tft monitor - which one should I choose?

[Guest]

Archived from groups: comp.sys.ibm.pc.hardware.video (More info?)

On Sat, 26 Jun 2004 19:23:35 +0000, Bob Myers wrote:

> "chrisv" <chrisv@nospam.invalid> wrote in message
> news😱efod053s0rmnegnu4gcuvuh42fr8660uk@4ax.com...
>> I'll tell you, Bob, right after you tell me what "big difference" the
>> thick glass on the front of a CRT makes.
>
> That one's pretty easy - the thick glass (required by the fact that the
> CRT faceplate has to resist a considerable amount of air pressure, and
> made thicker when you go to "flat-face" CRTs) has a couple of very
> objectionable optical effects - especially troublesome in desktop
> monitor use, since the typical viewing distance is on the same order as
> the screen diagonal. The first is the very obvious impact the glass has
> on the visual uniformity of the image - since you ARE using the thing
> with your eye relatively close to the glass, you're looking through a
> good deal more glass when you view the sides and corners of the image
> than you are in the center.

I calculate the difference due to this angle at 12% more glass at the
corners than in the middle, assuming viewing distance = diagonal.
This small difference in absolute brightness is compressed by the
logarithmic response of human vision. It's generally not noticeable.
No "big difference", Bob.

> Monitor CRT glass is pretty much always tinted - with a transmission
> generally in the range of 50-90% - as a contrast-enhancement technique.

LCD's have tint too, Bob.

> So, you wind up with the
> outside of the image looking quite a bit dimmer than the inside.

Bull. "Quite a bit dimmer" indeed.

> (And since the tint is never perfectly neutral, there are similar
> impacts on the color uniformity.)

Yet CRT's are still better than LCD's in this regard. From
http://website.lineone.net/~del.palmer/lacie.html

<quote>
It is also true that CRT monitors are at least 2 to 3 times more
accurate when it comes to displaying color than LCD screens even when
both are displaying 24-bit color and both are measured and calibrated
with a colorimeter. CRT is able to maintain color uniformity across
the screen 2 to 3 times better than an LCD as well. </quote>

No "big difference" in favor of the LCD, Bob.

> The second effect is refractive; since you are looking at the outer
> extremes of the image through the glass at an angle, the light from this
> portion is refracted differently (as you see it) than the center. This
> leads to a number of distortion effects - most notably, the appearance
> of a concave ("bowed inward") image when truly flat faceplate glass
> (which was very thick) was first tried a number of years ago (as in the
> old Zenith "FTM" tube design).

Using 15-year-old CRT designs to support your case, Bob? On my
(modern) CRT's, there is NOT any refractive distortion that any
reasonable person would describe as a "big difference". I can't
notice any at all. The geometric distortions that I CAN notice are
not caused by the "thick glass".

>> Doesn't surprise me at all, Bob, that you'd intentionally ignore the
>> point I was making (that spam_eliminator's statement was ridiculous),
>> and hypocritically attack it, with no mention of what spam_eliminator
>> said.
>
> I didn't ignore the point you were making - it was simply wrong, and
> spam_eliminator was right.

His point is technically correct in that an "ideal" monitor would have
nothing in-between the graphics and your eyes. However, I maintain
that his claim that the glass makes a "big" (presumably negative)
difference in the image quality vs. a LCD is ridiculous, since LCD's
are also non-ideal in this regard. Sorry, Bob, but my point is quite
valid.

> Thick faceplate glass is NOT
> desirable, for the reasons I gave above.

Will you now be fair and admit that LCD's are also imperfect in their
transmission of the graphics to your eyes, Bob? I'd like to see you
demonstrate your impartiality by posting a similar critique of the
LCD's light-transmission compromises.

> And I would have thought that a CRT expert such as yourself would have
> already been quite aware of such concerns, so they didn't need to be
> repeated here.

I know enough to have recognized spam_eliminator's biased, unfair
remark, Bob.

> Guess I was wrong about that, huh?

Once again, you (intentionally, I think) missed the point, Bob.

 

[Guest]

Archived from groups: comp.sys.ibm.pc.hardware.video (More info?)

"chrisv" <chrisv@nospam.invalid> wrote in message
news:vcqle0hl430pc0p9b5idh5vsqu8j7tqsjm@4ax.com...
..
> >- so what ELSE could we conclude, other than that
> >you believe the LCD is "being pushed"? Please offer another
> >alternative, if you have one.
>
> As we've already discussed, Bob, the "coolness" factor, and the
> "newer/flatter must be better" factor, are huge. The above factors
> are, in general, NOT tempered by consumer knowledge of the performance
> trade-offs involved.

And what, in YOUR view, is the source of the "coolness" factor,
or the notion that "newer/flatter" must be better - notions that you
again apparently disagree with, or at least do not consider to be
valid reasons for making this choice - if not "the LCD being pushed"?
The above makes it perfectly clear that YOU feel that the consumer
is ignorant of the "performance trade-offs involved" (could you
possibly be a little more specific?), and therefore is making a buying
decision based on what YOU consider invalid or irrelevant factors.


> >More naivete on your part, then
>
> Incorrect, and another completely unsupported charge from you, Bob.
> What I wrote is entirely reasonable and hardly evidence of any alleged
> "naivete" on my part.

On the contrary - what you have written appears to be ample
evidence of a lack of familiarity with the monitor market on your
part.


> > - it seems pretty clear from the
> >market response that the cost advantages of the CRT do not
> >outweigh its shortcomings in the minds of the buying public.
>
> More illogic from you. What's more popular is not a measure of
> goodness. GM sells more cars than Honda, for example. FWD cars
> outsell RWD cars by orders of magnitude - am I "naive" when I claim
> the RWD is better?

At the very least, you're being overly simplistic. If you make the
flat, unqualified claim that "RWD is better," then you force us to
ask the question, "better for WHAT?" How YOU define "better"
very likely does not apply to all usages or applications, and so such
blanket claims wind up being nonsensical.

Similarly, to make a blanket statement that "CRTs are better" would
be equally nonsensical - such things need to at the very least be
qualified with a statement regarding the specific application in question,
or noting that such a statement simply constitutes your personal
preference (which clearly is not subject to objective, quantitative
analysis).


> For various reasons, some of which are valid, others of which are
> related to extracting more money from our wallets.

There you go again - that unsupported assertion, that flies in the
face of everything we know about how this market evolved over the
past 15 years.


> My mind reels from the illogic, and the veiled insult that I could be
> foolish-enough to think any such thing. Do you think I'm an idiot,
> Bob?

Well, I certainly don't see anything to be gained by giving you an
honest answer to THAT one...you'll just get angry again.


Bob M.

 

[Guest]

Archived from groups: comp.sys.ibm.pc.hardware.video (More info?)

"chrisv" <chrisv@nospam.invalid> wrote in message
news:h5ule0hr56jnqnbfgf8p73q9m9oa6islhj@4ax.com...
> I calculate the difference due to this angle at 12% more glass at the
> corners than in the middle, assuming viewing distance = diagonal.

And also assuming, apparently, that the glass IS simply a flat sheet
rather than the somewhat more complex shape of the typical faceplate.
Further, you assume that the light emission from the phosphor screen
is spatially uniform rather than (at best) Lambertian.

Your figures would be a lot more impressive if they were measured
rather than simplistically calculated.


> > Monitor CRT glass is pretty much always tinted - with a transmission
> > generally in the range of 50-90% - as a contrast-enhancement technique.
>
> LCD's have tint too, Bob.

Gee, Chris, do you think I don't know that? However, please
check in with the glass makers and see what the transmission of the front
substrate glass itself is in the typical LCD panel. See what the
OVERALL transmission is of the entire stack past the LC material
itself, and compare that with the CRT faceplate.

> > So, you wind up with the
> > outside of the image looking quite a bit dimmer than the inside.
>
> Bull. "Quite a bit dimmer" indeed.

Well, since you're clearly possessed of considerable experience in
this area - what do you believe is the typical luminance uniformity
profile of a CRT? How does it compare with the typical LCD?


> Yet CRT's are still better than LCD's in this regard. From
> http://website.lineone.net/~del.palmer/lacie.html

I looked at the site; suffice it to say that anyone can post a web
page, and some know more than others what they're talking
about.


> It is also true that CRT monitors are at least 2 to 3 times more
> accurate when it comes to displaying color than LCD screens even when
> both are displaying 24-bit color and both are measured and calibrated
> with a colorimeter. CRT is able to maintain color uniformity across
> the screen 2 to 3 times better than an LCD as well.

Which is a rather easy statement to write in the absence of data.
Want to see some, from a real test lab?



> Using 15-year-old CRT designs to support your case, Bob? On my
> (modern) CRT's, there is NOT any refractive distortion that any
> reasonable person would describe as a "big difference". I can't
> notice any at all. The geometric distortions that I CAN notice are
> not caused by the "thick glass".

Would you care to quantify that? If not, again, would you like to
see some numbers? (I'm somewhat surprised that you're willing to
admit to any geometric distortions at ALL, but....)



> His point is technically correct in that an "ideal" monitor would have
> nothing in-between the graphics and your eyes. However, I maintain
> that his claim that the glass makes a "big" (presumably negative)
> difference in the image quality vs. a LCD is ridiculous, since LCD's
> are also non-ideal in this regard. Sorry, Bob, but my point is quite
> valid.

Let's see - we're agreed that glass makes a difference, and yet you
don't think there might be a "big difference" between a display
technology in which that glass is many millimeters thick and one
in which the comparable glass is under a millimeter thick. Wanna
try that one again?



> Will you now be fair and admit that LCD's are also imperfect in their
> transmission of the graphics to your eyes, Bob? I'd like to see you
> demonstrate your impartiality by posting a similar critique of the
> LCD's light-transmission compromises.

Of course; to use YOUR favorite debate tactic, it is "dishonest"
of you to imply that I have ever said that the LCD represents
anything remotely like a "perfect" display technology. However,
I am very certain that I have a considerably better understanding of
the sources - and likelihood of correction - of the imperfections
in the two technologies than you do.

Bob M.

 

[Guest]

Archived from groups: comp.sys.ibm.pc.hardware.video (More info?)

"Bob Myers" <nospamplease@address.invalid> wrote:

>> > So, you wind up with the
>> > outside of the image looking quite a bit dimmer than the inside.
>>
>> Bull. "Quite a bit dimmer" indeed.
>
>Well, since you're clearly possessed of considerable experience in
>this area - what do you believe is the typical luminance uniformity
>profile of a CRT? How does it compare with the typical LCD?

See below, Bob.

>> Yet CRT's are still better than LCD's in this regard. From
>> http://website.lineone.net/~del.palmer/lacie.html
>
>I looked at the site; suffice it to say that anyone can post a web
>page, and some know more than others what they're talking
>about.

The text I quoted from the link above is attributed to Lacie, a
respected manufacturer of professional display monitors. Hardly just
"anyone", Bob.

Here's something else - a quote from NEC/Mitsubishi, found at:
http://www.necmitsubishi.com/support/css/monitortechguide/index11.htm

<quote>
Based on the current core technologies, CRT monitors are able to
display a wider color space than LCD monitors and deliver more
consistent brightness uniformity throughout the screen.
</quote>

Gee, Bob, that seems to blow your entire argument, that the CRT's
thick glass causes it to have inferior luminance uniformity, right out
of the water.

 

[Guest]

Archived from groups: comp.sys.ibm.pc.hardware.video (More info?)

"Bob Myers" <nospamplease@address.invalid> wrote:

>"chrisv" <chrisv@nospam.invalid> wrote:
>>
>> >More naivete on your part, then
>>
>> Incorrect, and another completely unsupported charge from you, Bob.
>> What I wrote is entirely reasonable and hardly evidence of any alleged
>> "naivete" on my part.
>
>On the contrary - what you have written appears to be ample
>evidence of a lack of familiarity with the monitor market on your
>part.

Repeating your illogical, antagonistic conclusions does not make them
any more true, Bob. I note that, once again, despite my direct
request, you failed to actually dispute anything that I said in the
paragraph in question. Instead you again chose creative editing as
your "solution" of choice.

In any case, Bob, as much as you'd like to change to subject, the
issue of this little sub-thread is the allegations that you made
regarding my postion, when you wrote "Chris' standard response - CRTs
are losing the market just because the big bad monitor and system
makers pushed a clearly inferior product on an ignorant and
easily-duped public."

The rest is mostly your flopping-around trying to justify these
illogical, unsubstantiated, and untrue allegations.

>> > - it seems pretty clear from the
>> >market response that the cost advantages of the CRT do not
>> >outweigh its shortcomings in the minds of the buying public.
>>
>> More illogic from you. What's more popular is not a measure of
>> goodness. GM sells more cars than Honda, for example. FWD cars
>> outsell RWD cars by orders of magnitude - am I "naive" when I claim
>> the RWD is better?
>
>At the very least, you're being overly simplistic.

That's rather ironic coming from you, Bob, who has been attempting to
over-simplify the monitor market, such as when you wrote "There appear
to be only two broad possibilities, here - either CRTs are losing
marketing share because the LCD IS "being pushed," or they're losing
market share because the LCD is the superior product."

>If you make the
>flat, unqualified claim that "RWD is better," then you force us to
>ask the question, "better for WHAT?" How YOU define "better"
>very likely does not apply to all usages or applications, and so such
>blanket claims wind up being nonsensical.
>Similarly, to make a blanket statement that "CRTs are better" would
>be equally nonsensical - such things need to at the very least be
>qualified with a statement regarding the specific application in question,
>or noting that such a statement simply constitutes your personal
>preference (which clearly is not subject to objective, quantitative
>analysis).

No kidding, Bob, this is the same point I've repeatedly tried to make
about the monitor market, but you just come back with more of your
self-serving, illogical, oversimplifications. Go back and read my
last post (message ID: <vcqle0hl430pc0p9b5idh5vsqu8j7tqsjm@4ax.com>)
to see abundant evidence of what I am talking about.

I hope, now that you appear to understand these concepts, that we
won't see your erroneous oversimplifications again.

Maybe we'll also see an end to your misrepresenting my position, as
you did when you said that I think that LCD's are a "clearly inferior
product", and that I think that LCD's are gaining market share "just
because" they are being pushed.

Face it, Bob, no matter how much convoluted illogic you throw at it,
I've never said anything which would lead anyone to logically conclude
that I thought either of those things. The situation is obviously too
complex for "clearly inferior" and "just because".

>> For various reasons, some of which are valid, others of which are
>> related to extracting more money from our wallets.
>
>There you go again - that unsupported assertion, that flies in the
>face of everything we know about how this market evolved over the
>past 15 years.

LOL. Sure, Bob, it's got nothing to do with money. They have no
motivation at all to sell you a $300 monitor instead of a $100
monitor. It's simply not a factor. </sarcasm>

Sheesh. Talk about naive...

>> My mind reels from the illogic, and the veiled insult that I could be
>> foolish-enough to think any such thing. Do you think I'm an idiot,
>> Bob?
>
>Well, I certainly don't see anything to be gained by giving you an
>honest answer to THAT one...you'll just get angry again.

Angry if you arrived at yet another illogical, unsupported conclusion,
Bob? No, but I do think your subtle ad hominem attacks are worth
noting...

 

[Guest]

Archived from groups: comp.sys.ibm.pc.hardware.video (More info?)

"Bob Myers" <nospamplease@address.invalid> wrote:

>chrisv wrote:
>>
>> Will you now be fair and admit that LCD's are also imperfect in their
>> transmission of the graphics to your eyes, Bob? I'd like to see you
>> demonstrate your impartiality by posting a similar critique of the
>> LCD's light-transmission compromises.
>
>Of course; to use YOUR favorite debate tactic, it is "dishonest"
>of you to imply that I have ever said that the LCD represents
>anything remotely like a "perfect" display technology.

I never implied any such thing, Bob. I simply think it's less than
fair of you to nit-pick the CRT's flaws, while denying us the benefit
of your expert critique of the LCD's imperfections "in their
transmission of the graphics to your eyes".

The above is NOT encompassing of LCD display technology, much less a
claim that your omissions imply that you think LCD's are "perfect".

Really, any talk of something being "perfect" is absurd on it's face.
Obviously, I would never imply something so ridiculous, and that makes
your above claim ridiculous.

No, Bob, unlike many of the frustrated people that I face-off with, I
put forth a clean, honest, logical argument, which of course is what
frustrates the other guy, seemingly forcing him, in his eagerness to
"beat" me, into illogical arguments and sometimes worse.

>However,
>I am very certain that I have a considerably better understanding of
>the sources - and likelihood of correction - of the imperfections
>in the two technologies than you do.

No doubt. But anyone with a pair of eyes and some common sense can
reasonably disagree with spam_eliminator's "big difference" statement.

 

[Guest]

Archived from groups: comp.sys.ibm.pc.hardware.video (More info?)

"chrisv" <chrisv@nospam.invalid> wrote in message
news:2b2re0tv9f4kbbi83s6kpkgckvrg4jv3l7@4ax.com...
> >Well, since you're clearly possessed of considerable experience in
> >this area - what do you believe is the typical luminance uniformity
> >profile of a CRT? How does it compare with the typical LCD?
>
> See below, Bob.

Y'know, I looked and I looked, and I still didn't see anything
on what YOU think is the typical luminance uniformity profile
of a CRT or an LCD, or where you got that impression. You've
found a fair number of subjective comments on various web sites
- some of which appear to have been written by people who
know what they're talking about - but still not a shred of objective
data. So where is it?


> The text I quoted from the link above is attributed to Lacie, a
> respected manufacturer of professional display monitors. Hardly just
> "anyone", Bob.

Y'know, Chris, I generally hate to play this card, but I myself
am "hardly just anyone" either. What is it you think I DO for
a living, and for that matter have done for the past 20 years?

By the way, LaCie (not "Lacie," please) is NOT a manufacturer of
monitors. They are a reseller, similar in that business model to a very
many other companies. (What they manufacture is storage peripherals,
with displays as a sideline business.) The text you quoted is still
out of date in many regards, is a great oversimplification in others,
and in some places is downright erroneous - I don't care WHO
wrote it. Believe it or not, not everything you read on the web is
correct! Just for one example - what does "maintaining color uniformity
across the screen 2 to 3 times better than an LCD" MEAN, exactly?
As measured how? In what color space? Under what ambient
conditions? With what equipment, image content, display settings,
etc.? The statement as given is meaningless at best. The site you
referenced is full of such simplistic, nonsensical, and unsupported
pronouncements. Here's another one: "CRT monitors reproduce
color temperature much more accurately than LCD monitors. This
is especially important to those working in the 5000K range when
working with pre-press and color."

This statement is nonsense on the face of it - "color temperature"
refers to the setting of the white point of the display - the color
you see when the RGB inputs are all set to their maximum value.
CRT monitors may offer multiple white point settings, but the typical
default white of a CRT is 9300K - for the simple reason that this
gives a "brighter"-appearing white, by virtue of having excess blue.
It is a noticeably bluish white, and not one which corresponds well
to any industry color standard. TV, for example, is defined around
a 6500K point (specifically, the CIE "D65" illuminant, which is not
quite on the black-body temperature curve), while many document
or pre-press apps will use the 5000 or 5500K points as these are
more of a "paper white." But the notion that there is something
inherent in LCD technology that would prevent the use of these
points is just silly - the native white point of such displays, unlike the
CRT, is set by the white point of the backlight and the characteristics
of the color filters. So they CAN be designed to match pretty much
any point you need. LCD panels for the mainstream monitor industry
have generally tried to match the mainstream CRT defaults, and so
have been built to the 9300K or 6500K points - but there is certainly
nothing stopping someone from making 5000K or 5500K LCDs,
and - surprise, surprise! - such devices aimed specifically at the
pre-press and similar markets have done just that. In fact, as of
this year, solid-state backlighting is starting to make the LCD the
display of choice for color-critical document applications, as the gamut
of such displays will far exceed that which is possible with CRT
phosphors - out to well in excess of the NTSC standard gamut, and
a very close match to the Adobe RGB space.


> Based on the current core technologies, CRT monitors are able to
> display a wider color space than LCD monitors and deliver more
> consistent brightness uniformity throughout the screen.

Out of date information. You REALLY should check the actual
numbers on current products before you believe everything you
read.

> Gee, Bob, that seems to blow your entire argument, that the CRT's
> thick glass causes it to have inferior luminance uniformity, right out
> of the water.

And gee, NOW who's being - to again use your favorite word -
"dishonest" here. I never said that the CRT glass was the cause of
inferior luminance uniformity, only that this IS a valid concern that
comes with thicker faceplates. These DO contribute to perceived
uniformity problems - but they're not the only factor. If you understood
how a CRT operates better than you apparently do, you would
realize that there is an unavoidable loss of luminance from center to
edges, with the overall non-uniformity often as bad as 60-70%
(edges of the image as compared to the center luminance), and
THAT'S just in luminance alone. Next, if you like, we can discuss
a little more sophisticated measurement of perceived uniformity, such
as looking at the delta-E* measurements.

Bob M.

 

[Guest]

Archived from groups: comp.sys.ibm.pc.hardware.video (More info?)

"chrisv" <chrisv@nospam.invalid> wrote in message
news:49bre019iaa9svp6lvu2p6prp8aj8vf3g0@4ax.com...
> I never implied any such thing, Bob. I simply think it's less than
> fair of you to nit-pick the CRT's flaws, while denying us the benefit
> of your expert critique of the LCD's imperfections "in their
> transmission of the graphics to your eyes".

Ask and ye shall receive...

If we want to compare the two technologies in terms of their
performance re commonly-used metrics of image quality, I can
certainly do that.

1. Brightness: The CRT suffers from an inherent brightness/spot
size tradeoff; in other words, it is extremely difficult to produce a
tube which has both high light output and the spot size needed for
a high resolution images. CRTs suitable for high-resolution desktop
monitor use are thus generally limited to somewhere in the 100-200
cd/m^2 range, while it is not at all uncommon for LCD products
in this same market to at least double this. Further, increasing
LCD brightness is, within reason, simply a matter of putting "more
backlight" into the thing, and does not impact resolution, etc.. So
the advantage here is to the LCD.

2. Contrast: In current displays, this is an area basically dominated
by the ability of the technology in question to produce a good black.
LCDs, being "light valves," appear at first glance to be disadvantaged
here, as their performance would depend on completely shutting off
the backlight from transmission - whereas the CRT can actually be
set so as to emit no light. However, setting up a CRT so that the
"black level" of the input signal corresponds to a tube state below
cutoff is problematic, as it makes the overall response very non-linear.
To properly set up a CRT for best overall image reproduction in all
but the simplest black/white applications (such as straight text display)
requires that the "black" level actually be set slightly above cutoff.
The bottom line is that a CRT will rarely achieve a delivered contrast
ratio much above about 100:1 in typical applications, whereas an
LCD these days will have little problem doubling or tripling this.
In typical ambient conditions, the APPEARANCE of the two displays
may be very similar in this regard, despite the quantitative advantage
of the LCD. So - slight advantage for the LCD.

3. Geometry, Linearity, Focus: LCDs are fixed-format devices (they
actually have discrete, addressable physical pixels), whereas the CRT
is not. Therefore, the geometry, linearity, and "focus" of an LCD is
unchangeable and essentially "perfect." (Pixel size and location
errors are on the order of microns, and invisible to the viewer.)
Clearly, this is not and cannot be the case with the CRT. Advantage:
LCD.

4. Viewing angle: State-of-the-art LCDs offer viewing angles,
generally defined as the point where the contrast decreases below a
certain threshold (10:1 is common) of 160 degrees or better.
Further increases are possible, and will appear in the market soon -
however, clearly the LCD will never quite match the viewing angle
of a display which actually produces light at the screen surface, such
as the CRT. (Wait for OLEDs for the real viewing-angle champion!).
On the other hand, not many people use their displays at extreme
angle anyway. So a slight advantage here to the CRT.

5. Resolution: This is among the most misunderstood areas in this
discussion, and also one which does not lend itself to an apples-to-apples
comparison between the two technologies. It gets back to the LCD
being a fixed-format type, where the CRT is not. As a fixed-format
display, the "resolution" (in the proper sense of the term - how much
detail can be visibly resolved per unit distance) of the LCD is also
fixed; you get a certain number of pixels per inch, and that's that.
A CRT, which does NOT have physical pixels at all (the color
screen triads are NOT "pixels") does not have such a limitation;
its ability to resolve detail is limited primarily by the spot size, and,
in color types, by the pitch of the shadow mask and phosphor
screen (not QUITE the same thing). That the "look" of the two is
different, regardless of "resolution," is also indisputable, due to the
"soft" edges of the CRT spot vs. the sharply-defined features of the
CRT screen - but that is, to a very large degree, a matter of personal
taste. In terms of the absolute limits in resolution for the two
technologies
- color LCDs have already been demonstrated with resolution well in
excess of 300 pixels per inch, whereas color CRTs are definitely
struggling to get much above 150 ppi. (Very specialized monochrome
tubes have been made claiming about 300 ppi, but even those were
unable to properly resolve single-pixel details at that level.) Due to
the complexity of this issue, I won't state a clear winner here - but
there's the information as I have it. Draw your own conclusions.

3. Color convergence/purity: Short and sweet - the CRT is subject
to problems in these areas (and these are MAJOR areas of concern
in terms of the technology's susceptibility to external fields, see
below), and the LCD simply can't have these problems due to a
completely different operating principle. Advantage: LCD.

4. Color accuracy/gamut: Until fairly recently, the CRT was the
clear winner here, at least in terms of "accuracy" (which primarily
had to do with the overall luminance response curve of the technology).
However, this was not due to an inherent theoretical limit in the
LCD, and as of this year, LCD displays are being produced which
equal the CRT in response accuracy and greatly exceed it in
gamut. Advantage: For the moment, in terms of what's out in the
installed base and mainstream market, the CRT - but this won't
last.

5. Brightness/color uniformity: The CRT suffers from an inherent
loss of luminance as one moves from the center of the screen
to the edges (there's also a loss of "focus," which can be partially
compensated for), and is also more subject to convergence/purity
issues in the extremes as well. (All for the same reason - the
changing geometry of the electron beam with respect to the screen.)
In the LCD, color and luminance uniformity issues arise from
non-uniformities in the light profile coming through the backlight &
diffuser assembly, possible non-uniformities in the drivers, and to
some slight degree non-uniformities in the color filter response.
However, these do NOT inherently have the "good in the center,
worse at the edges" characteristic of the CRT, and so are often
less visually objectionable - and again, all can be addressed in the
design (if the customer is willing to pay the price), as opposed to
the CRT being inherently constrained by its basic operation.
Advantage: Unclear - it's easier to make this "look good" in a
relatively inexpensive CRT vs. a similarly bargain-basement
LCD - but the best performance at ANY cost will likely be from a
properly designed LCD product.

