Jump to content


  • Content Count

  • Joined

  • Last visited

  • Days Won


Posts posted by LDK

  1. PC Reports are a series of quick first impressions regarding the technical aspects of a PC game. This report was written by PCGamingWiki contributor LDK. For an up to date account of Fallout 4 fixes and improvements, please visit its respective PCGamingWiki article.


    The Fallout 4 is the next title in a series of a very popular post-apocalyptic RPG's developed by Bethesda Game Studios. The game key was generously provided by Gamesplanet. We are going to look at the technical quality of the PC version of the game.

    System Requirements


    • CPU: Intel Core i5-2300 2.8 GHz or AMD Phenom II X4 945 3.0 GHz
    • RAM: 8 GB
    • HDD: 30 GB
    • GPU: Nvidia GeForce GTX 550 Ti or AMD Radeon HD 7870 2GB of VRAM
    • OS: Windows 7, 64bit


    • CPU: Intel Core i7-4790 3.6 GHz or AMD FX-9590 4.7 GHz
    • RAM: 8 GB
    • GPU: Nvidia GeForce GTX 780 or AMD Radeon R9 290X, 3GB (Nvidia), 4GB (AMD) of VRAM
    The system requirements for Fallout 4 are a little bit on the demanding side. A quad core CPU is a must and the GPUs are from the older mainstream range. Recommended specs are even more demanding with high-end CPUs and GPUs.


    All the tests were done on a system with a Core i7-2700k clocked to 4.6GHz, 32 GB RAM and AMD R9 390 with 8GB of VRAM, Catalyst 15.11 Beta graphics drivers and version of the game. Testing was done at 1920x1200, and because there is no built-in benchmarking tool, a test run consisted of one minute of playing the game in one of the larger cities and included a short fight with a pack of ghouls. This resulted in very consistent frame rate measurements so only two measurements were averaged for each effect.


    Graphics settings

    menu options


    Most of the graphics settings have to be set in the launcher. This is rather inconvenient because there are very few graphical options directly in the game. Other that that the launcher offers a fairly standard arsenal of graphical settings and a few presets. The game has also auto detection system that sets options during the first launch of the game. A nice addition are the Windowed and Windowed borderless toggles.


    Field of View and wide screen setups

    Field of view is locked in low 80's by default and there is no direct option to change it from the game. Players have to edit configuration files to set a comfortable field of view.


    Same applies for ultra-wide and surround resolutions as these are not directly supported. Again, configuration files needs to be edited and even that can hide some of the GUI elements, thus rendering the game unplayable.


    For easy configuration file editing and tweaking a special configuration utility can be used.



    Overall performance and image quality

    Fallout 4 offers four image quality presets - Low, Medium, High and Ultra. There is only a 3% difference in performance between Low and Medium. High preset has a more significant performance drop of 33% and Ultra costs almost 40% of Low's framerate.


    The framerate is limited to 60 FPS by default, unlocking it via editing .ini files will unfortunately cause problems, as the game logic and physics are tied to the framerate. I've experienced severe framerate drops in major cities where the game's framerate seemingly randomly falls from 100 FPS to sub 30 FPS.


    The game also loads very slowly on mechanical hard-drives. The loading times were around 30-40 seconds, after moving the game on SSD loading times dropped significantly to only few seconds.


    During my testing I've found only one effect (Shadow Distance) to have significant performance impact on the game. Even the dreaded over-tessellated god-rays effect led only to a 4% drop in FPS. The rest of the effects have negligible or non-existent performance cost so I'm not going to include all the graphs here. Feel free to browse through the measurements though.


    graph   presets


    Visual quality on the Low preset is nothing special, shadows are visibly low resolution and cut not far from camera, vegetation is shown only very close to the player and textures are not very sharp. Medium preset improves shadow resolution and cutoff distance of vegetation. High preset further improves vegetation cutoff distance and increases cutoff for shadows and the scene seems to be less flat. Ultra preset further increases render distance. Overall the game doesn't look particularly impressive visually and the renamed Gamebryo engine very clearly shows its age. Full resolution screenshots: Low, Medium, High, Ultra.



    Texture Quality

    Texture quality controls the texture resolution and can be set in three steps - Medium, High and Ultra. Visual difference between High and Ultra is very small and performance impact wasn't clearly measurable. The textures are not very sharp to begin with so there is definitely room for improvement. Full resolution screenshots: Medium, High, Ultra.



    Shadow Distance

    Shadow Distance setting controls cutoff distance of the shadows and is the most demanding effect in the game. High setting causes almost 30% framerate drop, Ultra costs another 2% on top of that. Full resolution screenshots: Medium, High, Ultra.



    graphs   shadow distance


    The game offers only two anti-aliasing methods and these are only post process filters FXAA and TXAA. Both have major visual impact and horribly blurs the whole scene, vegetation lose its definition, broken lines are kept broken, textures are slightly blurred as well. At least these effects has only 3% performance impact. Full resolution screenshots: No AA, FXAA, TXAA.


    AA comparisons


    The game doesn't offer very large controls options. Mouse sensitivity can be set and the keys can be remapped but there are major problems with controls in Fallout 4. First of all there are multiple actions on one key that doesn't make sense and will cause trouble. Melee attack shares button with a Throw grenade function for example. There is also a system of favorites weapons and items, but slots are locked to the number keys and cannot be remapped. Scroll wheel also cannot be remapped (to navigating favorites for example) and controls camera distance which I find fairly useless.


    There is a mouse acceleration turned on by default and can be turned off only by editing an .ini files.


    Pip-Boy and GUI navigation is nothing short of a disaster. I was expecting it to be bad as horrible controls are Bethesda's trademark but this is a whole new level of awful. Nothing is consistent after more that 20 hours of playing I still have to think really hard about what key is doing what in the menus. For example, exiting or going back a menu is sometimes Tab, sometimes it's E and sometimes it is Esc. Worst is the settlement building option which suddenly has you using the arrow keys and the game doesn't even hint that Shift WSAD can be also used. Alternative GUI mod cannot arrive soon enough.


    On top of that tutorial is almost non-existent and fairly important game mechanics are never explained like VATS or how to assign settlers to jobs.



    Fallout 4 offers fairly standard audio options with many volume sliders. The game supports surround sound setups up to 7.1 and audio sound fairly good with one exception though. Voices have very low volume outside of cutscenes and very often are inaudible when speaking character is more than few meters away.



    Fallout 4 suffers from heavy consolitidis. The underlying game is fun and the world is rich but visually the game looks unimpressive and old, almost like modded Fallout 3. Performance is not terrible overall but there are framerate drops suggesting optimization issues, almost none of the graphics settings has any meaningful performance impact on the game.


    Keyboard and mouse controls are awful and should be redesigned from the ground up. Many of core PC features like FoV, widescreen support or unlocked framerate are absent and has to be edited in via configuration files. Another six month of polishing would greatly benefited the game, but obviously Bethesda is counting on modders to fix the game for free and that is not a good sign.


    PC Reports are a series of quick first impressions regarding the technical aspects of a PC game. This report was written by PCGamingWiki contributor LDK. For an up to date account of Fallout 4 fixes and improvements, please visit its respective PCGamingWiki article.


    Thank you for reading. If you enjoyed our article and want to us create more articles, more often, please consider donating to PCGamingWiki's Patreon campaign:


    Click here to view the article

  2. PC Reports are a series of quick first impressions regarding the technical aspects of a PC game. This report was written by PCGamingWiki contributor LDK. For an up to date account of The Witcher 3 fixes and improvements, please visit its respective PCGamingWiki article.


    The Witcher 3 is a continuation of a popular RPG from a Polish developer studio CD Projekt Red. We are going to look at the technical aspects of the game, test the performance impact of the effects and look at overall quality of the PC version.

    System Requirements


    • CPU: Intel CPU Core i5-2500K 3.3GHz or AMD CPU Phenom II X4 940
    • RAM: 6 GB
    • HDD: 40 GB
    • GPU: Nvidia GPU GeForce GTX 660 or AMD GPU Radeon HD 7870
    • OS: 64-bit Windows 7 or Windows 8


    • CPU: Intel CPU Core i7 3770 3,4 GHz or AMD CPU AMD FX-8350 4 GHz
    • RAM: 8 GB
    • GPU: Nvidia GPU GeForce GTX 770 or AMD GPU Radeon R9 290
    System requirements for The Witcher 3 are on the more demanding side. A newer quad core CPU is a minimum required processor and the GPUs are also not of the budget kind. Recommended hardware is pretty much the current lower high-end. Let's hope the visuals justify these high demands.


    All the tests have been done on a system with a Core i7-2700k clocked to 4.5GHz, 32 GB RAM and HD 6870 with 1GB of VRAM. The Witcher 3 was on version 1.02, Catalyst version 15.4 Beta. Testing was done at 1280x800, as no benchmarking tool is available in the game, a test run was a one minute ride through one village to the countryside. Results were very consistent between each run (difference was rarely more than half a frame per second), so only two measurements were averaged for each effect.


    Graphics settings

    menus   graphics


    The Witcher 3 offers two graphics settings menus - Graphics and Post Processing. The main Graphics settings menu offers a fairly standard set of visual options and sliders. A frame rate limiter is present, but can be switched off in this menu. Three display modes are available, borderless fullscreen is the default option. Apart from the Nvidia HairWorks setting, every other option can be set to four different levels - Low, Medium, High and Ultra. HairWorks only offers the options Off, Geralt and On. Four presets are available for less demanding users.


    The post processing menu features toggles for a few special effects, Ambient Occlusion and Anti-aliasing. There are also three further presets for easy setup. Every setting is applied in realtime and no restart is required making tweaking fairly painless.


    Field of view and wide screen setups

    The Witcher 3 does not offer a field of view slider. The default field of view on my screen seemed OK, but I would definitely like to increase it by a few degrees. The only major issue appears when using the Witcher Senses mechanic because the field of view gets narrower, which I found to be very disorienting. Hopefully there will be a mod for this in the future, but for now you can use a Cheat Engine configuration file, as always with caution.

    of the tweaked field of view.


    The game appears to keep vertical field of view locked regardless of the aspect ratio of the screen, so gamers with ultra-wide or surround setups (both are supported) will get a correctly wider field of view. Only cutscenes appear to be pillar-boxed on setups with very wide aspect ratios.


    The in-game HUD also has issues on these setups. The game offers HUD scaling slider in the menu, but that usually isn't enough, so manual tweaks are required. To change this, navigate to the %USERPROFILE%\Documents\The Witcher 3\ folder and open a file called "user.settings" in text editor. Locate the line "UIHorizontalScale 1" and change its value. After saving the file, set its properties to Read Only.


    Overall performance and image quality

    The Witcher 3 offers four graphics quality presets and three presets for post processing filters. Post processing filters do not have a very significant impact on the performance, at only around 14% going from Low to High. The graphics presets are a completely different story though. Choosing the Medium preset caused an 11% performance drop, which while noticeable is not huge.


    graph   presets

    The high settings hurt performance a lot more - choosing the High preset caused the frame rate to drop about 64% from Low. The Ultra preset causes an even bigger performance drop - 84% from Low and 63% from High. Please note though, that testing was performed on a system with an AMD GPU, so the HairWorks effect had a more significant impact.


    graph   post processing presets

    Visual quality on the Low preset is average at best, textures are muddy, vegetation is sparse and trees switch from 3D models to low resolution sprites not very far from the camera. On the Ultra preset the game looks a lot better, though it's nothing really special. I'd even say that last year's open world title, Far Cry 4, looks nicer in some places and definitely runs significantly better faster than The Witcher 3. Furthermore, there is the downgrade issue where supposedly developers got rid of many special effects to get the game running on consoles and didn't attempt to implement them in the PC version. The game is not performing very well even on high end hardware so I wonder if they also left out some optimizations. Full resolution screenshots: Scene 1: Low, Medium, High, Ultra, Ultra High post process. Scene 2: Low, Medium, High, Ultra, Ultra High post process.





    The Witcher 3 offers some tweaks for mouse and controllers. There are two sliders for mouse sensitivity, one for camera movement and another for menus. There is also a Hardware Cursor setting that toggles mouse raw input. These are awesome features to have, other developers take note! There are also toggles for inverting axes and sliders for controller sensitivity.


    menus   controls general

    The key bindings menu is pure disappointment. A quarter of the actions cannot be rebound at all and thus block many keys, and that includes movement, which is locked to WSAD. This is completely unnecessary, and just prevents players from leaving the menu until all important actions are bound to any keys.


    Since version 1.04 this issue has been fixed and all actions can be assigned to a custom keys.


    menus   controls binds

    The in-game GUI and menus can be easily navigated by a mouse or a controller, a few additional bindings would be welcome however. Camera movement in the game feels awesome and precise when controlling with mouse thanks to the raw input feature. Character movement on the other hand felt cumbersome and very inaccurate in tight spots or on a horse. Combat is nice and precise, and I have a feeling it takes a few hints from the Dark Souls series, but I haven't played previous Witcher titles so I can't really tell, if it's better or not.



    The Witcher 3 audio options offers just a few volume sliders and nothing more. The game supports surround sound systems very well and the positional audio uses all channels nicely.


    menus   audio


    Performance analysis

    In this section we are going to look at how each effect impacts frame rate and how does each setting looks on screenshots. Be sure to check full resolution uncompressed PNGs linked in each section - these are linked in each section for better comparisons as the images in the sliders are compressed. Be sure to check our gallery for additional images and graphs.


    Shadow Quality

    Shadow Quality setting controls the resolution of shadows and at what distance shadows are shown. The difference in visual quality is very apparent in direct comparison as the shadows are sharper and more detailed. Full resolution screenshots: Low, Medium, High, Ultra.


    Performance impact of Medium and High settings are on the low side (up to 10%), but the Ultra preset is very taxing and causes almost a 70% frame rate drop.


    graph   shadows

    Water Quality

    This options should control amount of tessellation on water surfaces and waves. Personally, I haven't noticed a major visual difference. Full resolution screenshots: Low, Medium, High, Ultra.


    Performance impact of Water Quality is negligible on Medium (3%), moderate on High (15%) and quite high on Ultra setting (40%).


    graph   water

    Foliage Visibility Range

    This option controls how many trees are rendered in the scene and how many real 3D models of trees are used. The scene is much richer on higher settings as more trees on the horizon are shown and grass is rendered much farther from the camera. Full resolution screenshots: Low, Medium, High, Ultra.


    Performance impact is moderate on the Medium and High settings, at 11% and 15% respectively. The Ultra setting causes a 40% drop in performance.


    graph   foliage

    Nvidia HairWorks

    This setting replaces default hair with physics-based hair. As this is Nvidia technology, AMD cards perform with it terribly. There are options to turn the effect completely off, use it only on the main character, or use it everywhere.


    graph   hairworks

    Grass Density

    This setting controls how rich the grass is. Visual impact is not very major, as not many grass models are added. Full resolution screenshots: Low, Medium, High, Ultra.


    Performance impact is negligible, up to 6% on the Ultra setting level.


    graph   grass density

    Ambient Occlusion

    Ambient Occlusion adds subtle shadows around objects and in corners. The Witcher 3 supports only two methods - SSAO and HBAO . Visual impact is very noticeable, there are shadows everywhere and objects are more defined. Full resolution screenshots: None, SSAO, HBAO.


    Usually, this is a very demanding effect, but performance impact isn't major, at only 8% for SSAO and 10% for HBAO .


    graph   ambient occlusion

    Texture Quality

    This option controls texture resolution, on Low settings textures are very blurry and barely recognizable (especially in cut scenes), higher settings makes them sharper and crisper. I haven't noticed any difference when comparing High and Ultra settings in 1920x1200 resolution, so the game is probably ready for using higher 2k and 4k resolutions. Full resolution screenshots: Scene 1: Low, Medium, High, Ultra. Scene 2: Low, Medium, High, Ultra.




    Performance impact is highly dependent on available VRAM and in my case there wasn't major dip in performance. If you are experiencing stutter especially when rotating camera, consider lowering the texture quality.


    graph   textures

    Terrain Quality

    This setting should control amount of tessellation applied to terrain, but I haven't noticed any visual difference. Full resolution screenshots: Low, Medium, High, Ultra.


    Performance impact was also negligible, up to 4% on Ultra.


    graph   terrain

    Detail Level

    This setting is a mystery for me, I haven't observed any visual difference and there is almost no performance drop.


    graph   detail level

    Number of Background Characters

    This setting controls how many people are in the cities and villages. I suspect performance depends on the CPU rather than the GPU with this setting, on my system there was almost no performance drop.


    graph   background characters

    Anti-aliasing and other post processing effects

    The Witcher 3 supports only one anti-aliasing method. It is some sort of post process AA filter that blurs whole scene and kills details on textures. Maybe that is why there is also a Sharpening filter that repairs the scene a bit. There is also a Chromatic Aberration filter, that blurs everything but center of the image. Why would anyone use these blur filters is a mystery to me. Performance impact of each filter is very small, turning on every filter will result in a 14% performance drop.


    post process comparison


    graph   post process effects



    I've been looking forward to this game, unfortunately the final product feels rather unpolished. There is a lack of features like proper anti-aliasing methods or a field of view slider. Character movement controls are cumbersome, very restrictive keyboard bindings do not help either. In-game menus are easily navigable and the game supports raw mouse input, so at least there's that.


    Performance is not great and the game has problems keeping a decent frame rate even on high-end hardware. Stable 60 FPS@1920x1200 on Ultra is only attainable on a GTX Titan X, barely. That wouldn't be such an issue, if the game looked like it did in the first trailers. It doesn't look particularly bad, but there is very much to be desired.


    PC Reports are a series of quick first impressions regarding the technical aspects of a PC game. This report was written by PCGamingWiki contributor LDK. For an up to date account of The Witcher 3 fixes and improvements, please visit its respective PCGamingWiki article.


    Thank you for reading. If you enjoyed this report and want to see more articles from us, more often, please consider donating to PCGamingWiki's Patreon campaign:


    Click here to view the article

  3. PC Reports are a series of quick first impressions regarding the technical aspects of a PC game. This report was written by PCGamingWiki contributor LDK. For an up to date account of Killing Floor 2 fixes and improvements, please visit its respective PCGamingWiki article.


    Killing Floor 2 is a continuation of a very popular coop zombie shooter from Tripwire Interactive. The game is just entering Early Access and we are going to take a look at its performance and overall technical quality.

    System Requirements


    • CPU: Core 2 Duo E8200 2.66GHz or Phenom II X2 545
    • RAM: 3 GB RAM
    • HDD: 10 GB
    • GPU: GeForce GTS 250 or Radeon HD 4830
    • OS: Win7 64-bit, Win8/8.1 64-bit


    • CPU: Core 2 Quad Q9550 2.83GHz or Phenom II X4 955
    • RAM: 4 GB RAM
    • GPU: GeForce GTX 560 or Radeon HD 6950
    System requirements seems to be very reasonable, minimum required CPU is a just dual core and a rather old one. Similar story with required GPU, which is very old mainstream. Recommended specs are little bit more demanding, as a quad core and a newer GPU is needed, but again nothing anywhere close to the current high-end hardware.


    All the tests have been done on system with Core i7-2700k clocked to 4.5GHz, 32 GB RAM and HD 6870 with 1GB of VRAM, Killing Floor 2 version was 1003, Catalyst version 15.3 Beta. Testing was done in a 1650x1050 resolution, the game doesn't offer benchmark tool, so test run was me playing the first round in a solo mode on the Burning Paris map and killing all zombies which took about one minute. Each effect was tested three times and results were averaged to eliminate inconsistencies between each run.


    Graphics settings


    Graphics settings menu offers wide variety of effects and setting to tweak. Some of the effects require restarting the game, so set them before you enter an online match. These effects are: Environment Detail, FX, Realtime Reflections and Shadow Quality. Other effects can be adjusted on the fly with immediate effect, but I've restarted the game after each change for benchmarks just to be sure.


    Killing Floor 2 supports any resolution, some more exotic ones can be accessed by switching Aspect Ratio to Any. The game supports Borderless, Windowed and Fullscreen modes, and Variable Framerate can be turned on for high refresh rate displays. I've encountered slight problem when lowering resolutions, but restarting game fixed it.


    Field of View

    Field of View slider is hidden under Game Settings options menu. It doesn't say exact degree of field of view, there is percentage adjustment instead with 100% - 125% range. Killing Floor 2 apparently uses vertical implementation of field of view. This implementation works much better on a variety of aspect ratios, ultra wide monitors and AMD Eyefinity or Nvidia Surround setups. Bellow you can see two comparisons, upper one is the game's field of view on ordinary 16:10 aspect ratio screen, lower one is field of view behavior on ultra wide screen. Even when slider is set to 125% on both cases, ultra wide screen properly offers more visibility. Surprisingly there was no measurable impact on framerate. Full resolution screenshots: 16:10@100%, 16:10@125%, 21:9@100%, 21:9@125%.



    Overall performance and image quality

    Killing Floor 2 offers four image quality presets - Low, Medium, High and Ultra. On the Low preset the game performed very well and the framerate rarely fell under 100 FPS. Switching to the Medium setting framerate dropped 30%, same drop was when the High setting was used. The Ultra preset offers the best image quality, but with severe framerate impact of 40% from High and more than 70% from Low setting. The game's performance scales very well with each preset.


    graph   graphics quality

    Performance also scales fairly well although not linearly. Doubling the resolution from 1280x800 to 2560x1600 results in 62% framerate drop.


    graph   resolution performance

    Image quality even on the Low settings is very good. Lighting is dynamic, some textures are blurry and there are only basic reflections. On the other hand, the Ultra setting have textures very sharp even on 1600p, environment has more objects and realtime reflections are also present and very apparent. Below are comparisons from three currently available maps. Full resolution screenshots: Paris Low, Paris Medium, Paris High, Paris Ultra; Outpost Low, Outpost Medium, Outpost High, Outpost Ultra; Lab Low, Lab Medium, Lab High, Lab Ultra.





