Unusual Frame Update Behavior with Philips 328P6VU

[mpabst]

Hi all, I hope this the right place to post but I've been conducting tests using a Philips 328P6VU monitor and came across an intriguing anomaly. My setup involves using a photosensor, which samples at a rate of 2048 Hz to understand what the monitor is doing. For comparison purposes, I ran my test three times, at different screen regions (top left, center, and bottom right). The monitor was operating at 60 Hz, cycling through 4 white frames and 4 black frames, effectively switching every 4 frames. To establish a baseline, I initially conducted a similar test on a different 1980x1200 monitor at 60 Hz. Below, you can find a plot depicting the photosensor's responses at the three distinct locations:



As expected, there is a delay between the moment I finish drawing a frame (flipping the front and back buffer in OpenGL). Frame pacing is indicated by the faint gray lines every 1/60 s. It appears that it takes precisely one frame to complete the frame drawing process (delay between top left and bottom right), which aligns with my expectations.

Now, when we examine the same plot for the Philips 328P6VU, a different pattern emerges. It appears that it takes wo frames to update the entire screen (the distance from dark red to green covers two of the faint gray lines):



Is this a known issue or behavior? This observation suggests that at any given time, there are three frames displayed on the screen, which seems unusual.

[Chief Blur Buster]

Backlight PWM dimming transients.

They're very very very annoying when you're trying to measure LCD GtG behaviors.

For GtG testing, use brightness 100% to avoid these scientific measurement interferences.

[mpabst]

Hey, could you expand on that? Both monitors where set to 100% brightness, so PWM should not be an issue...