There are a number of other areas of performance not directly
related to image quality which generally factor into the buying
decision. These include:

Cost: Clear advantage for the CRT.

Weight: Clear advantage for the LCD.

Physical size (depth): Ditto

Power: Clear advantage for the LCD.

Susceptibility to external fields: Clear advantage for the LCD
(and a MAJOR stumbling block for the CRT in many applications;
there are numerous cases where the CRT simply is NOT an
option for this reason alone).

Emissions: In terms of straight RFI, both CAN be noisy -
although I have generally found LCD designs easier to
"clean up" (your mileage may vary). However, in terms of
X-radiation and low-frequency electric and magnetic fields,
the LCD is the obvious winner (esp. in areas where these
are controlled by government regulation or the equivalent.)

The bottom line is that the LCD already has the advantage in
the majority of areas which are of concern to most customers, and
is the only possible choice in some. This has, over the past few
years, resulted in a split in the CRT market - the CRT is basically
vanishing as a "mainstream" computer display, and is being
pushed to the high and low ends of the market. At the high end,
it is being chosen (in specialized forms) for very high-resolution
applications where the LCD types are not yet cost-effective, and
(for a while yet) in color-critical applications (but again, these aren't
representative of mainstream CRT performance). And the
CRT remains the display of choice for very, very cost-conscious
markets (think China and India, for example), since it is still the
absolute cheapest display possible for PCs, and will be for some
time. Over the next few years, the LCD will begin to eat away at
those remaining high-end applications as well, leaving the CRT
as viable only in those markets where its cost advantages
outweigh all other considerations.

Bob M.

 

[Guest]

Archived from groups: comp.sys.ibm.pc.hardware.video (More info?)

"Bob Myers" <nospamplease@address.invalid> wrote in message
news:J9kHc.5579$3B3.4419@news.cpqcorp.net...
> Ask and ye shall receive...

Ooops - forgot one, and I really shouldn't've, since it's been
one of the major areas of concern re the LCD over the past
N years -

Response time: Again, the two technologies have very, very
different operating modes. In the case of the CRT, the image
is "drawn" essentially a pixel at a time (if the thing HAD pixels,
that is) by a scanned electron beam - whereas in the LCD, the
pixels are set to their desired light transmittance state a row at
a time, and pretty much left there until the next frame time.
Also, over most of their history, LCD materials have had
relatively slow response times compared to CRT phosphors
(at least low tens of milliseconds for the LCD, compared to
nanosecond "on" times for the phosphor followed by a few
hundred microseconds' persistence in the case of phosphors,
at least for most color CRT blends). All of this comes under
the broader heading of what a display technologist refers to as
the temporal response of the display.

Since the CRT phosphor does NOT stay illuminated for long,
it has to be driven very, very hard - the actual illuminated spot
(when the beam is striking it) is a good deal brighter than you
think, and then the light output falls off dramatically. Were it
not for this high initial spike making for a reasonable average
(over time) light output, and the persistence of vision, we would
see CRT images as just a flying spot of bright light, coupled
with a very low-contrast image. (Which some cameras DO
see when pointed at a CRT.) On the other hand, the LCD has
relatively flat light output over time, but until recently suffered from
much longer on/off transients.

The bottom line here is that the CRT, until recently, was a
better match to fast-moving imagery (frame rates on the
order of tens per second) vs. the LCD, which was best
suited to the display of relatively static images (which make
up a good deal of typical PC usage). The flip side of this
is that the CRT suffers from a severe flicker problem, which
gets worse with increased brightness (another reason that
PC CRT displays with the brightness of the current LCD crop
were never really a good idea) - whereas the LCD provides
a much more stable image, with essentially no flicker at all.

Over the last couple of years, however, LCD response times
(which confusingly are often given as the SUM of the on and
off times!) have fallen below the frame or field time of video
needed for good motion rendition (somewhere in the 20-60
FPS range), and are continuing to decline. Response times
in the 12-16 ms range are common now, and faster displays
(down to a few ms response time) have already been
demonstrated. At this point, the LCD is a very acceptable
display for video-rate imagery, as evidenced (and driven, for
that matter) by its entry into the television market).

Bottom line - advantage: until recently, the CRT, unless you
couldn't live with the flicker. From here on out, this
advantage will be basically gone in all but extremely
specialized applications.

To sum up, then, we have the following scorecard:

Brightness: LCD
Contrast: LCD
Geom., etc.: LCD
View. angl. CRT (slight)
Resolution: LCD (slight in mainstream apps)
Color conv./purity: LCD
Color acc./gamut: CRT at present; LCD within a year or two
Lum./color uniformity: Was CRT, now LCD in best case; essentially even now
for mainstream designs
Cost: CRT (and will be the winner here for a long time)
Weight: LCD
Size (depth): LCD
Power: LCD
Field susc.: LCD
Emissions: LCD
Resp. time: Even in majority of current applications.


Yeah, it's pretty hard to see why the market is switching over to the
LCD....🙂


Bob M.

 

[Guest]

Archived from groups: comp.sys.ibm.pc.hardware.video (More info?)

For what it's worth, the reason why I am working on TFT/LCD is the DVI-I /
DVI-D connector, it seems next to impossible to get a decent 1600x1200
resolution from videocards unless you try half a dozen, and I upgrade now
and then, and every single instance in the past 4-5 years the 1024x768 is
best these devices can do unless like "smoothed" signal, ie. blurry / poorly
defined image in higher resolutions. I know there are people with good
experiences with DB15 and CRT's, good for you! I had such experience myself
with Matrox Millennum and 486/DX4-100 back in 1990's.. since then power of
the graphics cards have increased and analogic image quality decreased..
feel free to disagree because I know you will be right, that is not in
question.

13w -> bnc gives a sharp, well defined image at 1600x1200 but I don't have
IA32 compatible PC with 13w->bnc with good dac's like those found in SGI or
SUN workstations do the job great. Since the work is on PC (Linux and
Windows) for me, this means DVI is way to go to get good, sharp image
inexpensively. It's just the price/performance that wins in that regard in
DVI's. I could buy $2000+ "workstation class" display card, but it would
cost more than rest of the workstation (A64 3000+, 1GB, ..)

Summary: good image quality at affordable price in resolution I like to work
with (more would be even better, but 1600x1200 has enough real-estate to get
the job done w/o stacking workspaces). I'm currently using dual 1600x1200...

It is inevidable that some smart ass steps in and says: "I got X brand CRT
and Y brand gfx card and 1600x1200 is sharp as razor, you're an idiot
because you don't know this" -- that sort of stuff happens novadays in the
Usenet ALL THE TIME. The point is that with DVI, I got one thing less to
worry about when choosing a graphics solution for next upgrade cycle. I know
the image will be crisp and sharp, and concentrate on features, performance,
robustness and price of the solution. It's not like there is abundance of
choise if we are realistic. ATI and NV, that's pretty much it. Matrox could
fix their drivers, they corrupt simple GDI rendering so that's not a very
good thing, they're off the game, zap.

3DLabs might be worth consideration, and Quadro's from NV but I don't really
need the features I would be paying for. DVI = inexpensive, good quality. It
seems to be somehow hard, difficult to make decent analogic high-resolution
high-refresh to work with decent image quality. If it is so damn hard, DVI
with digital connection is inexpensive and efficient solution. Why they
don't make CRT's with DVI connectors? Then the "last mile" in image quality
would be in the monitor and you could judge the image quality purely on the
monitor/display device, not:

- graphics card (dac mainly)
- cable (poor quality cables exist, ya' know?)
- display device (=monitor)

Too many things that can go wrong. With DVI it's only the "last mile" that
can go wrong (explanation: digital means something either works or doesn't
work at all-- ok shitty joke, flame me..)

I'm so worried about color being "the wrong shade", when I am not doing
printing or other color sensitive work... if red is not blue or green, I'm
fine. <- flame me for that aswell, very stupid thing to say.. now list
(statistically 2.89 reasons are mentioned to defeat silly argument in
Usenet) 3 reasons why I am a retard for not having brain to understand that
poor color definition is a Bad Thing for me.

Hehheh.

 

[Guest]

Archived from groups: comp.sys.ibm.pc.hardware.video (More info?)

"assaarpa" <redterminator@fap.net> wrote in message
news:ccu0ll$stn$1@phys-news1.kolumbus.fi...
> For what it's worth, the reason why I am working on TFT/LCD is the DVI-I /
> DVI-D connector, it seems next to impossible to get a decent 1600x1200
> resolution from videocards unless you try half a dozen, and I upgrade now
> and then, and every single instance in the past 4-5 years the 1024x768 is
> best these devices can do unless like "smoothed" signal, ie. blurry /
poorly
> defined image in higher resolutions. I know there are people with good
> experiences with DB15 and CRT's, good for you!