    Controls settings menu in the Killing Floor 2 offers basic mouse sensitivity settings, there is no mouse acceleration or smoothing toggle, but I haven't felt any additional mouse processing and aiming seems to be very precise. Keyboard binding menu offers basic key binds, only one key per action can be set. There was an issue with Caps Lock key, the key was binded to the Voice Chat by default and cannot be binded to anything else unless Voice Chat action is binded to other key first.

    Controllers are also supported, but buttons cannot be reassigned.




    Killing Floor 2 Audio Options menu offers three volume sliders for Game, Music and Voice Chat volume. The game supports surround sound setups and positional audio worked flawlessly on my 7.1 system. The audio quality is exceptionally good, the everything sounds very authentic. Killing Floor 2 is probably taking distance of the sound source and applying some filters on it for added realism. I was very surprised how good the game sounds.



    Performance analysis

    In this section we are going to look at how each effect impacts framerate and how does each setting looks on screenshots. Be sure to check full resolution uncompressed PNGs linked in each section. There are full resolution screenshots linked in each section for better comparisons as the images in the sliders are compressed. Be sure to check our gallery for additional screenshots and graphs.


    Ambient Occlusion

    Killing Floor 2 offers two ambient occlusion methods - SSAO and HBAO . SSAO adds very subtle shadows to the corners and around some objects. HBAO is much more pronoun and these shadows are much more noticeable. Full resolution screenshots: AO Off, SSAO, HBAO.


    Performance impact is around 10% for SSAO and almost 30% for HBAO from Off setting. As usual this is rather performance expensive effect, but not very noticeable with SSAO setting.


    graph   ambient occlusion

    Texture Resolution

    This setting affects quality of textures and can be adjusted in four levels. On Low setting some textures are very blurry and lacks any detail (glove on the screenshot for example), but other textures are surprisingly sharp and detailed even on Low. Larger quality boost comes with the High setting and the textures are obviously the sharpest at Ultra setting. Full resolution screenshots: scene 1 Low, Medium, High, Ultra; scene 2 Low, Medium, High, Ultra.



    Performance impact depends highly on amount of VRAM available, average framerate drop is only around 5% for each additional level. But heavy stuttering started to appear with High setting on my 1GB card.


    graph   texture resolution

    Shadow Quality

    Killing Floor 2 offers four levels of Shadow Quality settings. Each level increases resolution and view distance of the shadows. At Low setting some of the shadows are barely visible and the rest is very blurry. With each additional settings level, shadows are much nicer, sharper and farther visible. Full resolution screenshots: Low, Medium, High, Ultra.



    As usual, shadows have substantial impact on the performance. Medium setting causes around 7% framerate drop, High setting costs 16% and the most expensive Ultra setting costs 20% of performance.


    graph   shadow quality

    Environment Detail

    This settings affects how many objects are in the game world. These objects have no impact on the gameplay, usually they are there just for decoration (chairs, garbage..). Full resolution screenshots: Low, Medium, High, Ultra.



    Performance impact wasn't measurable, This is probably due to the fast CPU in the test rig. I'd imagine this setting will have much more pronoun impact on dual cores and CPU with weak single core performance.


    graph   environment detail

    Character Detail

    Character Detail setting affects model's level of detail. Polygon count is higher with each setting level, but it is not very noticeable directly in the game. Full resolution screenshots: Low, High, Ultra.



    Performance impact of Character Detail was only 5% for High setting and 8% for Ultra setting.


    graph   character detail

    Texture Filtering

    Texture filtering offers few filtering methods, unfortunately I haven't noticed any difference. There should be massive reduction of blurring on a distant textures, but the textures looks exactly the same. Performance impact is negligible for all but Anisotropic 16x setting where it is 13%. Full resolution screenshots: Bilinear, Trilinear, Anisotropic 4x, Anisotropic 16x.


    graph   texture filtering

    Depth of Field

    Depth of field setting offers only Off and On options. If turned On, DoF just blurs the gun when aiming down the sights. I'd like to see more of the blurring dependant on distance from player, but otherwise I much prefer this implementation. Some games have DoF forced on everything that is little bit farther from the camera just to hide horrible low resolution models or billboards however Killing Floor 2 is using DoF correctly. Performance impact is around 7%. Full resolution screenshots: Off, SAT DOF.



    Killing Floor 2 offers only FXAA as its anti-aliasing method. This is only post processing filter and not a proper anti aliasing method. Fortunately KF2 implementation is done fairly well and jaggies are moderately blurred without impacting sharpens of the textures. Performance impact is around 9%. Full resolution screenshots: Off, FXAA.


    Aa comparison

    Realtime Reflections and rest of the effects

    Realtime reflections adds reflections on some surfaces. This effects is very performance expensive, but it is well worth it as the game's world becomes much more realistic. This is the most demanding effect in the game, turning reflections On results in more than 40% framerate drop. Full resolution screenshots: scene 1 Off, On; scene 2 Off, On.


    Rest of the effects have negligible performance impact and I haven't noticed any visual difference.


    graph   misc effects


    PC version of Killing Floor 2 is very well done. The option menus are offering many setting to play with, controls are spot on and the sounds are awesome. Optimization is also very good as the game scales nicely and performs well. For a game, that is just entering Early Access, this is very polished product. Developers are actively pushing updates (I've downloaded two large updates just last week) and if first Killing Floor taught us anything, devs are not going to abandon Killing Floor 2 anytime soon.


    PC Reports are a series of quick first impressions regarding the technical aspects of a PC game. This report was written by PCGamingWiki contributor LDK. For an up to date account of Killing Floor 2 fixes and improvements, please visit its respective PCGamingWiki article.



    Thank you for reading. If you enjoyed our article and want to us create more articles, more often, please consider donating to PCGamingWiki's Patreon campaign:


    Click here to view the article

  4. PC Reports are a series of quick first impressions regarding the technical aspects of a PC game. This report was written by PCGamingWiki contributor LDK. For an up to date account of Pillars of Eternity fixes and improvements, please visit its respective PCGamingWiki article.


    Pillars of Eternity is an old-school RPG developed by Obsidian Entertainment, a very experienced developer of RPG games. The game is only released on PCs so naturally we are very interested in this title so dive in to find out, how the game performs, looks and controls.

    System Requirements


    • CPU: Intel Core i3-2100T @ 2.50 GHz / AMD Phenom II X3 B73
    • RAM: 4 GB
    • HDD: 14 GB
    • GPU: ATI Radeon HD 4850 or NVIDIA GeForce 9600 GT
    • OS: Microsoft Windows Vista 64-bit, OS X 10.6.3 Leopard 64-bit, Ubuntu 14.04 LTS 64-bit or newer


    • CPU: Intel Core i5-2400 @ 3.10 GHz / AMD Phenom II X6 1100T
    • RAM: 8 GB
    • GPU: Radeon HD 7700 or NVIDIA GeForce GTX 570
    The minimum system requirements are very low and Pillars of Eternity should run on vast majority of systems, including systems with integrated GPUs. Required graphics hardware is 7 years old and in mainstream range, but that is expected due the nature of the game. Download size is around 6GB, however the game takes almost 14GB on the harddrive. The only obstacle could be 64-bit operating system as the game is supposedly prone to crashing on 32-bit systems.


    Testing has been done on a Windows 7 system with an Intel Core i7-2700k clocked to 4.5GHz, 32 GB RAM and HD 6870 with 1GB of VRAM at the 2560x1600 resolution. The game key was purchased by Andytizer.


    Graphics settings



    Pillars of Eternity offers just a small number of usual options in the Graphics settings menu. There are resolution, vertical synchronization, gamma, fullscreen toggle and colorblind mode options. This is understandable due to the nature of the game as the Pillars of Eternity uses prerendered 2D backgrounds and only 3D models are characters, effects and moving objects.


    Multi-monitor setups are not directly supported by Pillars of Eternity, but there is a Cage Cursor option for these systems. This toggle restricts mouse movement to the game window. Unfortunately it is also a little bit broken and few times this feature didn't worked at all or restriction was completely off a real game window position. Alt-tab to the desktop and back to the game fixes these issues though.


    Resolution works little bit differently we are used to. The higher the resolution is, the more of the map is visible so on very high resolution (4k for example) screens some locations can be seen whole without need of scrolling. GUI in Pillars of Eternity scales on resolution nicely, for example inventory screen takes roughly 80% of screen regardless of resolution. This is a nice solution for very high resolution or DPI screens as some older games with high resolution patches has very small and almost unusable GUIs. There is also a Font Scale slider that affects some of the text windows and tooltips. This slider can be adjusted from 70% to 130%. Full resolution screenshot of resolution comparison here.



    Performance analysis

    Unfortunately there will be no benchmarks or lovely graphs for Pillars of Eternity due to the 2D/3D nature of the game and a fact that there are no graphical settings to test. The game on my main system runs between 50 to 80 FPS at 2560x1600 resolution, performance is highly based on how many characters are there on the screen as these are only 3D objects needed rendering. I've noticed slight performance drop the longer the game runs, but restarting game fixes it. Memory footprint is around 1,5GB RAM and fills completely my 1GB VRAM.


    I've also tested Pillars of Eternity on my very old system (Core 2 Duo E6750@2,66GHz, 4GB RAM, Radeon HD4850), and the game runs between 45 to 60 FPS at 720p resolution and is perfectly playable.


    Andytizer: I have tested this game on an old iMac (Late 2009, Core 2 Duo 3.06GHz, 4GB RAM, GeForce 9400M) on OS X and the game runs around 35 FPS at 720p and is very playable, especially considering this is below the listed minimum system requirements.



    Pillars of Eternity uses only MSAA as its anti-aliasing method and is always on by default. This can be a problem on a low-end or integrated graphics cards, fortunately MSAA can be disabled by a console command "msaa 0" without quotes. Console can be opened by a ~ key (key can be set in Controls menu).


    Update: The anti-aliasing option has been added to the Graphics option menu in the 1.0.3 patch. There are three options available in a form of s slider with Graphics Quality label, left most position is turning anti-aliasing completely off. Middle and right most options are different strengths of MSAA but I haven't seen any visual difference between those two settings.

    I haven't noticed any performance difference on my systems and I haven't noticed any image quality difference at first. Difference became clear after comparing screenshots in an image editor. Anti-aliasing is clearly there, but number of samples is very low or used method is not very good, changing number in msaa command doesn't seem to have any effect.


    There is a clear shimmering on a chainmail armors and horizontal line creep (for a lack of better term) on some helmets. This is just nitpicking, but I'd like to see implementation of SSAA only on 3D objects as this would greatly improve image quality and performance impact should not be that severe. Full resolution screenshots: MSAA default, no AA.


    Aa comp proper

    There is an another issue with anti-aliasing in Pillars of Eternity. If anti-aliasing is forced via graphics driver, the game suffers major image quality loss due to the massive background blur. This can be fixed very easily by disabling anti-aliasing in the drivers.




    There is no mouse setting or controller support in Pillars of Eternity, Controls settings menu offers detailed key binding though. Every major aspect of the game can be assigned to a primary and a secondary key, usage of Alt, Ctrl and Shift combinations is also supported.


    Additionally there are many other toggles and settings in the Game and Auto-Pause menus that tweak Pillars of Eternity behavior, difficulty and combat. I've also found some undocumented control features in the game, spells and special abilities can be assigned to otherwise unused keys by hovering mouse cursor over the spell/ability icon and pressing desired key. Every character can also be selected by specific number key and rest of numbers (as there are only six characters) can be assigned to character groups as in any real time strategy by using Ctrl number.




