CPU: Intel Core i5-2500 3.3 GHz / AMD FX-8320 3.5 GHz
RAM: 4 GB
HDD: 40 GB
GPU: NVIDIA GeForce GTX 560 / AMD Radeonâ„¢ HD 6870 1 GB
OS: Win 7 (64-bit), Win 8 (64-bit), Win 8.1 (64-bit)
CPU: Intel Core i5-4670K, 3.4 GHz / AMD FX-8350 4.0 GHz
RAM: 8 GB
HDD: 40 GB
GPU: NVIDIA GeForce GTX 780 / AMD Radeon R9 290 2GB
Test machine specifications
CPU: AMD FX 8320, 4.2 GHz
RAM: 8 GB
GPU: NVIDIA GeForce Gigabyte Windforce GTX 970 4GB
OS: Win 8.1 (64-bit)
Drivers: Nvidia Game Ready 347.25
I was very surprised when I initially turned the game up to the highest settings for testing, as there was a lot more aliasing than at lower options. The culprit is view distance, which is actually a LOD slider. You see, at higher settings, the game uses overly complex geometry considering how far away you are. For example, the distant trees and scaffolding under the highway created a lot of aliasing in motion and is unnecessary, so I actually prefer the lower settings for a smoother, simpler scene. Another thing to keep in mind is that this is biggest hit on framerates. At maximum view distance in the scene below, I got an atrocious 18FPS. At minimum I hovered at a stable 35. One thing to note is that zombies aren't affected by this, as you can see in the comparison, the distant zeds are identical. I definitely recommend turning this down regardless. Full resolution screenshots: Min, Ultra.
Texture quality only has two settings, medium and high and it has minimal impact on visual fidelity. If you're having issues running low on VRAM, turning down the textures isn't a horrific fate. Full resolution screenshots: Medium, High.
For anti-aliasing, the only option is a binary on/off. I'd wager with it's low performance impact that it's most likely FXAA. A little disappointing considering Dying Light is DirectX11 only and all DX11 games are capable of more advanced AA methods. Full resolution screenshots: Off, On.
Depth of field
The depth of field is an Nvidia only feature so I'm sure AMD users will be glad to hear that the setting actually makes no difference. It also doesn't affect performance. Full resolution screenshots: Off, On.
The shadow setting has a major impact on image quality but not framerate. I definitely recommend turning this up to atleast medium, as some scenes do look really great. Full resolution screenshots: Low, High.
Mouse smoothing by default is off and while the sensitivity slider isn't numbered, it has a lot of steps so it isn't difficult to find a sensitivity that you're comfortable with. Leaving the mouse menu sensitivity at default means that the in-game cursor behaves exactly like it would in windows, so I recommend leaving that where it is.
Dying Light is the first game from Techland that has an in-game functional FOV slider and its numbered too! This is a very nice surprise, as both Dead Islands and their latest Call of Jaurez required editing a text file with rather unreliable results. The FOV slider goes from 68 degrees to 104 so covers a nice range, however you can get higher (or lower) FOV by using this guide.
The key bindings in Dying Light are fairly good, with support for side mouse buttons. Every in-game action can be rebinded which is a good start.
The performance of Dying Light is a very odd case. All settings apart from view distance have very little effect on your framerate. What does matter is your processor's first core.
The picture above is why only view distance effects framerate, it's why GPU intensive settings don't make a difference and it's why framerate drastically varies from person to person. It's such a shame that a modern game running on a brand new iteration of Techland's Chrome Engine still can't get multithreading right, especially in an age where more cores seem to be every chip manufacturer's goal.
To illustrate my point, I conducted a few benchmarks. The first one is a comparison between max settings with view distance and it's highest and max settings with view distance at it's minimum. The difference is astounding, almost doubling my minimum FPS and gets the average FPS to a much better ~50.
To further highlight this single thread bottleneck, I'll show you the difference in performance between max settings with highest view distance and minimum settings with highest view distance.
As expected, the frame rate is almost exactly the same. The CPU intensive LOD and poor use of multiple cores combine to cause this serious issue.
In conclusion, the game's a fairly well rounded port excluding the multithreading and view distance issues. I can't wholeheartedly recommend Dying Light just yet as the performance issues really are disappointing, but if Techland are able to fix the CPU utilisation in the coming days or weeks, I think this would be a very good purchase.
Thank you for reading. If you enjoyed this report and you would like us create more articles, more often, please consider donating toPCGamingWiki's Patreon campaign: