Has anybody ever measured the input lag of emulators in relation to the real hardware?
When it comes to input lag within the same emulator according to different configurations, it's pretty easy to check:
With this method, you will realize that vsync in Direct3D fullscreen has about three frames of lag in relation to any configuration without vsync.
For some curious reason, Direct3D window mode, DirectDraw window mode and DirectDraw fullscreen all have only 0 to 1 frame of input lag when vsync is enabled.
This of course still doesn't say anything about the absolute input lag. For this, you would need to put an LED on the actual controller and use the same TV/monitor as for checking the emulator lag.
When it comes to input lag within the same emulator according to different configurations, it's pretty easy to check:
- Set your monitor to 60 Hz.
- Map a visible action (like a jump) to one of the keyboard keys that have an LED: Num lock, caps lock or scroll lock.
- Take a video camera that can record in 60 FPS. Point it at the screen, so that the screen and your keyboard are visible at the same time.
- Press the key to perform the action on the screen.
- Watch the video frame by frame and count the frames between the LED lighting up and the action on the screen becoming visible.
With this method, you will realize that vsync in Direct3D fullscreen has about three frames of lag in relation to any configuration without vsync.
For some curious reason, Direct3D window mode, DirectDraw window mode and DirectDraw fullscreen all have only 0 to 1 frame of input lag when vsync is enabled.
This of course still doesn't say anything about the absolute input lag. For this, you would need to put an LED on the actual controller and use the same TV/monitor as for checking the emulator lag.