    Pillars of Eternity offers in Audio settings menu sliders for Master, Music, Effects and Voice volume. The game doesn't support surround sound and output is only stereo. Sound mixing seems to be OK most of the times, but few times music started playing rather loudly into the dialog. There is no subtitle options, but every narrated dialog is also captured in log part of the screen. Majority of dialogs are only text based though.



    Pillars of Eternity is superb resurrection of long forgotten 2D isometric RPG genre. Technically there are only minor issues and the game performs nicely even on very old systems. Controls are detailed as it should be on a PC platform. I'd like to see more detailed anti-aliasing methods, but otherwise Pillars of Eternity looks and runs very well.


    PC Reports are a series of quick first impressions regarding the technical aspects of a PC game. This report was written by PCGamingWiki contributor LDK. For an up to date account of Pillars of Eternity fixes and improvements, please visit its respective PCGamingWiki article.



    Thank you for reading. If you enjoyed our article and want to us create more articles, more often, please consider donating to PCGamingWiki's Patreon campaign:


    Click here to view the article

  5. Awesome job! I do have some very minor suggestions:



    • add screenshot guide (make syntax with file name, hit preview, click on red link, upload)
    • multi-page settings (keybinds1, keybinds2...) merge into one screenshot
    • "For screenshots of external launchers or options menus, the window border should be visible." - Use alt + printscreen to capture just active window

    The Fixbox

    • link to several good examples from real articles?
    • Providing instructions, major configuration file modifications section - add collapsible section? is it even possible within fixbox?

    The Infobox

    • leave Engine section blank when engine is unnamed or in-house build

    Video Settings

    • fixes related to the video settings should be under the table, so all the "See Field of View" links and fixes are close together. I'd even include here fixes for extra visual settings like motion blur, depth of field etc. Same for Input and Audio Settings sections.
    • what each icon represents? "true" (green tick) means that there is in-game option, "false" means there is no option even when a game supports it (ex AF by default to 16 but without option to lower it would be "false")

    Issues Fixed

    • I wouldn't remove the whole fix when it is officially patched, because there can be people stuck on old version for some reason (modding compatibility for example). Just state with {{ii}} tag that the issue has been fix by patch and its version number.
    • I personally like to describe issue little bit with {{ii}} tag before inserting fixbox like so: "The game seems to have an issue with Razor gaming hardware, use this fix if you are having problems." This way I can keep name of the fix (inside ===Name=== tags) short and also include necessary information about identifying the issue.
  6. Dimmdrive is a small application that can create a virtual hard drive in your RAM and copy a whole game into it. Because a RAM disk can be more than 10 times faster than an ordinary SSD, games should load much faster. In this article, we are going to measure performance gains when using Dimmdrive against ordinary hard drive solutions.

    Dimmdrive overview

    Dimmdrive allows you to create a virtual hard drive in your RAM. This drive will behave just like an ordinary disk in your system but with a much higher speed, even comparing to SSDs. A lot of applications of this kind already exist, but Dimmdrive offers something made specially for gamers: automatic Steam library integration and very easy game migration to a RAM disk. Because RAM is a kind of volatile memory (volatile memory requires power to keep stored data), Dimmdrive also implements file synchronization so if your system crashes, no data will be lost.


    dimmdrive main


    How much RAM do you need? Obviously, the more the better, but Dimmdrive permits selecting only specific files and directories to copy to the RAM drive. This way, you can use it even on systems with limited amounts of RAM.


    dimmdrive less Ram


    Additionally, Dimmdrive also can be used with any applications and folders. Importing other content is as simple as drag and dropping it into the main window. File synchronization and other features are also used on data added this way.


    So, how does it work? Dimmdrive creates virtual RAM disk and then it uses the Windows implementation of the NTFS symbolic link function to create a virtual directory on your physical disk that links to the real files and folders inside the RAM disk. In essence, it is a shortcut to the folders and files at the file system level. Applications that will access the data through such symbolic links are going to be unaware of it, seeing the files at their original location.


    Testing and using Dimmdrive

    Testing was done on a system with an Intel Core i7-2700k clocked to 4.5 GHz and 32 GB of RAM @ 1333 MHz. A selection of games available on Steam was used to test load times on different storage solutions. Measured times are based on loading save games, using a stopwatch. In order to prevent the Windows Prefetch system from skewing the results, every game was tested only once, then the system was fully restarted and games were transferred to a different drive and tested again.


    graph hdds


    The graph demonstrates the speed of each of the drives. Sequential read speed was measured using AS SSD Benchmark. The RAM drive is unsurprisingly the fastest, with an incredible 4.9 GB/s read speed, followed by a Kingston HyperX 3k 120GB SSD with 400 MB/s. Second to last is a standard mechanical HDD, a Samsung SpinPoint HD103UJ F1 1TB with 92 MB/s. A worst case scenario is also included, a very old Western Digital Caviar WD200BB 20GB connected using a PATA to USB 2.0 bridge, with very low read speed of just 22 MB/s.


    graph games


    As expected the USB HDD had the slowest load times, followed by 1 TB mechanical HDD. SSD offers an impressively large improvement, but surprisingly, the RAM disk fails to give any significant gains in all but one game. For Crysis, Civilization V, Bioshock: Infinite and Splinter Cell: Blacklist the loading times on the RAM disk are practically the same as from an SSD, and only Stalker: Call of Prypiat loaded 7 seconds faster from the RAM drive.


    I was most surprised that the newer games were not very susceptible to drive speeds, and even using from older mechanical HDD, the loading times weren't bad at all. I suspect this is mostly because of the multi-platform nature of these games, as the previous console generation had little RAM and slow HDD/DVD/BD storage, requiring games to be a lot better at data loading optimisation. The Bioshock: Infinite loading time from a USB drive is absolutely astonishing, as it is only 4 seconds slower than a RAM drive. I have no idea what kind of sorcery or black magic developers used for this game, but it is incredible.


    Older and PC exclusive titles are a different story. The loading times for Stalker: Call of Pripyat scales perfectly on different storage speeds, and the story is similar with Crysis and Civilization.


    The Steam integration of Dimmdrive application failed for me completely, as it found only one game, and to add to that, incorrectly. I don't have my Steam installation in the default C:\ location, but Steam has correct registry entries for its placement on my F:\ drive. However, Dimmdrive instead relies on the config.vdf file found within the config folder to find Steam games so deleting this file would probably solve the issue, but I wasn't keen on messing with my Steam installation.


    Importing a game's folder to the APP tab worked flawlessly though. With a simple drag and drop to the Dimmdrive window, the application imported the game and I was able to use it without issue, as it showed up in the GAMES tab. Using the app itself is pleasant, as the GUI is clean, and every function is easily accessible. For more advanced usage the developer offers thorough video guides on their YouTube channel.



    Dimmdrive offers much better read/write speeds than any SSD, but testing didn't show any major load time improvements compared to ordinary SSDs. Considering the rather high price-tag of 28€/$29/£23, I certainly would not recommend purchasing it at full price. If you are considering buying it, try a free RAM disk application first to see if you gain any improvement in your games. There is also very nice free GUI application for creating symbolic links, removing any need to mess with command line.


    I can see some usage for power users though, because of the file synchronization function. Cache folders can be loaded to RAM, thus making the use of large data warehouses faster - for example, image libraries with lots of thumbnails.


    I've been playing around with RAM disks on and off for almost two years now, and this just confirms my previous experiences. It is not worth the bother for games, but rather for some specific applications.


    A review copy of the Dimmdrive application was provided by the developer. It can be purchased from the developer's website or the Steam store for 28€, $29 or £23.



    Thank you for reading. If you enjoyed this report and you would like us create more articles, more often, please consider donating to PCGamingWiki's Patreon campaign:


    Click here to view the article

  7. PC Reports are a series of quick first impressions regarding the technical aspects of a PC game. This report was written by PCGamingWiki contributor LDK. For an up to date account of Elite: Dangerous fixes and improvements, please visit its respective PCGamingWiki article.


    Elite: Dangerous is the resurrection of one of the oldest PC games which is ushering back the niche but popular space simulator genre. The game is developed by Frontier Developments, who don't have many PC titles in their portfolio, so we are very curious how they managed to adapt to the complex and diverse PC environment. As usual we are going to look at technical aspects of the game, how it performs, hot it looks and how it controls. As this is PC exclusive title and has been in public beta for a quite a while we are expecting nothing less than a very high quality PC version.

    System Requirements


    • CPU: Quad Core CPU with at least 2Ghz
    • RAM: 2 GB
    • GPU: DX 10 hardware GPU with 1GB VRAM (Nvidia GTX 260 or ATI 4870HD)
    • HDD: 4 GB
    • Internet connection
    The minimum system requirements are surprisingly low for a PC exclusive title and developer didn't specify recommended specs at all. Minimum GPU's are six years old so even current low-end graphics cards and laptops should run this game without significant problems assuming CPU has at least four cores. Elite: Dangerous unfortunately requires constant internet connection even for single-player game. When servers are not available, only tutorials can be accessed. Single-player version of the game should be playable even on low quality connections according to the developers.


    All the tests were performed on a system with an Intel Core i7-2700k clocked to 4.5GHz, 32 GB RAM and HD 6870 with 1GB of VRAM at the 1920x1200 resolution. The game doesn't have built-in benchmarking tools, and achieving repeatable results wasn't easy. Test run consisted of three one minute fights in the Wolfpack Tactics training and averaging the results. I've found this test to be a good benchmark for gathering data because it takes place in asteroid field with many simulated objects with a large effect on performance. In other parts of the game, flying in open space effectively doubles the framerate as there are fewer objects being simulated.


    The game key has been provided by the developer.


    Graphics settings

    Graphics option menu is very well done and offers some unusual settings. The game supports windowed and windowed borderless mode and additionally there is a monitor selection for systems with multiple screens. Wide variety of resolutions are also supported including ultra-wide and 4K screens.


    menu graphics

    There are three settings for managing framerate. Vertical synchronization, monitor refresh rate and frame rate limiter. Vertical synchronization and frame rate limiter can be disabled to achieve maximum frame rate on those lovely 144Hz screens. The game also features few options for stereoscopic display (3D). On my system I could set only anaglyph 3D (I'd need glasses with green and red lenses) and side by side. I'd imagine it is prepared for VR support (Oculus Rift should be supported already).


    And at last there are several graphics quality options, field of view slider and preset selector.


    Overall performance and image quality

    I was pleasantly surprised how the Elite: Dangerous handles multiple cores. As you can see in the screenshot from Process Explorer, the game creates many threads. Two threads are doing majority of the work, then there is a thread by graphics driver (atidxx32.dll) and a bulk of additional threads. I suspect the game is heavily parallelized any many game processes are divided into separate threads. CPU utilization by the game fluctuated between 10 to 20% depending on what was on the screen. Dual-core processors with strong single thread performance would probably run it OK but need for quad-core is justified.