No, you point out another important distinction between the
two technologies. This again gets back to the fact that the LCD
is a fixed-format technology (which doesn't mean that it is somehow
"inherently digital" - it's not, and in fact LCDs are analog devices
at the pixel level), which requires very precise timing, at the pixel
clock level, in order to properly sample the video data stream
(whether it's sent in analog OR digital form). The classic "VGA"
analog interface simply doesn't provide this level of timing information,
whereas the "digital" interfaces DO (they have to - as you noted,
they wouldn't work at all without it). There IS a VESA standard
in the works which would address this shortcoming of the VGA
interface, but it's uncertain if it will be adopted by the industry or
if we'll all just wind up making the painful transition to digital.


> Why they
> don't make CRT's with DVI connectors?

Well, quite simply because there would be no advantage to it.
"Digital" CRT monitor designs have been proposed, but the CRT
by its nature simply doesn't need that level of timing accuracy,
since it could not possibly care less about where the "pixels" are
in the analog video stream. And, believe it or not, it really IS
the timing that is responsible for most of the problems you see
with analog interfaces on LCDs, not the "bandwidth" or other
possible image quality factors. (On the other hand, God knows
there is certainly no end of the truly lousy VGA cable
assemblies that are out there, and they do NOT do you any
good in this regard at all.)


Bob M.

 

[Guest]

Archived from groups: comp.sys.ibm.pc.hardware.video (More info?)

> workstations are solid engineering unlike most consumer graphics cards..
and
> you notice it on the pricetag!)

And that, my friends, is the crux of the matter at hand. DVI is inexpensive
way to eliminate the errors from the graphics card and cabling. Only the
display device now have to accurately present the signal. TFT does this
pretty well, due to their nature. CRT's would, agreed, get less benefit from
this arrangement.. but.. consider this:

If a sgi->13w->bnc cable can do 1600x1200 without breaking a sweat and
geforce6800u->db15->db15 has trouble coming up with adequate 1280x960, using
_same_ display (1200nf) on both cases, what's wrong with this picture?

I tell you what's wrong: it seems to be prohibitively (literally) expensive
or difficult to make a good digital-to-analogic converter, this is why $8000
workstation does the job properly and $650 graphics card doesn't. This is
logical. Now look at DVI: inexpensive way to do what with analogic signaling
takes $8K workstation and by defnition is not something $650 graphics card
can be expected by default to do. Not for years.

If you claim that it is just as infeasible to put a good digital-to-analogic
converter into a CRT than it is to put one in graphics card, what makes you
think CRT then can handle the analogic signal any better than the graphics
card, considering there can be a shoddy cable between the two to add into
the confusion? Wouldn't it make much more sense to avoid unnecessary signal
loss as long as possible, especially when tranfering the signal digitally is
now cheap?

Let me guess. It would require a shift which market is not 'prepared' for,
in otherwords, it is uncertain that market would follow. Why you think Intel
is still selling products based on x86 architechture? Because it is good
enough and the market is used to bying that, in short, "that's the way
things are" -- it is VERY difficult to change "the way things are".. human
mind resist change, they really do.

"Technological progress", is not change, it is just rehashing of the old
ideas for most part. Graphics cards are still based on the same principles
they were 10 and 20 years ago. This is not surprising, "so why don't you
think of something better?", indeed that would be interesting question. The
current graphics architechture 3D cards implement is fundamentally flawed
when it comes to handling of translucent primitive, for example. Raytracing
for instance solves this elegantly, however, it is architechture which is
not as well suited for on-chip implementation as it involves hierarchical
database with *random*, arbitrary dependent queries! GPU's are executing the
same instructions for all pixels, this is very deterministic and reasonably
easy to do in silicon and in parallel.

Remember when NV last time tried curved surfaces? Way before GeForce's or
TNT's? What did the market and developers say? They said nothing or at best
they said, quote, "DUH?" -- me included. It didn't even fundamentally change
or break the existing pipeline! This was just example of the ways human
mind, or atleast communities have natural resistance to new ideas (both good
and bad, what is needed is luck and good presentation

DVI + CRT? DUH? What a stupid idea, it been proven that it doesn't offer
anything over DB15. <- ring a bell? I say it would when done properly. But
you see the market doesn't care so the vendors don't care. The market has
TFT's and they sell well. Market wants to re-sell the same thing for us over
and over again so that profit can be made and people can be employed and
that economy stays in good shape and everyone will have a good time.

How many VCR's did you guys buy? What's wrong with the previous ones? They
did break down didn't they? Or some new fancy feature you just had to have
was introduced. You know what I mean. How many of you still using VCR? How
many even have one still in the house, or praytell, plugged in? How many
find himself buying the movies again in DVD they had as VHS already, because
of better image, better sound (AC3 baby! =), extras on the DVD, whatever?
How many thinks he won't be buying some of these movies again when HIGH
DEFITION DVD (or whatever the next Big Thing in Video will be)? Oh, that
many? I see a lot of people good at deceiving themselves.

The point is that there is no incentive to introduce a line of products with
CRT+DVI, because there isn't market for it. Few exceptions like me don't
make a market, unless the prices are orbital. Now you think I would pay
orbital price for something I can get for peanuts? Not very many intelligent
people would! And _that_ is what I think about this, thanks for trying to
"prove" me wrong, but you see this is just opinion and opinions don't
necessarily be 'correct' to be valid.. I ask you guys one question, please
answer, I been supplying a LOT of extra information to back up WHAT
PRECISELY my opinion is BASED on. Now, this extra information is taken apart
with clinical accuracy, which I have no objection to. However, it is ironic
that the original point is completely and utterly avoided and/or not even
understood (I would be biased towards 'ignored', though).

 

[Guest]

Archived from groups: comp.sys.ibm.pc.hardware.video (More info?)

"Bob Myers" <nospamplease@address.invalid> wrote:

>Ask and ye shall receive...

Well, actually I asked you to nit-pick the LCD the way you did the
CRT, but instead I got a summary devoid of extreme nit-picking, which
I suppose suits me fine, anyway.

>(snip summary)

I note that there was nothing in the summary regarding disadvantages
of the CRT due to it's thick glass front. Apparently the effects are
so small as to not be worthy of your reasonably complete summary.
This, of course, bolsters my case that the effects of the thick glass
do indeed NOT cause a "big difference" relative to the LCD.

In other words, I've been right all along, and you knew it, Bob.

 

[Guest]

Archived from groups: comp.sys.ibm.pc.hardware.video (More info?)

"assaarpa" <redterminator@fap.net> wrote in message
news:ccudag$hmb$1@phys-news1.kolumbus.fi...
> > > Why they
> > > don't make CRT's with DVI connectors?
> >
> > Well, quite simply because there would be no advantage to it.
>
> Why not? I get crystal sharp 1600x1200 with 13w -> bnc with 22" SyncMaster
> 1200NF, same monitor, dozen difrent DB15 -> BNC or DB15->DB15 cables with
> over $400 PC graphics cards: image is blurry, out-of-focus, etc. It seems
> that the analogic is great if the analogic sources are great.

Or if the analog cabling is "great" - a major problem is that a
LOT of VGA (HD15) cabling isn't, and in fact isn't even really
fit to carry baseband TV video. Noting that you can get "crystal
sharp" images with a 13W3 or BNC connectors should be
taken as an indication that it's not necessarily the analog signal
sources or the analog interface standard itself that is the limiting
factor here.

Again, a VERY major factor in the perceived benefit of the
"digital" interfaces with LCDs and similar fixed-format displays
is the presence of much better pixel-level timing information, which
such displays absolutely require for optimum performance. The
CRT has no such requirement, and so does not clearly benefit
from these interfaces. (It makes little difference that the video
information itself is transmitted in "analog" or "digital" form in
either display technology; 8-bit accuracy, which is all that's provided
on the current digital standards, isn't that hard to achieve in
analog video. And note that both the CRT and LCD are
fundamentally analog-controlled devices - it's just that in the case
of the LCD, a digital-to-analog conversion typically takes place
within the display panel itself.)

This is not to say that there aren't poor analog video sources -
there clearly are - however, it really doesn't take all that much
to clean these up, IF the designer knows what he or she is
doing.


Bob M.

 

[Guest]

Archived from groups: comp.sys.ibm.pc.hardware.video (More info?)

"assaarpa" <redterminator@fap.net> wrote in message
news:ccuvdf$66d$1@phys-news1.kolumbus.fi...


> I tell you what's wrong: it seems to be prohibitively (literally)
expensive
> or difficult to make a good digital-to-analogic converter, this is why
$8000
> workstation does the job properly and $650 graphics card doesn't. This is
> logical. Now look at DVI: inexpensive way to do what with analogic
signaling
> takes $8K workstation and by defnition is not something $650 graphics card
> can be expected by default to do. Not for years.

The term, first of all, is usually "analog" - I don't believe I've
ever encountered the construct "analogic" before.

