On Tue, 02 Oct 2007 19:54:39 -0400 kony <spam@spam.com> wrote:
| On 2 Oct 2007 12:54:39 GMT, phil-news-nospam@ipal.net wrote:
|
|
|>| Of course lower is technically possible but not a practical
|>| design goal for a PC video card. Thus, video cards really
|>| don't go lower because you'd have to change the default
|>| behavior including the driver settings, which nobody in
|>| their right mind would do to output from a PC to an LCD.
|>
|>Why not? If the video card has an upper limit on pixel clock AND you
|>want a larger geometry, then slowing the horizontal (to get more pixels
|>per line) and the vertical (to get more lines per frame) is simply going
|>to be what you do.
|
| No you don't _want_ to do that, because there's no gain in
| doing it, only degradation in visual quality. This is until
| the data link bandwidth is exceeded, at which point a second
| data link is used. It is used for a reason - to preserve
| visual quality and/or allow for it, instead of degrading it
| with your proposed solution.

There is no second data link. A card with a second data link would cost
more money. The geometry can be achieved with a lower frame rate with
the same video card, with no degradation in quality for a great many
uses of graphical displays, including most business uses.


|>|>FYI, high definition TV signals range from 60 Hz at the highest down to
|>|>a low of just above 23.976 Hz. All LCDs should be required to go down
|>|>to as low as 23.976 Hz.
|>|
|>| You're talking about frames per second, yes? That's not
|>| what is being referred to previously by 56Hz, it does not
|>| only signal at the rate of frame change and no PC use LCDs
|>| need to support below 60Hz to display HD.
|>
|>Then what do you think the 56 Hz refers to? The horizontal? The
|>rate your eye blinks?
|
| Fair enough, it is the framerate but the 24Hz is a lower
| limit one doesn't want to cause, it is only due to the
| format of the video content. There is no need to cause
| this state.

The motion picture industry does not seem to agree with you. Of course
a motion picture projector would produce flicker if run at 24 shutter
openings per second. The figured out ages ago that by opening the
shutter 48 or 72 times a second, timed so that the openings were of the
same timing between them, where 2 or 3 openings would have the same
frame, that this eliminated the flicker (even though it did not change
the total accumulated time the shutter was open) for most eyes.

Were it the case that motion that changed only every 1/24 of second
were annoying, too, they would have change the system to a higher frame
rate. They did for special cases (action sports), but not for the
average movie.


|>What the TV world refers to as frame rate is what the computer world
|>refers to as vertical frequency. A "1080p24" HD TV signal has a frame
|>rate of 24 frames per second (or 1000/1001 of that rate if locked to
|>legacy NTSC reference clock timing), and that is 24 Hz or 23.976 Hz.
|
| This doesn't matter, it is only a lower limit. Slowing down
| one end to the slowest the other would ever require is a
| backwards concept. On a PC (thus using LCD PC monitor),
| both ends can support faster than that.

Making a video card that has a faster pixel clock rate costs more
money. Since these clock rates are already well into the UHF range,
it's not just a matter of faster oscillators. It requires better
quality components overall. It requires better silicon technology
to shift data bits around faster. It requires faster DACs in the
case of analog output. It requires the second DVI link because doing
such high rates over one link costs even more than two links costs.

Making a monitor capable of running slower (lowering the low end,
while keeping the high end where it is at, of the sampling range)
costs very little. The engineering design changes would require
parameters in the software to know of the slower rate, and change
the synthesizer to handle it by increasing the divide ratio range
by merely adding 1 or 2 bits. It may be the case that the parts
can already do this, and it is the software that imposes the limit.


|>Many video cards have pixel clocks with limits in the low 100's of MHz
|>and plenty enough memory to run with geometries of 2048x1536 or more.
|>With the max pixel clock and full utilization of memory to get higher
|>geometries, you do get lower frame rates (vertical Hz).
|
| Why would you use an old video card that can't support 60Hz
| with a newer high resolution LCD? That degrades the quality
| regardless of a video/HD content scenario.

It can support 90 Hz at the geometries of the day it came out. It
also is flexible enough that it can support extremes in horizontal
geometry or extremes in vertical geometry. And it was an open
architecture not encumbered by trying to hide how to use the card
in software other than their own.

Now if I combine extreme horizontal AND extreme vertical ... it works.
That is, the _card_ works. But logically, when you divide the clock
by the extreme horizontal and then again by the extreme vertical, you
get something below 50 Hz (the lower end of most LCD monitors it would
seem).


|>I have an _old_ Matrox Millennium card that can run up to 1280x1024.
|>But at that geometry, even with the pixel clock set to the maximum the
|>card supports, the frame rate (vertical Hz) is so low it flickes like
|>mad on a CRT. Switch to LCD and it's great ... if the LCD can go that
|>low.
|
| So the problem is not refresh rate, it's trying to use
| ancient hardware. Same could be said about trying to use
| any very old low performance hardware for a modern more
| demanding use.

It's trying to use open architecture hardware that can do things that no
modern card can do ... work with independent software.


|>CRT is on the way out. LCD is on the way in (although OLED may bump
|>that in a few or so years). Old assumptions about flicker just do not
|>apply, anymore.
|
| Never said they did, but one could call the jerkiness of
| 24Hz flicker-like as well. 24 FPS is visually noticable on
| an LCD.

Tell me which LCD monitor you have seen 24 fps video displayed?
Manufacturer and model, please.


|>Some people _need_ high frame rates for legitimate purposes, such as
|>gaming, and watching high action video.
|
| Most people will benefit from more than 24FPS in typical
| motion video uses. Most people don't have the problem you
| suggest and attempt to solve by lowering the rate to 24.

They can benefit by less costly hardware, whether older or newer, by running
the video at 24 fps when their type of use of the computer is compatible with
that slow frame rate.

When the type of use changes to something that needs a higher frame rate, I
most certainly expect the LCD monitor to handle it. I've never suggested
that the upper limit frame rate be reduced. Although broadcast video is
limited to no more than 60 fps (50 fps in Europe), as I understand it,
double that rate looks even better for certain high action video, such as
the race cars I like to watch. So I look forward to the day when the can
cover a NASCAR race with 1920x1080 progressive video at 120 fps. This will
need about 20 MHz of spectrum with 8VSB modulation as used in over the air
TV, and about 10 MHz of spectrum with 256QAM modulation as used over many
cable systems. This would pose problems for them since they slice the
spectrum up in 6 MHz chunks. Satellite, however, may be able to do it.

--
|---------------------------------------/----------------------------------|
| Phil Howard KA9WGN (ka9wgn.ham.org) / Do not send to the address below |
| first name lower case at ipal.net / spamtrap-2007-10-02-1924@ipal.net |
|------------------------------------/-------------------------------------|