Re: Best tv for pc resolution

On Thu, 27 Sep 2007 00:50:25 +0100, Roderick Stewart
<escapetime@xxxxxxxxxxxxxxxxxxxxxx> wrote:

In article <2p7lf314dfrpkargfhslhrsi0oqvjhk078@xxxxxxx>, Colin Stamp wrote:
I'm looking at a 1600x1200 screen right now. I'm sitting quite close to it 
looking at quite small text. I can't even be sure without checking the cables 
whether it is fed from the digital or analogue output from the computer, 
because I have tried both and cannot see any difference at all.

We've been though this before. You only have one system, you haven't
said whether it's LCD or CRT, and 1600 is significantly less than
1920.

You're right. Checking back I see that I didn't actually specify the type of
monitor in those terms, though I did say that my *monitor* was 1600 x 1200, which
would not be the case if it was a CRT, because the displayed resolution would
depend on the signal fed to it. Evidently that didn't make it as clear as I assumed
it would. Sorry for the misunderstanding.

It was indeed a bit too obscure for me. Apologies for missing it.

It is actually an LCD type

Well, that's that cleared up then.


However, I would dispute "significantly". Yes there will be a difference in the
frequency response required of the video amplifier, but it is not significant. I
haven't worked out exactly what it would be, but you suggested 20% and that sounds
about right. Hardly enough to make a difference unless the video amplifier had a
fairly steep HF filter somewhere between the upper limits of the two video signals,
which would be highly unusual as the filter itself would do more damage than
anything else.

I haven't worked out the bandwidth difference either. It's 20% more
pixels per line. You'll have to agree that there is an upper limit to
the horizontal resolution that the interface can handle reliably. We
can also safely assume that the limit varies wildly from one
implementation to another.


I also wonder how you think you can say that I have only one system. We are not
aquainted outside this newsgroup, so how could you possibly know?

Ah, it was an assumption on my part, based on you not having mentioned
any others. I was only referring to systems which are (supposedly)
relevant to this discussion. I certainly didn't mean to belittle your
overall computing power.

The best
experiment I've been able to try was on my main system, as described, because both
the graphic card and the monitor can handle digital and analogue signals, so it
really is an "all other things being equal" situation, but it is not the only
system I have, and certainly not the only one I have seen. I actually have three
working computers at the moment, but must have built dozens over the years, for
myself and other people. I am familiar with the behaviour of CRT and LCD monitors,
video amplifiers, and various ways of connecting them together, and have used them
for a variety of purposes.

It doesn't need to be an all-other-things-equal job. If it works on
VGA then that's one point for the "VGA always works" camp, but it's
still not 1920 pixels per line.




If a video amplifier works with one monitor but not 
another, or if it handles one video signal but not another in which the upper 
frequency limit is only 20% different, then it's badly designed.

And you don't think these "badly designed" ones are out there in
significant numbers then? The vast bulk of video cards haven't been
designed by broadcasters.

I'm sure there's a lot of rubbish out there, but it doesn't tell us anything
fundamental about the relative merits of digital versus analogue connections
between computers and monitors through 2 metres of cable. If something doesn't work
very well because it's badly designed, then that's the reason, not the fact that
it's analogue, or whatever.

This is the crux of our disagreement, and it seems to be bogus. I'm
not saying, and never have said, anything about the relative merits of
analogue versus digital video connections in general.

I'm talking about one particular interface, which became the de-facto
standard for PC monitors many years ago, and has been independently
developed by numerous manufacturers over the years. In the beginning,
there were only CRT displays, so the interface was designed without a
pixel-sync. Each entire line of pixels has to fall exactly into the
correct slots (using accurate timing alone) for a flat-panel to
display it properly. This was never an issue with CRTs.

All this has resulted in a very untidy situation where the VGA
interface can work in some situations but not others. Hence my
original comment that it doesn't really work properly. It's not
because it's analogue, it's because nobody expected it to be used for
displays with discrete pixels.

Like you say, there's a lot of rubbish out there, but it's real-world
rubbish and we have to take that into account when selecting displays
and signal sources that must be inter-operable.

If I were to buy a 1080 TV and a PC which had to work together, I
would make sure the PC had a DVI output and the TV had a spare input
to plug it into. Wouldn't you do the same?


If a video amplifier gives a performance that deteriorates as it warms up, then 
it's *very* badly designed. I'd suspect a fault in the monitor first.

Nope. It's not the TV. The effect is independent of the TV
temperature, but dependant on the PC temperature. I guess the video
card is a bit crap It was cheap after-all. It's disappointing that it
doesn't work perfectly at 1380 horizontal, but not a huge surprise.

Well, there's your answer - not a very good video card.

Indeed. Not very good and probably not very atypical.


Incidentally, the same PC worked apparently perfectly at 1600X1200 on
a CRT monitor for some time before being switched to TV duty.

I'd suggest that agrees with the above.

I'd suggest that it supports my assertion that a slightly ropey video
signal can show up more on a display with discrete pixels.


True, but I don't think this can have anything to do with the video amplifier 
or the cable.

Of course it does. The interface can be a bit dodgy and it'll just
marginally reduce the sharpness on a CRT screen - nobody will be any
the wiser. Use that same interface to drive an LCD display at the same
resolution, and the deficiency will be really obvious.

Assuming both monitors - LCD and CRT - are capable of displaying the full bandwidth
of a given video signal, can you explain the mechanism which will make a reduction
in HF of the signals fed to them more apparent on one display than the other? I
can't think how this could happen, and have never seen the effect.

Well, if you want to concentrate only on bandwidth, then Reducing the
HF of the signal to an LCD will, after a certain point, cause
individual pixels to interact with their neighbors. This is very
noticeable on displays which are expected to be razor sharp, with
perfectly rectangular pixels. On a CRT it'll produce a softening
effect which many people might not notice, or might happily tolerate.

But, of course, there are many more distortions which might affect the
signal other than bandwidth limitation. Now I think about it further,
the system that goes dodgy when the PC warms up is probably down to a
cheap crystal oscillator on the graphics card drifting with
temperature. Something that a pixel sync would completely sort out,
and something which would only show up on a CRT as a minute change in
picture size.

Cheers,

Colin.
.

Quantcast