Second - again, for the CRT (or for that matter, the LCD), the
digital information must be turned into an analog voltage at some
point. The difference between the two technologies in this area is
that for the CRT, this conversion MUST occur at the pixel rate
(whereas in the LCD, it occurs at essentially the line rate). It is
clearly not less expensive to include the D/A function in the CRT
than in the graphics card - and in fact, it will generally be MORE
expensive, all else being equal, since in the monitor this function
would have to be done via a discrete component rather than being
integrated into existing silicon. Further, the present digital interfaces
have very limited capacity compared to what can be achieved with
even moderately-good analog video implementations; single-link
DVI tops out a 24 bits per pixel at a 165 MHz pixel rate - enough
for 1600 x 1200 at 60 Hz, but not much beyond that. It's not
hard at all to get well over 200 MHz pixel rates via analog video
(to 2048 x 1536 and beyond, and further better-than-8-bit/color
grayscale performance is achievable with careful design and
implementation. It's generally better, then, in the CRT case to permit the
CRT to be as flexible as possible, and place the burden of decent
output design on the graphics card designer. But those makers will
deliver only what the market requests, you know.

As to the "$8000 workstation vs. $650 graphics card" example -
where do you think the workstation manufacturers get their graphics?

Bob M.

 

[Guest]

Archived from groups: comp.sys.ibm.pc.hardware.video (More info?)

> No, I'm not "being vague". Both VNC and Terminal Services let you use a
> Unix box as a display for a Windows PC.

Sounds quite a bit more expensive solution for general user to buy SGI
workstation than just getting on with the program and eliminating shitty
DAC's in PC consumer gfx cards completely out of the picture with DVI and
whatever you can plug into the other end. Inexpensive and simple solution.
Even simpler would be that the DAC's and cables were good to begin with, but
practise has shown that is not the case in the real world we live in: if you
want high quality you pay. DVI means high quality is affordable, thanks to
the technology that is being used to solve the problem.

Since graphics cards now come with DVI as pretty much standard, why would it
be 'unreasonable' to get rid of the DB15 completely, especially considering
that TFT's are slowly getting rid of them while we speak? If I connect a PC
to a Plasma Display, which input port I choose, you think, the DVI or DB15?
Definitely not the component or scart, s-video or good riddance, composite.
HDMI is also beginning to show it's head in home electronics, some DLP
projection screens have it and some plasmas have it. In a few years from
now, who knows, we might begin to see it in PC graphics cards as option
aswell.

Your position of defending DB15 is a bit silly, I understand it from the
backward compatibility point of view but let's face it the market is moving
away from it slowly but steadily. DVI-I still includes DVI-A, so DVI-I
compliant devices are still able to give out DB15 for years to come. This
means extra cost, so I don't think the vendors will put too much effort into
making it crips and the primary interface will be DVI-D anyway for most
users, especially in the coming years.

This is why I am "whining" about it.. ( I myself prefer to call it
"observing a trend" ).. why we are moving into this direction is cost
efficiency and high quality. The trouble with DVI is the BANDWIDTH. For
3800x2400 you need two DVI channels, if you want something like that you are
going to need two separate, or one non-standard cable and connector. But for
mass market uses, DVI is adequate. In upper limits of the bandwidth it wins
the analogic signaling hands down and does only be precise to the required
tolerances, after that it's fair game to malfunction: that's what digital is
all about, tresholds and tolerances.

Analogic display device getting digital signal would indeed introduce
inaccuracy of it's own: you cannot change voltage in instant from one to
other, the voltage will go through all values between the two, how sharply
and how the signal handles at ends is up to the quality of the components.
Making that precise square waveform is, expensive, "or so it seems", if I
may quote my earlier posts.

Next thing we know, I will get another angry response about "stop trolling"
and "stop whining", ohwell..


> > Is this so hard to accept? Give me more than one good reason to switch
> > back to CRT with PC.
>
> Why would I care if you "switch back to CRT with PC"? You're the one
> whining about being unable to do so.

I am not whining, pay attention: I am rejoicing that it is inexpensive and
easy to get 1600x1200 crystal sharp image with PC gfx cards. You seem to
have problem that there is a solution that works out for me, or why so
critical?

It's LOTTO to get a decent 1600x1200 with CRT and anything that is
associated with DB15. That's a beautiful point: no need for VNC, Terminal
Services, etc.. buy cheap off-the-shelf hardware and get on with the
program. Simple and cheap. The way it should be.

 

[Guest]

Archived from groups: comp.sys.ibm.pc.hardware.video (More info?)

assaarpa wrote:

>> No, I'm not "being vague". Both VNC and Terminal Services let you use a
>> Unix box as a display for a Windows PC.
>
> Sounds quite a bit more expensive solution for general user

I wasn't talking about the "general user", I was talking about _you_.

> to buy SGI
> workstation than just getting on with the program and eliminating shitty
> DAC's in PC consumer gfx cards completely out of the picture with DVI and
> whatever you can plug into the other end. Inexpensive and simple solution.

Again you fail to demonstrate that what can't be done in an "inexpensive"
video board can be done in an "inexpensive" monitor.

> Even simpler would be that the DAC's and cables were good to begin with,
> but practise has shown that is not the case in the real world we live in:
> if you want high quality you pay. DVI means high quality is affordable,
> thanks to the technology that is being used to solve the problem.

What "technology" do you believe "is being used to solve the problem"? DVI
with an LCD display works aroudn the problem, it doesn't solve it.

> Since graphics cards now come with DVI as pretty much standard, why would
> it be 'unreasonable' to get rid of the DB15 completely,

How does changing the connector improve the situation?

> especially
> considering that TFT's are slowly getting rid of them while we speak?

If you believe that you might want to check the stock at CompUSA and the
like.

> If I
> connect a PC to a Plasma Display, which input port I choose, you think,
> the DVI or DB15?

What of it? A plasma display can process a digital signal directly, with
the only analog conversion being the intensity at each pixel.

> Definitely not the component or scart, s-video or good
> riddance, composite.

So? The only data-grade monitors I know of whcih have coponent, s-video,
composite, or scart inputs are projectors.

> HDMI is also beginning to show it's head in home
> electronics, some DLP projection screens have it and some plasmas have it.
> In a few years from now, who knows, we might begin to see it in PC
> graphics cards as option aswell.

Why would one need an audio interface on a PC graphics card?

> Your position of defending DB15 is a bit silly

What position is that? I am not "defending DB-15". I am stating the _fact_
that putting the DAC in the CRT and putting a DVI interface on it has not,
in the real world, proven to be a satisfactory solution. If you believe
that it is possible to implement such a monitor satisfactorily and that
there is a market for such monitors, then I suggest you start taking
advantage of your superior insight into the market and form a company to
produce them, thus making yourself fabulously wealthy.

>, I understand it from the
> backward compatibility point of view but let's face it the market is
> moving away from it slowly but steadily. DVI-I still includes DVI-A, so
> DVI-I compliant devices are still able to give out DB15 for years to come.
> This means extra cost, so I don't think the vendors will put too much
> effort into making it crips

"crips"? What does a street gang have to do with anything?

> and the primary interface will be DVI-D anyway
> for most users, especially in the coming years.

By which time LCDs will be so highly perfected and so cheap that nobody will
want a CRT anyway.

> This is why I am "whining" about it.. ( I myself prefer to call it
> "observing a trend" )..

Observing at great length and in ad nauseum detail. Fine, there is a trend
from analog interfaces on non-CRT monitors to digital interfaces. So what?

When someone tells you that a few years ago there was a similar trend with
CRT monitors with the results being universally dismal and with a resulting
reversal in that trend you accuse them of "defending DB-15".

If you don't like the message there is little point in shooting the
messenger.

> why we are moving into this direction is cost
> efficiency and high quality.

It's only "cost efficient" if you can eliminate digital-to-analog
conversion. With a CRT, unless you redesign the thing from the ground up
along different principles from those that have become established in the
industry over the past 70 years or so, you cannot eliminate that
conversion.

> The trouble with DVI is the BANDWIDTH. For
> 3800x2400 you need two DVI channels, if you want something like that you
> are going to need two separate, or one non-standard cable and connector.

That's also the trouble with analog. An analog cable that can carry
3800x2400 without ghosting is not cheap. What of it?

> But for mass market uses, DVI is adequate.

Who has claimed otherwise?

> In upper limits of the
> bandwidth it wins the analogic signaling hands down and does only be
> precise to the required tolerances, after that it's fair game to
> malfunction: that's what digital is all about, tresholds and tolerances.

Are you Amish?

> Analogic display device getting digital signal would indeed introduce
> inaccuracy of it's own: you cannot change voltage in instant from one to
> other, the voltage will go through all values between the two, how sharply
> and how the signal handles at ends is up to the quality of the components.
> Making that precise square waveform is, expensive, "or so it seems", if I
> may quote my earlier posts.

So what?

> Next thing we know, I will get another angry response about "stop
> trolling" and "stop whining", ohwell..