    CPU utilization

    The game features four quality presets - Low, Mid, High and Ultra. Performance scaling seems to be balanced fairly well. At Low the game performed nicely and on my ageing GPU the framerate never went below 60 and was averaging around 80 FPS. With Mid preset around 18% performance was sacrificed and the minimum framerate fell under 60 with average around 65 FPS. High preset resulted in almost 40% performance loss from Low preset, but the game was playable with average around 48 FPS and with drops to 38 FPS. Performance drop with Ultra preset was only 7% from High preset and 43% from Low preset and the game was playable with framerate abode 35 FPS.


    Graph preset performance

    Unfortunately there is little difference in image quality. Polygon edges are smoothed by anti-aliasing, asteroids have slight glow from their insides and ambient occlusion is visible on some objects. Nevertheless the game looks very good even on the lowest settings. Full resolution screenshots: Scene 1 Low, Mid, High, Ultra; Scene 2 Low, Mid, High, Ultra.



    Field of view

    Elite: Dangerous features field of view slider but in its default state it is fairly limited. Its maximum value is 60 degrees vertical (85 degrees horizontal on 16:10). Fortunately is can be changed in config files.


    To change field of view:

    • Navigate to %localappdata%\Frontier Developments\Elite Dangerous\Options\Graphics\
    • Open file Settings.xml and change FOV value on line 8.
    • Elite: Dangerous is using vertical field of view, so use field of view calculator to find out horizontal equivalent.
    Full resolution screenshots: FOV on min, FOV on max.



    I haven't seen that complex controls option menu for a very long time. The game supports vast variety of joysticks, ordinary controllers and last but not least mouse and keyboard. I was very surprised that my obscure 3D mouse was immediately recognized and I could assign every one of the six axes the device offers. And yet some bigger titles has problems with common mice...


    menu controls Alt

    The game offers few very basic presets, but majority of players are going to customize controls anyway. Every analog input has invert axis toggle and its own deadzone slider, if there is no analogue input, digital key presses can be assigned instead. Some of the ships functions can be set to key press or to keep key pressed to function.


    Mouse implementation in Elite: Dangerous shames almost every other game. Not only the game offers direct input, but there are sliders for two sensitivities, deadzone, relative mouse rate and mouse power curve. Both axes can be inverted independently for ship control and head movement. I have one small issue with mouse though. The game offers two control schemes for mouse movement: relative and absolute. I've found absolute mode much better for navigating with flight assist on, but pretty much unusable with flight assist. Opposite with relative mode. I would like to see another toggle, where I can assign different mouse modes to flight assist toggle.


    Keyboard support is also almost spotless and the game allows assignment of every key and if that is not enough, you can assign combination of keys up to four keys (Shift Q W for example). There is only one primary setting so only one key or one combo can be assigned to one action though.



    Audio options menu is similarly complex and features several volume sliders and plenty of toggles, there is no master volume slider though. I'm very happy to see Dynamic Range setting with options for standard speakers/headphones, home theatre with high dynamic range and night time for low dynamic range. The game also support surround sound systems and positional audio on my 7.1 system was spot on. Frankly the Elite sounds awesome and I think I'm enjoying sound of the game more then anything else.


    menu audio

    Performance analysis

    In this section we are going to look at how each effect impacts framerate and how does each setting looks on screenshots. Be sure to check full resolution uncompressed PNGs linked in each section. Or visit gallery for list of all screenshots for this article.



    The game offered me several anti-aliasing methods, post process filters FXAA (by Nvidia), MLAA (by AMD) and combination of post process and sub pixel sampling called SMAA. Performance drop is the worst with the FXAA, activating this methods results in 19% framerate drop. MLAA has considerable smaller footprint and performance cost is 4 and 8% respectively. SMAA performance drop is around 8%.


    Graph anti aliasing performance

    As usual FXAA softens edges fairly well but also washes textures making this method pretty much unusable. MLAA 4x has the same strength of edge blurring as FXAA and the textures are left almost intact. SMAA has similar image quality as MLAA 4x. Unfortunately the game really suffers from disconnected lines (shown at the top of comparison screenshots) as the game is filled with many one pixel wide lines and curves. This can be avoided by using any true anti-aliasing methods like MSAA of SSAA. Full resolution screenshots: No AA, FXAA, MLAA 2x, MLAA 4x, SMAA.


    Anti-Aliasing comparison


    Ambient occlusion

    The game only offers toggle for ambient occlusion effect. Performance drop is only around 3% and image quality impact is not very strong due to the nature of the game. Full resolution screenshots: Off, High.


    Graph ambient occlusion performance


    Material quality

    Material quality affects some materials and on screenshot additional sub surface lighting can be seen on the asteroid. Performance drop is around 6% from Low to Ultra. Full resolution screenshots: Low, Ultra.


    Graph Material Quality performance


    Shadow quality

    Shadow Quality affects resolution of shadows and has substantial performance impact of almost 30% from Off to High setting. Again by the nature of the game I haven't found strong difference apart from ship self shadowing ship on the cockpit. Full resolution screenshots: Low, High.


    Graph shadow quality performance


    Model draw distance

    This setting should control object cutoff distance from the camera, but I haven't found any difference. I could not even measure performance impact and framerate fluctuated around normal levels. Full resolution screenshots: Min, Max.


    Graph model draw distance performance


    Galaxy Map Quality

    Galaxy Map Quality should enhance details of galaxy map but I haven't found any difference. Performance drop is 30% for both Medium and High settings.


    Graph Galaxy Map Quality performance


    Rest of the effects

    Rest of the effects has small or negligible performance and visual impact. Visit gallery for unpublished graphs and screenshots.

    • Texture Quality - no measurable change in framerate, I haven't noticed any visual change. Maybe textures are already high enough for 1920x1200 and the game is prepared for 4k.
    • Environment Quality - no impact of Medium setting, 10% drop of High setting.
    • FX Quality - No performance impact from Off to Low, 3% drop from Off to Medium or High.
    • Reflection Quality - 5% framerate drop from Low to High setting.


    Frontier Developments have managed to release a very well done PC exclusive title. Elite: Dangerous' main strength is its unbelievably complex controls menu with support of many joysticks, controllers or whatever analogue input device works with Windows. Keyboard and mouse support is also very well done: the game features mouse direct input, complex mouse settings and advanced keys assignment.


    Performance of the game could not be better and the game should run on majority of current system without any problems. Image quality impact of some of the settings is questionable, but the game looks very nice even on lowest settings.


    With only few minor issues Elite: Dangerous is nice example of how PC title should look like from the technical stand point. The game is also still in very active development and developer listens and interact with the community on daily basis. Again, it's a prime example of what PC game development should be like, and not the "release, two patches and forget" tactics of many AAA studios.


    PC Reports are a series of quick first impressions regarding the technical aspects of a PC game. This report was written by PCGamingWiki contributor LDK. For an up to date account of Elite: Dangerous fixes and improvements, please visit its respective PCGamingWiki article.


    Thank you for reading. If you enjoyed our article and want to us create more articles, more often, please consider donating to PCGamingWiki's Patreon campaign:


    Click here to view the article

  8. PC Reports are a series of quick first impressions regarding the technical aspects of a PC game. This report was written by PCGamingWiki contributor LDK. For an up to date account of The Crew fixes and improvements, please visit its respective PCGamingWiki article.


    The Crew is online racing game from studios Ivory Tower and Ubisoft Reflections. It takes place in large open world and lets player to explore simplified version of USA. This article will analyze quality of the PC version, how it performs, how it looks and how responsive are the controls.


    The game was released on December 2, 2014 for Windows and is available through several digital distribution channels.

    System Requirements


    • CPU: AMD Athlon II X4 620, 2.6 GHz or Intel Core 2 Quad Q9300, 2.5 GHz
    • RAM: 4 GB
    • HDD: 15 GB
    • GPU: AMD Radeon HD 4870 or Nvidia GeForce GTX 260 with 512 MB of VRAM


    • CPU: AMD Phenom II X4 940, 3.0 GHz or Intel Core i5 750, 2.66 GHz
    • RAM: 8 GB
    • GPU: AMD Radeon HD 6870 or Nvidia GeForce GTX 580 with 1024 MB of VRAM
    The minimum system requirements are nothing to be afraid of, required CPU's are five years old and GPU's even more so the game should run on majority of today's hardware without any problems. Similar with recommended hardware specs, CPU's are also five years old, but newer architecture with higher clocks. Recommended GPU's are two generations newer then minimum and even those are in the mainstream range.


    All the tests have been done on system with Core i7-2700k clocked to 4.5GHz, 32 GB RAM and HD 6870 with 1GB of VRAM. The testing was done in 2560x1600 resolution to eliminate 60 FPS limiter. Benchmark consisted of driving one lap in test drive mode in New York car dealer during the day.


    Graphics settings

    menu video

    The game is running on the Babel engine developed in-house by Ivory Tower and offers few tweaks in the graphics options menu. Every common wide-screen resolution is supported and that includes 4k and ultra wide-screen resolutions. 4:3 aspect ratio resolutions are not supported at all and even if you force it by editing configuration files or by using borderless windowed mode, the picture will be stretched from wide-screen resolution.


    There is field of view slider, but for some strange reason this slider is not enabled for single screen setups. There is a cvar in the configuration file and editing it moves the slider in the menu, but it has no impact on the game. It should work for multi-monitor setups with AMD Eyefinity or Nvidia Surround technologies, but I was unable to confirm it on my system.


    Another issue is forced framerate limiter. It can be set to 30 or 60 frames per second, but higher framerate is not available. Other options in this menu are common effects that can be tweaked or turned off completely.


    Overall performance and image quality

    The Crew has four quality presets - Low, Medium, High and Ultra. On the Low setting not every option was turned off so be sure to check them if you have performance problems. Also Ultra preset is not setting anti-aliasing to maximum and keeps it to MSAA 4x instead of MSAA 8x.


    Graph - Presets performance


    At the Low preset the game performs very good on my aging GPU considering it has to run at 4Mpix resolution. And even on Low the game doesn't looks half bad. Medium preset offers better shadows and higher resolution of environment mapping and costs around 12% of performance.


    Performance drop of High preset is above 18% and further enhances shadows and mirroring on the car. Overall level of detail is better as cutoff for objects to disappear is further. Ultra preset means almost 32% performance drop from Low but I haven't noticed much difference from High. The testing was done with only FXAA instead of presets default anti-aliasing, because I was unable to get consistent results. More about that later.


    I've experienced stuttering on some skill races. The game runs on stable framerate, but it stutters a bit when I reach specific point in a few skill races. Other than that, the game runs very nicely. Full resolution screenshots: Scene 1 Low, Medium, High, Ultra; Scene 2 Low, Medium, High, Ultra.




    The Crew is racing game and as such it offers variety of control options. You can drive with keyboard, controller or a steering wheel and each can be separately configured. Even controller buttons can be remapped. Unfortunately I don't own steering wheel so I cannot comment on that, but controller and keyboard controls seemed pretty well done. Button prompts on the screen are changing on the fly and depends on the last input from controller or keyboard.


    menu controls


    Plugging the controller in the middle of the game can cause problems though. The game crashed once when I turned controller (wireless 360) on and commonly D-pad was not responding after the controller was connected after the game loads.



