uses MSAA.
MFAA uses some temporal MSAA too, so it can conflict with games' own temporal solution.
 
I'll say this, super glad I invested in a GSync monitor.
Even though I'm hitting as low as 48 fps I haven't felt it at all.

same. This might actually be the first game that I felt about that tech being beneficial since I bought the monitor. It doesn't really help much if you can get 60+ fps for a game :P
 
If you don';t use shadowplay, try uninstalling geforce experience. That software used to make my computer crash all the time. Now I just install the driver and physX only. You can use FRAPS to monitor your framerate.

I do use Shadowplay all the time but I'll give it a try.
 
These are the settings I'll be keeping from now on (with my pc specs) I am getting a rather solid 60fps. There is some stuttering/framerate drops still but not as bad as before. I used the nvidia guide to check out what each setting looks like and I am quite happy with how the game looks. Really happy with the visuals/performance now. Time to get my teeth into the game now :)

VF2JoDu.jpg
 
970, i5-6600k here

How big a hit are reflections and shadows? I usually lower them to High/Medium in games
 
Close to the settings in the video above.

Now, that's a $1200+ PC (not talking about SLI) against a $400 console. 60 FPS aside, it IS remarkable.

I don't really agree because you can get that kind of performance from a much cheaper PC. A core i3-equipped budget PC with something like the RX470 should provide similar or better performance.
 
Got a nasty gameplay bug where I was in control of a camera and I couldn't get out. Just stuck with no option for anything. Pressing B to back out was no good either. Only option was to go to main menu. Silly bug to have left in. Maybe it was specific to this room at Stanford Uni where the door is locked and a case inside you can retrieve.

So a bit of fiddling and 4K, 100% pixel density, Shadows at Very High, SSR off, SMAA and everything else Ultra except that Detail Level Boost at 0% though I tried 10% and lost 2 or 3 fps. Also I play at 50hz/50fps.
 
Just experienced the windows explorer bug. This is insane. I bet it's the anti cheat tool causing all this weirdness.

And of course lost config of a tool I had open due to hard reset. Protip: Just close everything but the game client before you start this game smh
 
Top Bottom