Since you won't accept any suggestions on how to get the result you want
with the hardware you have, you clearly have no interest in doing anything
but complaining idly. It may come as a shock to you, but nobody who has
the power to implement the changes you want in the monitor industry is
reading this newsgroup with bated breath waiting to gain the benefits of
your superior insights.

>> > Is this so hard to accept? Give me more than one good reason to switch
>> > back to CRT with PC.
>>
>> Why would I care if you "switch back to CRT with PC"? You're the one
>> whining about being unable to do so.
>
> I am not whining, pay attention: I am rejoicing that it is inexpensive and
> easy to get 1600x1200 crystal sharp image with PC gfx cards. You seem to
> have problem that there is a solution that works out for me, or why so
> critical?

If there is a solution that works for you then why are you creating these
vast posts in which you express your dissatisfaction with the hardware that
you can buy?

> It's LOTTO to get a decent 1600x1200 with CRT and anything that is
> associated with DB15.

Hardly "lotto". Just not cheap if you want SGI quality. Most people don't
want SGI quality, and to tell you the truth if I liked to read tiny little
type I would be quite happy with the 1600x1200 that I get out of my $50 ATI
board and CRT monitor.

> That's a beautiful point: no need for VNC, Terminal
> Services, etc.. buy cheap off-the-shelf hardware and get on with the
> program. Simple and cheap. The way it should be.

So figure out a way to make "cheap off the shelf hardware" that does what
you want and get on with it.



--
--John
Reply to jclarke at ae tee tee global dot net
(was jclarke at eye bee em dot net)

 

[Guest]

Archived from groups: comp.sys.ibm.pc.hardware.video (More info?)

> program. Simple and cheap. The way it should be.

Oh, and the point: if CRT vendors *1) are interested in my money, they
should think about how I could transmit a crystal clear image from the
latest ATI and NV graphics products *2) to the display 1600x1200 minimum.
Before that is adequately solved, don't bother.

Then a comment about VNC. How you propose I transfer the framebuffer
generated by GPU to the SGI workstation? You DO know that reading from the
GPU local memory to system memory is very, very slow, right? Then
transfering the data over LAN, you don't think there would be any latency (I
have Gigabit switch and ethernet adapter in my Windows box, but the Octane2
only has FastEthernet). All in all, I wouldn't expect performance in par
with working locally. I could, ofcourse, jump back and forth between the two
workstations from one room into the other.

But why bother, when I get crisp output with DVI-D? I still at loss what you
are suggesting me to do when I have everything I need sorted out neatly and
cheaply! Doesn't make any sense why you are having such a difficult time
with this?


1) As-if we could say there are dedicated "CRT" and "TFT" vendors, maybe
there are one or two, or maybe you could call vendors who dropped CRT's out
of their product lineup "TFT" vendors -- I wouldn't. The "CRT vendor" in the
context of above sentence merely means any vendor who is selling CRT's and
want to compete for my money against DVI based solution such as TFT display.
I don't claim to be relevant individual, but I wouldn't say that my point of
view would be unique, that would be flattering myself.

2) As we all know most ATI and NV based graphics products in < $1000
price-range have inferior image quality for analogic output. I don't say I
couldn't go for other vendors but looking at their offerings they are either
very expensive or don't have the other criteria met these two vendors have,
too bad about their DAC's, though.

 

[Guest]

Archived from groups: comp.sys.ibm.pc.hardware.video (More info?)

"assaarpa" <redterminator@fap.net> wrote in message
news:cd11l6$d16$1@phys-news1.kolumbus.fi...
> Oh, and the point: if CRT vendors *1) are interested in my money, they
> should think about how I could transmit a crystal clear image from the
> latest ATI and NV graphics products *2) to the display 1600x1200 minimum.
> Before that is adequately solved, don't bother.

However, this brings up another point which was addressed much
earlier in this thread - there is VERY little new development going
on these days in CRT displays, since they are rapidly being displaced
in all segments of the market (with the exception of the very low-cost
end) by other technologies. And clearly, those low-end displays are
NOT going to be an area where you see the introduction of a lot
of gee-whiz, expensive new features.

Bob M.

 

[Guest]

Archived from groups: comp.sys.ibm.pc.hardware.video (More info?)

Please clarify something.

Is everything in your life is perfect or is there something you want to
change?

If everything in your life is perfect then what are you on about?

If there is something you want to change please state clearly, and
succinctly, in one paragraph of less than five lines what it is, so that
someone of my meager intellect can comprehend the issue.

assaarpa wrote:

>> program. Simple and cheap. The way it should be.
>
> Oh, and the point: if CRT vendors *1) are interested in my money, they
> should think about how I could transmit a crystal clear image from the
> latest ATI and NV graphics products *2) to the display 1600x1200 minimum.
> Before that is adequately solved, don't bother.

I suspect that the CRT vendors couldn't care less if you personally bought
from them.

> Then a comment about VNC. How you propose I transfer the framebuffer
> generated by GPU to the SGI workstation? You DO know that reading from the
> GPU local memory to system memory is very, very slow, right? Then
> transfering the data over LAN, you don't think there would be any latency
> (I have Gigabit switch and ethernet adapter in my Windows box, but the
> Octane2 only has FastEthernet). All in all, I wouldn't expect performance
> in par with working locally. I could, ofcourse, jump back and forth
> between the two workstations from one room into the other.

Have you tried it?

> But why bother, when I get crisp output with DVI-D? I still at loss what
> you are suggesting me to do when I have everything I need sorted out
> neatly and cheaply! Doesn't make any sense why you are having such a
> difficult time with this?

If you don't have a problem then why do you keep going on about your
difficulties interfacing your CRT to a PC?

> 1) As-if we could say there are dedicated "CRT" and "TFT" vendors, maybe
> there are one or two, or maybe you could call vendors who dropped CRT's
> out of their product lineup "TFT" vendors -- I wouldn't. The "CRT vendor"
> in the context of above sentence merely means any vendor who is selling
> CRT's and want to compete for my money against DVI based solution such as
> TFT display.

What leads you believe that anybody but you wants to do this?

> I don't claim to be relevant individual, but I wouldn't say
> that my point of view would be unique, that would be flattering myself.

Are there enough like you to pay for product development?

> 2) As we all know most ATI and NV based graphics products in < $1000
> price-range have inferior image quality for analogic output. I don't say I
> couldn't go for other vendors but looking at their offerings they are
> either very expensive or don't have the other criteria met these two
> vendors have, too bad about their DAC's, though.

If it is your intention to become a market analyst for the display industry,
don't quit your day job.

--
--John
Reply to jclarke at ae tee tee global dot net
(was jclarke at eye bee em dot net)

 

[Guest]

Archived from groups: comp.sys.ibm.pc.hardware.video (More info?)

> fundamentally analog-controlled devices - it's just that in the case
> of the LCD, a digital-to-analog conversion typically takes place
> within the display panel itself.)

J Clarke says that is infeasible to have the da converter in the display,
comments on that?


> This is not to say that there aren't poor analog video sources -
> there clearly are - however, it really doesn't take all that much
> to clean these up, IF the designer knows what he or she is
> doing.

A friend of mine "cleaned" his GeForce4 by removing resistors from his
card's BCP few years back, which lead me to look into the articles on the
topic back then. The generic idea in the article was that the resistors were
there to dampen the signal so that the cabling wouldn't emit too much
interference, RFC regulation related anyway, the article might still be
online somewhere.

I'm not disputing what you say, you are easy to agree with.

 

[Guest]

Archived from groups: comp.sys.ibm.pc.hardware.video (More info?)

> As to the "$8000 workstation vs. $650 graphics card" example -

Octane2 is using SGI's own design, V6,V8,V10 and V12.


> where do you think the workstation manufacturers get their graphics?

I think they get the graphics from 3DLabs, nVidia, ATI, E&S, IBM, .. one of
the differences to consumer-priced products would be the quality of
engineering for hardware and software (drivers), and diffrent feature-set
enabled in the chip (either different mask or externally disabled
functionality either in the BCP, firmware or driver) and little details like
that.

Also if you want to run VERY high resolution display you're better prepared
to spend some cash. Help, mercy! 1600x1200 is practical for consumers now,
that's why I am happy not complaining like JC claims.

 

[Guest]

Archived from groups: comp.sys.ibm.pc.hardware.video (More info?)

"chrisv" <chrisv@nospam.invalid> wrote in message
news:koh8f0pg7tlbv8kj35npn8rlqllim7slnf@4ax.com...
> Well, actually I asked you to nit-pick the LCD the way you did the
> CRT, but instead I got a summary devoid of extreme nit-picking, which
> I suppose suits me fine, anyway.

If you can think of some relevant "extreme nit-picking" to
add, please be my guest. If, on the other hand, you think
anyone here is in any way required to write whatever you
request, think again.

> I note that there was nothing in the summary regarding disadvantages
> of the CRT due to it's thick glass front. Apparently the effects are
> so small as to not be worthy of your reasonably complete summary.