    Audio settings menu has several sliders to adjust volume of different aspects of the game's audio. The Crew is also supporting 7.1 surround sound, however surround sound is utilized only for few effects such as bypassing cars, roadwork etc. Radio, voices and car engine sounds only from front speakers. When driving, you can choose from a few radio stations and skip the songs on the fly.


    menu audio


    Performance analysis

    In this section we are going to look at how each effect impacts framerate and how does each setting looks on screenshots. Be sure to check full resolution uncompressed PNGs linked in each section. Or visit gallery for list of all screenshots for this article.



    I've got a few issues with anti-aliasing in The Crew. The main problem is very unstable performance regardless of resolution. MSAA 4x was running fairly OK with expected framerate drop when my car was stationary. When the car moved, the framerate plummeted to unplayable levels. Same with MSAA 8x but much more unstable. This is the reason, why there is no benchmark for each AA method. The FXAA had only small FPS drop as it is just a post process filter. I wasn't able to select TXAA on my AMD card.


    Sml gallery 13 56 4129


    As expected the game is heavily aliased and running lines are very much visible on the car without any AA applied (middle section of the comparison above). FXAA offers much better image quality and the textures are almost not affected, however there is slight moire around shard edges on the textures. MSAA preserves texture quality, but the edge blurring is not that strong and I've noticed light glow on some edges (same issue I've encountered in Far Cry 4). Full resolution screenshots: Off, FXAA, MSAA 4x, MSAA 8x.


    Ambient Occlusion

    The Crew features only two ambient occlusion methods: SSAO, SSAO plus or you can turn AO completely off. SSAO is much stronger than SSAO plus, but not as precise. Performance drop is only 8% for SSAO, but almost 30% for SSAO plus. Full resolution screenshots: Off, SSAO, SSAO plus.


    Graph - Ambient Occlusion performance


    Dynamic shadows can be turned off or set to one of four level. Each level is increasing shadows resolution and Contact hardening Soft Shadows should blur some of the hard edges, but I haven't noticed any major difference from High setting. Performance drop noticeable from Off to Low (around 10%), but negligible for additional levels (around 2%). Soft shadows option is exception with performance drop of 21% from no shadows. Full resolution screenshots: Off, Low, Medium, High, Contact Hardening Soft Shadows.


    Graph - Shadows performance


    This setting is managing polygon resolution of the game's assets and object's cutoff distance from camera. Performance impact is not very substantial and from Low to Ultra the FPS drop is only around 6%, image quality is much better though. Full resolution screenshots: Low, Medium, High, Ultra.


    Graph - Geometry performance


    Textures is the only setting, that requires resetting the game (that is why the car in the screenshots has different color). Impact on the game is rather questionable. From the screenshot the roadway on the right side of car has little bit more details on Low setting and I haven't found any difference on other textures. But that could be affected by weather or the bump map could be generated every time the game is loaded. Nevertheless other textures stayed pretty much the same regardless of texture setting and average framerate also stayed at the same level. Full resolution screenshots: Low, Medium, High, Ultra.


    Graph - Textures performance

    Environment Mapping

    This setting affects resolution of real time reflections on player's car. The game looks much better when the effects is set to Ultra as the reflections are very sharp and almost at the same resolution as the rest of the world. Performance impact is around 10% for Ultra setting, 8% for High and 2% for Low. Full resolution screenshots: Low, Medium, High, Ultra.


    Graph - Environment Mapping performance

    Depth of Field

    I haven't noticed this effect during gameplay but it is heavily used during cutscenes and in the garages. Performance impact is around 5% for Low and 10% for Ultra. Full resolution screenshots:Off, Low, Medium, High.


    Graph - Depth of Field performance



    Grass setting controls amount of grass billboards in the world. Performance impact of Medium setting is almost none, High setting will cause 5% framerate drop. Full resolution screenshots: Low, Medium, High.


    Graph - Grass performance


    PC version of The Crew is running very well and offers some graphics tweaking. Controls can be heavily adjusted and the game supports controllers and steering wheels apart from standard keyboard and mouse combo (which is more than capable).


    There are some issues though. Anti-aliasing was bugged on my system, field of view slider is not enabled for single screen setups and the game stutters occasionally (not as badly as in Far Cry 4, but it is there). Then there is the whole always online thing. The game will not start at all when the servers are offline. This is not good, because every race can be played in single-player and what happens, when Ubisoft decides to pull the plug?


    Overall The Crew is an average port without game-breaking issues. Which sadly is success after Ubisoft's disastrous this year season.


    The game key has been provided by the publisher and contained the game, several cars and the season pass.


    Click here to view the article

  9. PC Reports are a series of quick first impressions regarding the technical aspects of a PC game. This report was written by PCGamingWiki contributor LDK. For an up to date account of Far Cry 4 fixes and improvements, please visit its respective PCGamingWiki article.


    After not very impressive Watch_Dogs port and almost unplayable Assassin's Creed Unity, Ubisoft is trying its luck for the third time this year with another open world game called Far Cry 4. Is it better then its predecessors? To find out, we are going to look very closely on the performance of each effect, controls, audio and overall quality of the port.


    Far Cry 4 was released on November 18, 2014 for Windows and is available through several digital distribution channels.

    System requirements


    • CPU: Intel Core i5-750 2.5 GHz or AMD Phenom II X4 955 3.2 GHz
    • RAM: 4 GB
    • HDD: 30GB (31GB installed, 26GB without localization files)
    • GPU: Nvidia GeForce GTX 460 or ATI Radeon HD 5850; 1 GB of VRAM, DirectX 11 compatible


    • CPU: Intel Core i5-2400S 2.5 GHz or AMD FX-8350 4.0 GHz
    • RAM: 8GB
    • GPU: Nvidia GeForce GTX 680 or AMD Radeon R9 290X; 2 GB of VRAM
    The minimum system requirements are rather modest with five years old hardware, quad-core CPUs and mainstream GPUs from that time is needed. Recommended hardware is different story though and developer suggests using current generation quad-core CPUs and high end GPUs with at least 2 gigs of VRAM.


    All the tests have been done on system with Core i7-2700k clocked to 4.5GHz, 32 GB RAM and HD 6870 with 1GB of VRAM. The testing was done in 720p resolution to eliminate stuttering caused by the lack of video memory and streaming problems. Benchmark consisted of one minute running at the top of the bell towers in the game during sunny weather, again to eliminate stuttering during streaming additional data from HDD test run stayed only at one location.


    Graphics settings and overall performance

    options graphics

    Graphics and video options in Far Cry 4 are rich and allows to set many different levels for each effect. Display modes are also nicely done and the game supports many resolutions, common aspect ratios and even borderless windowed mode. Unfortunately the game is forcing 16:9 aspect ratio by using letterboxing (black bars) on different aspect ratios. Far Cry 3 didn't have letterboxing and it is shame to use it in this title.


    The game unfortunately doesn't support multi-monitor setups very well. There are problems like field of view slider not working, stretching 16:9 aspect ratio, letterboxing and other similar issues. More information about it can be found at Widescreen Gaming Forum.


    Additional setting contains field of view slider and few modes of vertical synchronization with option to turn on 30 FPS frame limiter to smooth out performance.


    Overall performance and image quality

    The game offers five quality presets - Low, Medium, High, Very High and Ultra. Ultra preset does not set maximum anti-aliasing but leaves it at post process SMAA instead of MSAA 8x. Although MSAA 8x has massive performance impact so I can understand this reasoning.


    graph preset performance


    At low details the game runs very well even on older GPU. Medium preset cost about 10% of performance but it looks much better with added ambient occlusion effect. High preset is rather costly and prepare for losing about 30% of framerate. In this preset ambient occlusion is much stronger, textures are sharper and mountains are much more detailed. Anti-aliasing is also turned on. Very High preset is not very different from High preset. Shadows are in little bit higher resolution and billboards of distant trees are more detailed. Performance drop is about 5% from High preset. Ultra preset is turning on several advanced volumetric and post process effect that has very strong performance impact of 20% from Very High preset and almost 50% from Low preset. Performance scaling in different resolution is rather linear.


    graph resolution performance


    Here we can see drastic difference between Low and Ultra presets. At Low details the image is very flat without any ambient occlusion effect, textures are blurry due to lack of anisotropic filtering. At Ultra the game comes to life and show beautiful scenery with detailed environment.


    The game unfortunately suffers from heavy stuttering during faster movement in game's environment. Developer suggest to install it on SSD but I've went further and created virtual hard drive in system memory and copied all the game's files directly into RAM. But even that didn't leave a dent and game kept stuttering.


    After few tries I've managed to completely eliminate stuttering after all. The key was lowering VRAM heavy effects and resolution. Anti-aliasing, ambient occlusion turned off, textures kept at medium helped a lot so I suspect the game is not very economic with its graphics memory utilization. I suspect this is due to bad optimization from console code as the current generation of console have unified pool of system and graphics memory that is much larger (4-5GB) then graphics cards today normally have.


    After getting rid of stuttering the game performed flawlessly on my system with framerate around 60FPS with few effects turned to moderate settings.


    Full resolution screenshots: Scene 1 Low, Medium, High, Very High, Ultra; Scene 2 Low, Medium, High, Very High, Ultra.


    Field of view

    We are very happy to see Far Cry 4 featuring field of view slider. Unfortunately the implementations is not consistent and on few instances it can completely break the game. First of all the slider is not working at all in windowed borderless mode and on multi-monitor systems. There are reports, that field of view slider completely disappeared with 1.4.0 patch for some users, but personally I have not encountered that.



    Then there are annoying issues with this slider. The slider is working nicely when character is running in the game, but any other activity is on default and very low field of view. Activities like riding in a vehicles, using grappling hook and in a cutscenes.


    And then there are game breaking bugs: Conquering bell tower starts cutscene where camera fly though few locations but after that field of view reverts to ridiculously low value about 50 degrees. Messing up with slider does not help, only remedy is getting back to main menu and reload save.


    mortar Fov Bug


    Completely opposite story is flying with wing-suit where field of view jumps to ridiculously high value above 150 degrees. With this jump game obviously stutters for a moment as it has to render much more of the scene often resulting in players death. Same thing with mortar aiming where field of view is much, much higher.


    graph field Of view performance


    There is also no indication of what field of view values slider represents but its maximum value seems to me to be just about right for a single large screen. Performance impact is noticeable with about 20% of framerate loss from lowest setting. Full resolution screenshots: FOV on min, FOV on max.



    The game offers nice controls customization even for gamepads although no full button bindings. Keyboard keys are fully customizable though with the exception of Tabulator key which can be fixed. There are few multiple actions on one key scenarios, but I haven't found them problematic. Very nice addition is plenty of toggles for GUI elements.


    options controls gamepad

    options controls


    What is problematic though is mouse acceleration. There is a switch and a slider that can turn off mouse acceleration, then there are additional entries in the configuration file but none of these can turn mouse acceleration completely off, only reduce its strength. But at least there is no negative acceleration as it was in Far Cry 3. Additionally the mouse movement felt jumpy and not smooth at all although I got used to if fairly quickly.



