Phil Weldon wrote:

I am just indicating that rating performance by frames per second may allow
comparisons, but is it a USEFUL comparison, especially since we all seem
perfectly satisfied by movies at 24 frames per second and television
displays at 30, 50, or 60 frames per second (I'm sorry, but PAL and SECAM at
25 frames per second gives me a headache.)

Yeah. PAL and SECAM 25 FPS (50Hz refresh) DOES flicker, their claims to the
contrary notwithstanding. I notice it too.

While I'm not sure I buy the whole theory, partly because it's not
something I've spent a lot of time on, there IS research which shows a
perceptible difference with frames rates 'too high to see'. I.E. faster
than the monitor refresh interval.

The postulated reason is that, with movie and TV, you're taking a
'snapshot' of real life movement, not a 'frozen in time' stagnant image
(things don't stop moving for your snapshot to take place), so there is
'smearing' of it over the frame interval and, they think, this provides
additional cues to the eye.

With computer generated frames, however, they ARE simply one stagnant image
after another, computer shifted the 'right amount', 'full frame' at a time,
to the next image to simulate movement. The idea is that faster than the
refresh rate frame generation creates a more lifelike 'moving picture' that
the refresh rate is then taking the 'snapshot' of, kind of like how 'real
life' is moving all the time as the frame is taken.

Seems to me that, if it were 'perceptible', it would appear more like
tearing, since it isn't as if the entire image were moving, only 'part' of
the frame would be in the 'new position', but then I haven't run actual
human tests so I would be speculating whereas others claim to have observed
it. Also, by perceptible they don't mean consciously observable, just that
the observers seem to feel that the 'too fast' frame rates are 'more
realistic'. Maybe the eye compensates for the 'partial' smear just as it
does for flicker and in recreating full color from 3 primaries. That would
make me think there is some minimum multiple (maybe an odd multiple so it
cycles through the image) before the effect would be effective, again, like
a minimum rate to remove flicker.


As for my personal use, I like the price on display adapters two generations
behind the bleeding edge. Paying $400 US or $500 US for performance that is
only helpful for a handful of 3-D game programs is a very expensive
performance boost for a very limited use. Money spent on boosting the
performance of your entire system for a wide range of uses is better spent.