Obviously, getting into every detail of both technologies would
require a book. I've done a book before, and it's not all that
much fun - and it's certainly not something I'm going to be doing
for the sheer hell of it here.

> This, of course, bolsters my case that the effects of the thick glass
> do indeed NOT cause a "big difference" relative to the LCD.

Boy, when you get hung up on something, you really get hung
up on it, don't you?

Is it as "big" a difference as, say, the problems with color purity
or convergence or susceptibility to external fields? No, not in the
minds of most customers. Still, the optical effects of the thicker
faceplate IS a very obvious and visible difference (and yes, it IS
one that I have known to be a deciding factor in the purchasing
decision in some cases). But, as with all things, it's going to depend
a lot on what a particular customer considers to be "big."

> I've been right all along, and you knew it, Bob.

Hey, Chris, if it makes you feel good to keep score on this, and
you think you "got one" here, by all means - knock yourself out.
It makes very, very little difference to me one way or another.
You're certainly "right" in that you do not see this as a significant
difference. You are equally certainly wrong if you believe that
everyone trying to decide between these types shares that view.
All I can do is to try to explain what differences exist, where they
come from, and what if anything can be done to compensate for
them.

Bob M.

 

[Guest]

Archived from groups: comp.sys.ibm.pc.hardware.video (More info?)

"assaarpa" <redterminator@fap.net> wrote in message
news:cd1sdf$muh$1@phys-news1.kolumbus.fi...
> Excuse the point was that DVI is inexpensive way to get a crisp signal to
> the display device. All the better if the device is digital.

Within its limits, yes. However, I would have to at this point note
that many of the devices you may think are somehow "digital"
in reality are not, and are not strongly tied to digital interfaces.
(Would you be surprised, for example, to learn that a major
manufacturer of LCD panels and monitors for years produced
panels which kept the video signal in analog form all the way through
to the pixel level?)


> Elimination of two parts in signal degration can be eliminated: the source
> (which is primary problem these days) and the cable (which is minor
problem
> overall compared to the signal source being poor to begin with).

That's not an altogether correct assessment. There are poor
sources and good sources; there are poor cables and good cables.
In my experience, there are far more extremely poor cable
assemblies out there than there are analog-video-capable ICs
which are beyond all hope.


> Now the display could have a shot at doing best possible job it can
> presenting the digitally accurate data that comes in.

Digital does not automatically equate to "accurate." You
have to look at the specifics of the given implementations
being compared, both "analog" and "digital."


> Making square waveform at high frequency is very difficult hence
expensive.
> Analogic signal will be smoothed and inprecise because of this.

Why do you believe that a "square waveform" is desirable? Pixels
are not, in terms of the data itself, to be considered as
"little squares," but rather as point samples of the intended "original"
image (even if that image is CG). All that is required of either
transmission
system is that the intended luminance value for each video channel
can be unambiguously determined for the sample in question - which
generally means simply that the signal be at that value at the defined
sample time. (Once again, this shows the importance of pixel-level
timing to a fixed-format display.)

> Digital
> signaling has tolerance and treshold which determine how the signal is
> interpreted: high or low voltage. This is what digital means.

Well, no, actually what you have described is a particular form
of "digital" encoding. Most generically, "digital" simply means
that the transmitted information is being provided in the form of
numeric values, as opposed to levels analogous to the original
information (which is what "analog" means). How those numeric
values are encoded and transmitted varies depending on the
specific implementation in question. (For instance, what a
modem works on is not simply "high or low voltages"; nor is
that a good description of what goes over the air in digital
television systems. Both of these are undeniably "digital" in
operation, though.)

> Still with me?
> When we want to design and manufacture a circuit which does this is with
> current manufacturing processes inexpensive, the problem solved itself
when
> the manufacturing processes got better and faster chips could be produced
at
> cheaper prices.

Yes, but there's nothing particularly efficient about simple binary
encoding, nor is it necessarily easier to make a chip which will
transmit information in that form at the required rates (which is
why, for instance, a single-link DVI connection still doesn't provide
the data capacity of some analog RGB video connections).


> How much you think a Silicon Image DVI controller costs? How much you
think
> accurate digital-to-analogic converter would cost?

The total system costs of these two are actually not all that
different. (Remember that a DVI link running at the spec maximum
limit is carrying serial data on each of its three pairs at a raw rate
of slightly over 1.6 Gbits/sec; that is NOT a simple trick to pull
off...)


> > What of it? A plasma display can process a digital signal directly,
with
> > the only analog conversion being the intensity at each pixel.
>
> The what that DVI is gaining ground even on television sets.

Ummmm...I'm not sure what you meant to say there, but if
it was a supposedly rhetorical questions as to why DVI is
"gaining ground" in the television market, the main reason is
the fact that it supports a fairly robust content-protection
system. It is not because the video is better carried in digital
form (especially for CRT-based televisions).


> > So? The only data-grade monitors I know of whcih have coponent,
s-video,
> > composite, or scart inputs are projectors.
>
> I was talking more in the context of television use for the Plasma
Displays,
> for instance the DVI input has better image quality than the DB15. Infact
> the DB15 image quality is much worse than s-video from DVD set-top-box!

Apples and oranges. An HD-15 (what you're calling a "DB15")
connection - more commonly called just a "VGA output" - will
hardly ever be seen carrying video at less than a 31 kHz line rate
and about a 25 MHz pixel rate, and generally will be outputting
something a good deal faster. However, an S-Video output is
by definition not carrying anything beyond baseband "TV" video,
which is well under 10 MHz total bandwidth (closer to 5 MHz,
tops, in fact).


> > Why would one need an audio interface on a PC graphics card?
>
> HDMI is a digital A/V connector, sorry to burst your bubble, you might
want
> to check out:

Well, yes and no. The version of HDMI appearing on
certain CE equipment carries audio. There's also a dual-link
version that was intended for the PC market, which has to date
never seen much usage, which doesn't necessarily provide
audio support. HDMI, when it does appear on PC products
these days, is generally there as an interface to "TV" products,
much as PCs have in the past provided S-Video outputs.


> That may be true from the *technical* perspective, but I was steering away
> from the DB15 not because of the theoretical specs of the connector but
the
> poor quality of real-world implementations. My claim is that DVI doesn't
> leave any room for error in the signal carrying media like connector or
> cabling.

If the spec is followed, no - but then, that's true of any system.
And Gawd knows there have been plenty of examples of DVI
implementations that didn't play well with others.


> The display would be one package. Either it would be poor quality, or it
> would be high quality. Now the burden is put into three distinct parts:
> display, cable, graphics card. Each of which have to use high-quality
> components so that none of them are the weakest link. DVI allows
inexpensive
> components to be used due to nature of the technology used.

Not really, with either technology - and it certainly does not
reduce the total system cost for CRT displays, for reasons
previously discussed.


> > That's also the trouble with analog. An analog cable that can carry
> > 3800x2400 without ghosting is not cheap. What of it?
>
> DVI cable that can, is inexpensive. QED.

Except that you currently need TWO dual-link DVI
cables (note that this is a total of eight discrete connector
components, and two cable assemblies each including
seven individually-shielded pairs) to support this; I can
guarantee you that neither interface type is "cheap" when
it comes to supporting these sorts of data rates. Neither
interface type has an inherent edge over the other in this
regard. Please note that DVI cable assemblies which
meet the specification requirements are NOT as
inexpensive as the cheapest VGA assemblies on the market;
conversely, for what you spend for a DVI cable, you
can get a VGA assembly of comparable performance.


Bob M.

 

[Guest]

Archived from groups: comp.sys.ibm.pc.hardware.video (More info?)

"assaarpa" <redterminator@fap.net> wrote in message
news:cd1sne$nc2$1@phys-news1.kolumbus.fi...
> > fundamentally analog-controlled devices - it's just that in the case
> > of the LCD, a digital-to-analog conversion typically takes place
> > within the display panel itself.)
>
> J Clarke says that is infeasible to have the da converter in the display,
> comments on that?

Specifically, he said that it wasn't advantageous to have a
DAC in a CRT monitor, and he is correct about that. In the
case of an LCD panel, the "DAC" function happens in the
column drivers - it's a whole different sort of animal, and these
run much more slowly than the pixel rate due to the nature of
how an LCD panel is driven.


> A friend of mine "cleaned" his GeForce4 by removing resistors from his
> card's BCP few years back, which lead me to look into the articles on the
> topic back then. The generic idea in the article was that the resistors
were
> there to dampen the signal so that the cabling wouldn't emit too much
> interference, RFC regulation related anyway, the article might still be
> online somewhere.

Actually, what was removed was most likely small capacitors
(which in surface mount form, are difficult to distinguish from
SMT resistors). This can in some cases improve signal rise times,
but on the other hand it's certainly possible to put EMI-reducing
filter networks on video outputs that have no visible impact at all
on the signal (I've done this in several designs). Removing
RESISTORS from the outputs of video cards, if that's really what's
being done, generally isn't a good idea, as it is changing the
cable termination at the source end (which can lead to some
severe "ghosting" problems in many cases).

Bob M.