    The audio options menu is very disappointing. There is only one slider for master volume and toggle for turning off music. Surround sound is supported but only up to 5.1, additional channels are ignored. Positional audio is nicely done though and the game otherwise sounds very good, I'm particularly impressed with weapon's sounds as each gun has very distinct and rich sound with slight echo.


    options general And audio


    Performance analysis

    In this section we are going to look at how each effect impacts framerate and how does each setting looks on screenshots. Be sure to check full resolution uncompressed PNGs linked in each section. Also visit gallery for additional benchmark and screenshots.



    This setting is little bit complicated as it contains several different effects. I haven't found increase in resolution with higher levels of this options, however there is increase of anisotropic filtering that results in much sharper textures. Another effect bundled in "textures" is parallax mapping that adds very strong 3D relief to the textures. I don't know, why these effects cannot be separated as both have very different performance and memory impact.



    Performance impact is around 8% up to High level, Very High and Ultra cost 12% and 15%. This setting is very dependent on graphics memory and even if your framerate is high and stable in static scenes, not enough VRAM will cause the game to stutter when camera is turned quickly or character moves fast. Full resolution screenshots: Scene 1 Low, Medium, High, Very High, Ultra; Scene 2 Low, Medium, High, Very High, Ultra; Scene 3 Low, Medium, High, Very High, Ultra.


    graph textures performance



    This option sets up shadow resolution and distance of shadows level of detail. There is special level of Soft Shadows that blurs edges of shadows to appear more natural.


    graph shadows performance


    Performance impact is surprisingly not very high and Soft Shadows are causing only 10% FPS drop. But as with Texture option, this is very memory demanding effect that can cause stuttering. Full resolution screenshots: Low, Medium, High, Very High, Ultra, Soft Shadows.



    This sets terrain details. Low and Medium options are almost identical, very strong difference is on High level and Ultra offers very little improvement. Performance impact is very little, up to 3%. Full resolution screenshots Low, Medium, High, Ultra.


    graph terrain performance



    The game offers few anti-aliasing methods, three levels od MSAA (up to 8x), SMAA and Nvidia specific TXAA which I was unable to test.


    Anti-Aliasing comparison


    MSAA is very performance heavy and surprisingly does not offer very good image quality. MSAA is for some reason creating slight bright outline around some edges that can be noticeable. SMAA on the other hand soften edges very nicely and does not blur textures at all. Additionally SMAA also blurs edges on foliage as these are in a form of transparent texture that is ignored by MSAA. Performance impact is not very strong with SMAA, as it is form of post process filter. Full resolution screenshots: No AA, MSAA2, MSAA4, MSAA8, SMAA.


    graph anti aliasing performance


    Ambient Occlusion

    Far Cry 4 also offers three methods of ambient occlusion - SSAO, HDAO and Ubisoft brand new SSBC. Ambient occlusion adds subtle shadows around objects and in corners that creates much richer scene. Performance impact is moderate from 5% of SSAO and SSBC to 11% of HDAO . Full resolution screenshots AO Off, SSAO, SSBC, HBAO.


    graph ambient occlusion performance



    This option creates light shafts in a fog and dust. Additionally it tweaks contrast of sky and clouds resulting in very nice dramatic sky. Performance is between 13 to 14%. Personally I prefer option Volumetric Fog instead of Enhanced.


    graph godrays performance


    Rest of the effects

    Rest of the effects only quickly as they didn't seem to improve image quality dramatically. Post FX and Water quality performance drop is around 5%, but without noticeable visual impact, same with Vegetation and its 2% FPS drop. Geometry affects level of details of further objects and has 11% FPS drop from Low to High and 16% from Low to Ultra. Fur is switching advanced fur on animals with 16% FPS drop, but the effect is noticeable only when animal is very close. Trees Relief should turn on tessellation on trees, but I haven't found any changes. Performance impact is 11%.



    Far Cry 4 is a step in right direction, there are many toggles and sliders for tweaking, the game runs rather well and looks very nice. Unfortunately as with the previous Ubisoft games, Far Cry 4 suffers from premature release and severely lacks polish. There are unnecessary problems with field of view implementation, mouse acceleration, multi-monitor support (or lack of), limited audio option and stuttering. Sad thing is, these problems could have been easilly avoided if the game was in development little bit longer.


    Click here to view the article

  10. If you are experienced PC player, you have probably came across a broken or malfunctioning mouse. We are going to look at the unwanted double-click problem and provide detailed guide with many macro photos and video how to fix it regardless of mouse model. Without any soldering and with just basic tools you can bring back to life your older or aging, but once expensive, gaming mouse.

    The problem

    Double-clicking in single click, inability to create continuous line in paint application, difficulty to move camera in RTS, iron sights not sticking in FPS - these are all symptoms of a broken mouse button. Culprit in majority of cases is a component called miniature snap-action switch, or micro switch. This switch contains three leads and a metal with little spring. This metal is jumping between two contacts thanks to the spring and bridging connection between contacts.


    This spring can become tired so it isn't providing pressure necessary to maintain connection. With combination of dirt (dust) and oxidation on the contact areas, it is a perfect recipe for unstable connection and unwanted double clicks.

    Switches difference

    Fig. 1 - types of micro switches


    Tools and tips

    You will need:

    • screw driver (type depends on your mouse model, Phillips and flat head should be enough
    • tweezers
    • safety pin, nail scissors or something pointy
    • contacts cleaner or rubbing alcohol
    • any type of glue
    • camera (cellphone will suffice)
    • little bit of time


    • Make several photos of each step, it will help you tremendously in reassembly.
    • Don't use much force, there can be hidden screws or latches and you will break something if you are not careful.
    • Disconnect your mouse from computer and pull out batteries first.
    • Take your time, you do not have be finished in ten minutes.
    • Every photo in this guide is in high resolution, visit gallery for additional details.
    • And as usual, PCGamingWiki is not responsible for any damages caused by following this guide.

    1. Mouse disassemble

    This step is unfortunately rather vague, because of each mouse having vastly different design. Be extra careful and look for hidden latches and screws when pulling parts apart.


    First of all you have to find screws on your mouse. Usual places are under rubber feet, in battery compartment and under stickers. Consult Google with a name of your mouse and word "disassemble".


    Screws location

    Fig. 2 - possible screw locations


    After getting rid of all the screws, upper part of the mouse should come apart. Now it is a good time to clean any oil and dirt from many crevices in the upper cover with rubbing alcohol or soap water (in case upper cover does not contain any electronics). Let it dry thoroughly. Example of mouse insides are displayed on figure 3.


    Logitech MX700 naked

    Fig. 3 - Logitech MX700 without cover


    Now look for micro switches that needs fixing. You have to have clear access to them in order to properly fix them so additional disassemble could be required. Usually there are few screws, that holds everything to the plastic base, internal connectors, scroll wheel, that can be easily put away etc. In some cases there can be hard connection between boards that cannot be non destructively disassembled. Again take a lot of photos during your work for future reference.


    For example in figure 3, Logitech MX700 is very complicated mouse with three layers of PCB's, two of which has soldered hard connection (4 pins directly above scroll wheel and on the left of D-3 component) and soldered flat cable between two other (upper right corner, white cable). There are five fixable switches on this mouse and three unfixable.


    Vast majority of manufacturers are using Omron type switches for main buttons (see figure 1 and 4). There is a possibility of finding smaller, four pin surface mounted switches, that cannot be opened and repaired without soldering. These can be also found in Xbox 360 controller.


    Omron detail

    Fig. 4 - Omron type micro switch


    3. Opening the micro switch

    So you have located faulty switch and ensured enough access to it. Opening it is fairly straight forward. There are usually two latches, that has to be lifted separately (figure 5). Some manufacturers has these latches on the long sides of the switch (figure 1).


    Omron with arrows on tabs

    Fig. 5 - location of latches on the Omron micro switch


    Insert your open safety pin or point of scissors under the latch to unlock it and lift one side of the switch cover slightly (figure 6). Do not try to lift whole cover up just yet, you'll break the second latch.

    Omron with one tab open

    Fig. 6 - one latch unlocked and cover slightly lifted


    Do the same on the other side of the switch so you'll get situation displayed on figure 7. Rotate the mouse (or part with the switch) so the white pin on the top of the switch is pointing down.


    Omron with both tabs open

    Fig. 7 - both latches unlocked and both sides of cover slightly lifted



    Now you can pull cover away from the base. The white pin is loose in the cover and can be easily lost if not handled upside down. At the end, you'll end up with situation displayed in figure 8.


    Omron completely open

    Fig. 8 - uncovered switch and detail of the white pin


    4. Metal spring fun

    You can see insides of the switch. The part, that needs fixing, is the long metal with spring in the middle and It needs to come of the rest of the switch. This is another not very difficult task, lay the switch flat, as shown on figure 8, hold the part without the spring with two fingers and twist the metal to the side. It should go fairly easily and you'll end up with bare metal with spring (upper left side of figure 9).


    Detail of metal spring and fixing it

    Fig. 9 - Metal with spring and prying spring


    As the fix itself you'll need to pry the spring little bit. Lay the metal piece flat, hold it as flat possible with tweezers and pry the spring up just a little bit. There is very little strength required, just fraction of a millimeter should be enough. Clean the contacts on the metal with alcohol, do the same with pins on the base of the switch.


    Now for the tricky part: you have to put it back together. See figure 10 in full resolution for more details. There are two grooves on the two metal pins on the base of the switch (numbered 1 and 2 on the figure 10). These grooves are corresponding with cutouts on the metal, also numbered 1 and 2. The hammer needs to go between two most right pins.


    Grooves in contacts with description

    Fig. 10 - grooves on contacts and metal spring detail


    The easiest way I've found is to place the hammer first, then place the metal into the groove 1. Spring is resting on the middle pin, just above groove number 2. Situation on the top left, top right and bottom right part of figure 11.


    Setting ut spring

    Fig. 11 - reassembling metal with spring on the switch base


    Now for seating the spring into its groove. Take the tweezers or small screw driver and push the spring down into its place (rest of figure 11). This is especially tricky move, as you have to push with enough force to push it down, but not enough to over push. If you over do it, you'll have to start over. When you manage to seat the metal properly, try to click it few times so you know, it is working OK. Seated metal is displayed on figure 12.


    Finished switch

    Fig. 12 - properly seated metal spring on the switch base


    Whole process of opening and fixing the switch is captured on a video. Be sure to turn on HD for best details.



    4. Testing and reassembly

    Close the switch and make sure, you are putting the cover the right way. White pin should not be directly above the spring. Consult your photos for proper orientation. Unfortunately you have to fully reassemble your mouse to do proper testing. Screws are putting the right amount of pressure on the micro switches and without them you could get another faulty clicks.


    Sometimes sound of the click and force needed to achieve button press can change slightly depending of force used on prying the spring. Try to open it again if you are not satisfied with results.


    The best way to test success of the fix is in any paint application. Create a blank document, pick any color and create any random shape with the button, you've just fixed. You should get continuous line if the fix is successful. Glue back the rubber feet and enjoy your fixed mouse.

    View full article

  11. PC Reports are a series of quick first impressions regarding the technical aspects of a PC game. This report was written by PCGamingWiki contributor LDK. For an up to date account of Dark Souls 2 fixes and improvements, please visit its respective PCGamingWiki article.


