Tomb Raider Underworld: Frame Rate Analysis

The video below was originally going to form part of the Eurogamer Xbox 360 vs PlayStation 3 Round 17 feature that went live yesterday, but something bothered me about it - the occasional spike to 32fps shouldn’t really happen in a v-locked game that runs at 30fps. You might expect a certain amount of jitter in any form of video capture and bearing in mind that 11,000+ frames are being analysed on each platform, a certain, negligable amount of dropped frames wouldn’t be unreasonable.

Yet the capture software reported none, and more than that, if there were dropped frames, you would expect some of the unique frames to be dropped at an equally consistent rate, resulting in a lower frame rate (so as many 28fps readings as 32fps results). And yet, Xbox 360’s results speak for themselves.

So, before publishing the video, I thought it best to verify that the results were totally sound. I tried the same test with a Blackmagic HDMI capture card (albeit at 16-bit precision - the best it has, versus TrueHD’s truly lossless 24-bit) and saw the same results, thus ruling out any issues with TrueHD drivers or hardware. After that, the real test: Project Gotham Racing 4, my benchmark for a v-locked 2VBL game. Frame analysis was consistent there: 12,644 frames captured in one test, exactly 6,322 duplicate frames - pure 30fps, and the testing method vindicated. So here’s the ‘missing’ video.

So there we go. Significantly prettier on PS3 - almost PC-like, even - but consistently smoother on Xbox 360. Most of the time the Sony platform keeps pace at 30fps, but when the frame rate drops, you really feel it in the controls. As I said in the face-off: one looks better, one plays better, but both are very cool games.