Testing the 3D Performance of Your PC With Quake 4
By Gabriel Torres on June 26, 2007
One of the best ways to benchmark the performance of your PC is running a heavy 3D game, since it will pull the maximum performance your video card, CPU, memories and hard disk drive can deliver. Quake 4 fits nicely in this role for PCs found on the market today. In this short tutorial we will teach you how to use Quake 4 to measure the performance of a PC.
After installing Quake 4, you need to install the latest game patch, which will update Quake 4 to its latest version. Even though the latest patch can be found on Quake 4 official website (http://www.quake4game.com), we suggest you to go to this other website instead, since the official website delays ages to be loaded: http://www.gamespot.com/pc/action/quake4/index.html
Quake 4 is based on the same engine as Doom 3, so the way to activate its benchmarking mode is the same as we described in our tutorial Testing the 3D Performance of Your PC with Doom 3 and Far Cry: enter the console mode by pressing Control, Alt and ` (tilde) simultaneously and type:
The problem is that, contrary to Doom 3, Quake 4 doesn’t come with any prerecorded single player demo. It comes, however, with a prerecorded multiplayer demo after you have upgraded it to the latest patch, that can be run through the following command:
If you are using patches 1.2 or 1.3, the name of this demo is id_demo001. If you installed a patch under the 1.4 series, then the name of this demo is id_perftest. If you installed patch 1.4 and over, you need to enter Quake 4 in multiplayer mode, otherwise the demo won't run, returning a "file not found" error. This can be done by using the "Quake 4 Multiplayer" shortcut that the patch installed on your desktop, using the regular Quake 4 shortcut won't work. If you deleted by mistake this shortcut, you can create it again by adding a shortcut to Quake4.exe with the following parameters: +set fs_game q4mp.
After running timedemo or playnettimedemo, just call the console again to see the results, measured in FPS (frames per second). The higher, the better, of course. We strongly suggest you to run the same demo three or four times and calculate an arithmetic average of the obtained results (i.e., add them up and divide by the number of times the demo was run), since they vary a little bit.
Keep in mind that you cannot compare the results taken using different patch versions.
You can also record your own demo or even download a demo made by others from the Internet (downloaded demos must unpacked and saved under c:Program Filesid SoftwareQuake 4q4basedemos directory – you will need to create it). If you do so, you will need to make this demo available if you are posting a review, so your readers can use the same demo you used check the performance of their system compared to the published data. That’s why we always prefer to use the demos that already come with the game.
If you want to record your own demo, type Recorddemo on Quake 4 console, play the game, and then type Stoprecording to stop the demo recording. After this procedure the demo file will be available under c:Program Filesid SoftwareQuake 4q4basedemos. Play it with Timedemo .
Another option is to install HardwareOC Quake 4 Benchmark tool. This program comes already with some demos and can help you a lot configuring Quake 4 settings, like resolution and image quality, and also programming Quake 4 to run several different benchmarking session with just one mouse click (batch configuration).
Talking about resolution and image quality, we must remember you that you cannot compare results taken with different resolution and image quality settings. I.e. you cannot directly compare the fps rate taken on a machine running at 1600x1200 with image quality set as “high” with the fps rate taken on a machine running at 800x600 with image quality set as “low,” for example. That’s totally unfair, since the higher the resolution, usually the lower the performance. The same goes for image quality, the higher the settings, the lower the performance. In fact, that’s the difference between high-end VGAs to cards from other segments. Usually high-end cards can sustain a high performance even at high resolutions and/or high image quality settings.
To compare image quality, you can hit the F12 key to take a screenshot. Of course you need to take screenshots on the same location on Quake 4 map for the comparison to make sense. The screenshots will be saved under c:Program Filesid SoftwareQuake 4q4basescreenshots as a TGA (Targa) file. You will probably want to reduce the file size by editing them on Photoshop or similar software, but keep in mind that editing them will probably change their quality, thus compromising the final results. If you don’t need a high quality (and big) file output, you can type screenshotJpeg on Quake 4 console to capture screenshots using the Jpeg format, which creates smaller files.
A final and very important word about benchmarking: keep all the hardware parts and driver versions the same except the one being measured. For example, if you are benchmarking VGAs, all other components must remain the same: motherboard, CPU, memory, hard disk drive, etc. Of course since different CPUs may use different socket types, when benchmarking CPUs you may need to change the motherboard. However, keep the other components the same.
If you are going to publish the results, always disclosure all hardware parts, driver versions, resolution and image quality settings used, and also the patch version number.