    Dark Souls 2 is third installment of the "Souls" series but only second game on PC platform. First Souls game on the PC, Dark Souls, has been very rushed port with some controversial limitations, locked 720p resolution, locked framerate and awful mouse and keyboard support. Developer From Software promised proper PC version of Dark Souls, we are going to look, if they succeeded this time.


    Dark Souls 2 was released on April 25, 2014 for Windows and is available through several digital distribution channels.

    System requirements


    • CPU: Intel Pentium Core 2 Duo E8500 3.17 Ghz, AMD Phenom II X2 555 3.2 Ghz
    • RAM: 2 GB
    • HDD: 14 GB
    • GPU: Nvidia GeForce 9600 GT, ATI Radeon HD 5870


    • CPU: Intel CoreTM i3 2100 3.10 GHz, AMD A8 3870K 3.0 GHz
    • RAM: 4 GB
    • GPU: Nvidia GeForce GTX 750, ATI Radeon HD 6870
    The game is asking for a rather old components in its minimum requirements so it should run on a variety of older machines. Even recommended hardware specifications are not very menacing.


    Testing methodology

    Testing was done on a system with Core i7 clocked to 4.5 GHz, 32 GB RAM and AMD Radeon HD 6870 in 2560x1600 resolution. Unfortunately the game has framerate limiter set to 60 frames per second and this system is fast enough to run stable on 60 FPS regardless of settings thus making standard FPS measuring rather inconclusive.


    To circumvent this problem I was forced to abandon framerate measuring and measure GPU utilization which is in percents.


    In normal conditions (framerate not limited by the game, fast CPU) GPU is utilized close to 100% and framerate is fluctuating on complexity of the rendered scene. In this case however, maximum framerate is locked and my machine is able to keep framerate on maximum value of 60 FPS. To keep this framerate, GPU has to wait after it finishes a frame and this is lowering utilization.


    For example low end GPU is able to barely keep 60FPS so its utilization is 100%. Much faster GPU hasn't got many problems to achieve 60 FPS and it has to wait after each frame making its utilization only 60%.


    Keep that in mind when reading graphs in this article as lower value means better performance.


    Measuring was done by utility GPU-Z, version 0.7.7, with sensor refresh rate set to a half a second. The game lacks internal benchmark so test run in game was done in the main hub called Majula by running from obelisk to the camp fire, then to the pit and back to the obelisk. This run takes around a minute and produces roughly 120 data points.


    Recently released new Durante's utility GeDoSaTo wasn't used in this test as I was unable to produce consistent results with it and using it created game instability. And since game started using VAC as anti-cheat solution, there is a slight chance, that you can get banned. Developer of this tool considers that highly unlikely though.


    Video settings

    Basic video option consists of resolution, full-screen mode and auto detection. Windowed borderless mode is not supported out of the box and external utility, such as Borderless-Gaming, is needed. Again be careful of using any third party application as Dark Souls 2 is using VAC.


    List of available resolutions is taken from operating system, so vast variety of single monitor resolutions are supported (presumably even 4k). Unfortunately the game is again locked to 16:9 aspect ratio so multi-monitor support is achieved only by using Flawless Widescreen utility, which can get you in trouble with VAC.


    menu video


    Additional graphics settings can be found in the next menu. List grew quite a bit since its predecessor which contained only anti-aliasing and motion blur toggles. Dark Souls 2 features settings for variety of popular effects, some of them can be set in a different levels.


    Almost every effect can be adjusted directly in game apart from texture quality and high-quality character rendering. These two are applied after reloading your save.


    Unfortunately there is no field of view setting but default field of view is set rather reasonably and didn't cause me any discomfort even after few hours of playing.


    Mute on focus lost cannot be set and is disabled by default and the game simulation is not even paused when alt-tabbing, which is highly unusual behavior. Although alt-tabbing is working perfectly from fullscreen mode and is not very slow.



    Dark Souls 2 performed on my system very well as it has been outlined on methodology paragraph. CPU-wise the game is working on simulated dual-core system without any performance impact, however when limiting it to only single core the game was unplayable.


    Allocated RAM never exceeded 900MB, same with VRAM, even on max details. Loading a level is fast, although cold start has several annoying notifications requiring user's input.


    graph preset


    The game features three presets: Low, Medium and High. Again bear in mind these numbers are not representing frames per second, but rather GPU utilization. On Low preset the GPU is working only 50% of the time. High preset comes with 30% performance cost as the GPU has to work at 82%.



    Apart from massive background blur and ocean quality, image difference is not very apparent. Frankly I got a feeling, that some of featured graphical effects are there just to mask horrible objects in the distance and other low resolution/low polygon assets (see rocks and castle in this screenshot). PNG screenshot of preset in higher resolution: Low, Medium, High.


    Texture quality

    Texture quality can be set to Medium (presumably a console resolution) and High. There is very little (2%) performance impact when setting it on High. VRAM utilization also rose only by a 40 MB (620 MB on everything Low, 660 MB with TQ on High).

    graph texture quality



    Surprisingly there is almost no image quality difference between Normal and High. Textures are significantly better on character models when viewed from up close, but that only happens on cut-scenes or very rare instances in the game. Some textures are very nice and seems to be in high resolution but many other textures are not which created bad contrast when these two resolution assets are placed close to each other. Tiling one texture on a large area is also very common unfortunately. Screenshots in full resolution: Medium, High, placement.



    textures different resolutions


    Shadow quality

    Shadows are the most performance heavy in the game taking almost 20% of performance when turning on High. There are also Medium, Low and Off states.


    graph shadow quality



    Higher setting means higher resolution of shadows, but only characters and few objects are casting shadows and only few items are considered light sources. I haven't found specific pattern what qualifies objects for shadow casting and lights as a light sources. Generally characters are casting shadows, but I've encountered few torch stands and pillars that also casts shadows. One pillar is casting shadows correctly, but pillar just next to it does not. And this inconsistency is prevalent also with light sources. Sun is considered light source, same with player lit torches, but camp fires and many other torches are not. Screenshots in full resolution: Off, High, inconsistency.


    slider shadows inconsistency


    Water surface quality

    Water surface quality is very performance cheap but very nice looking effect. With quality set to low, ocean is just dark and bland. Increasing quality to Medium or High increases GPU utilization just by 2% and ocean surface looks tremendously better.


    graph water surface quality


    Unfortunately this effect is applied only to ocean and I haven't noticed any difference on any other water bodies. Screenshots in full resolution: Low, Medium, High.



    Model quality and HQ character rendering

    I haven't noticed any difference in model quality settings, performance impact was only 1%.


    High-quality character rendering is turning on cloth animations, otherwise there is no noticeable difference. Performance impact is below 1%.



    The game features only FXAA which is low quality post process blur filter. There is some jaggies reduction, but also fine details on textures are lost as whole picture is blurred. It is the cheapest anti-aliasing solution performance-wise, turning it on costs only 7% performance. Samples on the right are anti-aliased. Screenshots in full resolution: Off, On.


    Aa comparison


    Ambient occlusion

    Ambient occlusion setting has only two states - on and off. Implemented solution is not performance demanding (costing only 3%), but also doesn't look particularly good.



    Again I've noticed some inconsistency with this effect. On some scenes ambient occlusion is hardly noticeable, in others it is disturbingly strong. Screenshots in full resolution: scene 1 On, Off; scene 2: On, Off.



    Depth of field

    Dark Souls 2 is perfect example how depth of field effect should never be used. Depth of field can be nice and very powerful effect that adds realism to binoculars, camera viewfinders, gun sights and other devices, that uses lenses. Here we have blur filter on everything little bit too far or close without any context. Unfortunately Dark Souls 2 is not the only game, that is using DoF completely wrong. At least it does not affect performance very much.



    By that comparison it is apparent, that this effect is used mainly for masking low resolution textures and object on the skybox. For better comparison open these screenshots in full resolution: On, Off.


    graph misc effects



    Dark Souls 2 has been developed with controller in mind and Xbox controller works quite nicely right away. There are some problems with controller autodetection and multiple HID devices but that can be easily resolved by disabling every controller but one. There are few settings hidden in Game Options menu such as X, Y axis reversion, sensitivity and vibration strength. The only controller rebindable action is Jump, no other buttons can be changed.


    menu game


    Keyboard and mouse support is much better than its predecessor, but it is still light years from perfect and miles from just usable. First of all there are Xbox button prompts regardless of having the controller or not. That is making learning the game extremely difficult as tutorial offers only information like "Pushing RB LS makes triple jump". You can use our guide for some of the prompts.


    Every menu and option screen is screaming controller and is very hard to navigate with mouse as there is zero consistency. There are hidden sections that are activated by just hovering mouse cursor over them and getting from that section properly is also a nightmare. Let me give you an example with key mapping menus:


    menu nightmare 1


    Hitting escape key frees mouse from camera control and shows first menu and cursor. Hovering over icons shows drop-down sub menu. Clicking on Key Bindings brings appropriate menu. This menu consist of two sections: top selector of action groups and bottom section with actions and assigned keys. Hovering over top icons is switching between them nicely, but hovering over assigned keys locks this section and hovering over top icons is no longer working.


    menu nightmare 2


    So how to escape this menu a save changes? Logical solution would be hitting escape key with hopes, that would bring a dialog asking for saving any changes. Unfortunately escape key immediately shuts down every menu and all the changes are lost. Properly closing menu with saving is by action called "Cancel" or by clicking right mouse button and selecting back.


    This confusing behavior is prevalent in every menu screen. Technically you can use your mouse to navigate through them, it is just extremely frustrating.


    Another lovely new feature is double click binding. Some actions can be assigned to the mouse button double click. Interesting idea, horrible execution. Light left and right arm attacks are assigned to single click left and right mouse button respectively. Unfortunately after the game receives single click, it waits for potential second click, that could never come, thus delaying actions assigned to single clicks by a significant amount of time. Combat with these delays is very imprecise and again very frustrating.


    This issue has been briefly mentioned in the Official Bandai Namco Support Thread on Steam Community forums, but it has been since deleted from the first post. That suggests this will never be officially fixed, but few fixes are available in our Wiki article.


    Camera control with mouse is precise, and I haven't had any problems with it at all. Mouse acceleration is disabled by default and the game even ignores acceleration set by operating system so at least something is finally done right.



    Audio options are accessible only from the game and not from main menu. The game features voice chat, that is thankfully disabled by default. Volume of music, sound effects and voices can be adjusted separately.


    Surround sound is supported by default, however there is no fine tweaking. On my system the game took surround sound setting from the operating system and correctly used 7.1 configuration. Sound mixing is little bit iffy, some effects are sent only to front speakers and ignores position of camera or player. Overall surround sound is implemented very well.



    Performance in the game is very good but that comes with cost of not very striking visuals. Graphical settings offers only disputable visual changes. Resolutions are working correctly. Keyboard and mouse controls are barely usable and require extensive tweaking to work right so controller is again highly recommended. It seems that From Software haven't learned much from first Dark Souls and offers classic port from consoles.


    PC Reports are a series of quick first impressions regarding the technical aspects of a PC game. This report was written by PCGamingWiki contributor LDK. For an up to date account of Dark Souls 2 fixes and improvements, please visit its respective PCGamingWiki article.


    Click here to view the article

  • Create New...