PlayStation Move controller lag detected, analyzed
By Tim Stevens 
posted March 18th 2010 10:28AM
posted March 18th 2010 10:28AM
When Sony unveiled its PlayStation Move (nee Arc) controller at the GDC last week, it came along with some impressive promises: it would only cost developers 2MB of system memory (out of 256MB on offer) and it would respond to user inputs within a single frame of animation. Our own experiences with the thing felt a bit more laggy, and now Eurogamer is echoing those impressions and putting a bit of science behind them courtesy of a 60fps Kodak Zi6 camcorder. By filming the controllers and their on-screen representation, the site's tireless statisticians calculated an actual lag (including that of the display) of 113ms -- 
@spin cycle
Exactly. If this game is double-buffered, which it probably is, then you need to take 2 frames of lag before you can even start to calculate the controller input lag. Additionally, I'd really like to see the video calibration results from Guitar Hero or Rock Band on this display. There are TVs out there with 100ms+ video lag. Show me the same video on a CRT or a certified low-lag TV and we can start to draw conclusions.
This video only shows the entire accumulated lag of the entire system, including the display (which is easily the biggest variable). Not just the controller lag.
@spin cycle To isolate the lag from the Move, what they need to do is to take a game that works with the standard controller and with the Move, and then measure both and compare the two numbers.
Also, lag can be compensated with good motion prediction. However, good prediction requires a low latency and a quick measurement cycle, which should run independently of the rendering cycle.
This should serve as a reminder: Set your TV to "game" mode!
Most late-model TVs do various image-processing operations to make the picture look better. This can add a frame or two of latency ("lag") to the display. However, they usually have a "game" mode which turns off this extra processing and displays the input image as quickly as possible. For fast-action games, you should always enable that mode on your TV.
And there you go PS3 fans~
I would think the problem is the Playstation Eye. I mean that thing is just a glorified VGA webcam, nothing fancy. The fact that Sony could pull this using that thing is probably a feat by itself. Sony should've re-designed the camera too.
@pika2000 The PS3 Eye is more than a VGA webcam, and Sony did indeed work with the design of it. It's capable of 60 FPS VGA video, and 120 FPS QVGA video. The video is uncompressed. It has optical zoom (if only 2 settings) and much better light sensitivity than most average webcams. It also has an array of 4 microphones, rather than just one.
I'd prefer to see a test done with a CRT just to remove any display latency from the equation.
Hey, is the Move controller ready for launch yet? No? It still has a while? Then STFU about the lag.
Useless experiment, pretty aficionado btw.
the screen give a lag, the camera give a lag, the human response give a lag..
What's the lag from a regular PS3 controller? If it's similar to the Move we don't really have anything to worry about. Any ideas?
While enjoyable as the game was, I'm sure it beats out the awful controller lag that Darksiders had me endure. So yea, not to worried if it even beats out GTA VI, which I didn't seem to notice enough to tell.
Jump Stupid! JUMP!
@Merikov
GTA IV*... Romans and their accursed numerals