Screen Flickering

[Lord Vetinari]

Anyone explain why you do not see screen flicker at the cinema with a projector running at 24 frames/sec whilst you do with a CRT running at 60HZ especially as the projection screen has no image persistence whilst CRT phosphors do?

[Lord Vetinari]

Nobody have an explanation?
shameless bump!

[headbox]

actually if you are sitting close to the movie screen and the shot pans accross some scenery it can be quite painful to your eyes.

I have read that our eyes can detect up to about 70 images a second, but will settle for 24. I'm guessing on a monitor that is very bright with sharp lines, it is easier to see the flicker.

...then again I dont see flicker on an LCD monitor at 60hz, so who knows.

[Moridin]

Two factors come into play. The first is persistence. While I don’t know exactly how this applies to a movie screen/projector, the idea is to have the surface continue to “shine” while there is no new signal or light going to it.

My guess is that with a movie, the time between frames is very short, so it doesn’t require the image to be maintained for very long. On a monitor, the story is a little different, since only 1 pixel is lit up at a time, they need to “save” their value for a much longer period of time relative to a movie projection.

The other factor is that movies double flash their images, so the actual projection rate ie 48 frames per second.

[Ramuman]

Originally posted by headbox
actually if you are sitting close to the movie screen and the shot pans accross some scenery it can be quite painful to your eyes.

I have read that our eyes can detect up to about 70 images a second, but will settle for 24. I'm guessing on a monitor that is very bright with sharp lines, it is easier to see the flicker.

...then again I dont see flicker on an LCD monitor at 60hz, so who knows.

I wonder how short a period of time the eye can discern. On a SLR film camera of mine, its possible to see a flash (not the actual camera flash, which only syncs upto 1/250th of a second on my camera, but an actual pulse of light from the surroundings) of an image with the back door open at speeds of 1/2000th of a second on faster, its impossible to see the actual image since its so short, but you definitly know the shutter opened. I wonder how short a pulse the human eye could detect. I suppose it depends alot on the display method. There's probably a cutoff at which the eye (or brain rather) begins to mesh a series of pictures together to create motion, but the actual shortest period of time in which the eye can detect a change is likely much shorter -- though I presume it just tapers off from being able to clearly identify an image to detecting a pulse or flash of light to seeing nothing at all.

[Lord Vetinari]

[QUOTE]Originally posted by Moridin
[B]My guess is that with a movie, the time between frames is very short, so it doesn’t require the image to be maintained for very long. On a monitor, the story is a little different, since only 1 pixel is lit up at a time, they need to “save” their value for a much longer period of time relative to a movie projection.

Surely the screen is lit up only every 24th or 48th sec and has no persistence whilst each phosphor pixel is lit up every say 60th sec and has a persistence approaching the same value (ie 60th sec) so it should not flicker.

I wonder if its got anything to do with relative brightness ie cinema is very dark and screen is very bright whilst CRT screen is only moderately bright but room is normally light as well.

[Adisharr]

Motion blur captured on film is another reason...

[Moridin]

Originally posted by Lord Vetinari
[B

Surely the screen is lit up only every 24th or 48th sec and has no persistence whilst each phosphor pixel is lit up every say 60th sec and has a persistence approaching the same value (ie 60th sec) so it should not flicker.

[/B]

There will be some persistence based on the surface you are projecting the movie on.

The point I was making is that the duty cycle is likely much longer for a movie. Unlike a monitor, a movie picture is projected all at once, and only needs to have a very short off time between frames. Thus the time you go with no light projected can be shorter for a movie then for a monitor despite the lower frame rate.

Each pixel on a monitor is off for over 16 ms. If a movie projector can get a 75% duty cycle, the off time with no light being projected on the screen, would be around 4 ms, and I would not be surprised if it was lower then that.

Certainly lighting will make a difference as well though.

[russ_watters]

Actually for movies the persistance isn't in the image on the screen, its in your eyes. Since a movie is all reflected light intermittently blocked by a shutter there is exactly ZERO image persistance on the screen. But its like flash-blindness and thats why you see spots after looking at the sun and can write letters in the air with a sparkler or glow stick. And thats why movies are best viewed in the dark. The contrast between the dark background and the bright movie causes the light receptors in your eye to overload and continue sending a signal even after the light is gone. And I think if you turn down the intensity of the light and raise the house lights you can see the flickering.

This is different from jerkiness caused by framerate though. I sat too close to a screen in a crowded movie last week and I could see the stuttering when objects moved quickly across the screen. Thats a matter of field of view vs framerate. Its like resolution actually - sitting too close is like too low of a resolution. When objects move across the screen they cover more ground faster when you sit closer, so the steps also need to be closer together. Motion blur for digital effects helps, but you see stuttering even in non-digital motion.