30-bit Color on 24-bit Hardware

[Bob Myers]

This debate is recurring. The last one was not long ago. The
numbers in Stephen's reference have been discussed over and
over, with the same conclusion every time. The reference's
numbers show (correctly) that CRTs have orders of magnitude
more dynamic range, especially for near blacks, then it proceeds
to claim that this isn't an advantage, because of "typical office
lighting conditions" -- which is completely irrelevant to the issue.

Sorry, but the proposition you made wasn't about
dynamic range, it was about gamut. You've been asked
at least twice now (well, three times, counting this) to
provide values that would back up YOUR claim of a wider
gamut for CRTs vs. LCDs, and we're still waiting. Further,
the notion of "orders of magnitude" more dynamic range is
clear nonsense in the first place - even if that WERE the
subject under discussion here.

Those who claim LCDs have anywhere near the same gamut as
a decent CRT are generally the same people who've never even
heard of a light hood.

Possibly in your experience, but also irrelevant here. I
personally have over twenty years' experience testing
displays, including both CRTs and LCDs, and have very
recent test data (from products currently on the market,
of both types) which shows that the claim of wider gamut
for the CRT is nonsense. It was certainly true in the past,
but has not been true for some time now.

So again - I would invite YOU to either provide some
numbers, or tell us just what oddball definition of the term
"gamut" you're using here.

Bob M.

 

[Rick]

Sorry, but the proposition you made wasn't about
dynamic range, it was about gamut. You've been asked
at least twice now (well, three times, counting this) to
provide values that would back up YOUR claim of a wider
gamut for CRTs vs. LCDs, and we're still waiting. Further,
the notion of "orders of magnitude" more dynamic range is
clear nonsense in the first place - even if that WERE the
subject under discussion here.

I've referenced the thread that gives these numbers. It's in
Stephen's own "research paper". You trimmed that reference,
and now are claiming one haven't been given. So you're either
lazy, stupid or (most likely) just another pathetic troll like
Stephen.

Look, I learned a long time ago there no accounting for
intelligence. If you believe LCDs have an equal or greater color
gamut to CRTs, then by all means use them. You'll just be one
more moron over which professionals will have a competitive
advantage.

Rick

 

[James Akiyama]

Actually, I believe the on-axis viewing of LCD's can be as good if not
better than a CRT. The problem with LCD's are that they tend to suffer
chroma and luminance shifts when being viewed off axis (viewing angle
issues). Apple Cinema displays have come a long way in addressing this, but
it's still not fully addressed. One particular problem here is that when
you're viewing a very large display close-up, there is a considerable change
in your effective viewing angle across the screen. All of this leads to
difficulties where absolute color-accuracy is needed.

James

 

[usenet]

[Followups set back to rec.photo.digital only - please don't help this
troll spray this disinformation all over Usenet.]

Kibo informs me that westin*[email protected] (Stephen H.
Westin) stated that:

In any reasonable lighting level, a good LCD will be at least
competitive in dynamic range with any CRT.

I (& others) debunked this claim of yours the last you brought it up
here. The only time that statement is true is under bright office
lighting, where the reflected glare on CRTs reduces their effective
contrast down to below that of LCDs under the same lighting. Anyone
who's serious about colour work operates under subdued lighting to
prevent that effect.

The days of all CRT's being
better than all LCD's are past.

As Rick pointed out, the limitations of LCD backlighting currently make
LCDs greatly inferior to CRTs WRT to black levels & shadow detail. I
don't know of any LCD on the market that can match even a generic CRT in
that area.

 

[Stephen H. Westin]

This debate is recurring. The last one was not long ago. The
numbers in Stephen's reference have been discussed over and
over, with the same conclusion every time. The reference's
numbers show (correctly) that CRTs have orders of magnitude
more dynamic range, especially for near blacks, then it proceeds
to claim that this isn't an advantage, because of "typical office
lighting conditions" -- which is completely irrelevant to the issue.

Those who claim LCDs have anywhere near the same gamut as
a decent CRT are generally the same people who've never even
heard of a light hood. I invite you to look up these previous
threads in Google's Usenet archive, e.g. http://tinyurl.com/623mb

Rick

This debate is recurring. The last one was not long ago. The
numbers in Stephen's reference have been discussed over and
over, with the same conclusion every time. The reference's
numbers show (correctly) that CRTs have orders of magnitude
more dynamic range, especially for near blacks,

First, a CRT with a totally black screen will be much darker than
an LCD with the same image. Unfortunately, that gets a lot brighter
if you actually display anything useful on the screens. Orders of
magnitude brighter, so that the dynamic range of a CRT with reasonable
image content isn't dramatically better than that of an LCD. As I recall,
the figure was around 4000:1 range for black screen vs. white screen,
but only about 350:1 when the majority of the screen is displaying
middle gray.

Second, "dynamic range" is the ratio between white and black, so
I don't know what "more dynamic range, especially for near blacks"
can possibly mean.

then it proceeds
to claim that this isn't an advantage, because of "typical office
lighting conditions" -- which is completely irrelevant to the issue.

Because, of course, you only use your CRT in a photographic darkroom..

Those who claim LCDs have anywhere near the same gamut as
a decent CRT are generally the same people who've

... actually measured them.

never even
heard of a light hood.

Oh do please explain how the color gamut of a display is
strongly affected by stray light.

As Bob pointed out, you may not understand what the term "display
gamut" means. It primarily refers to the chromaticities produced; as
Bob also pointed out, there doesn't seem to be an advantage of CRT's
over the best LCD's these days.

I invite you to look up these previous
threads in Google's Usenet archive, e.g. http://tinyurl.com/623mb

Hmm. You might actually want to read them yourself.

I recall my reference to the Munsell Color Science Lab at RIT and your
vehement misinterpretation as meaning the Munsell Labs subsidiary of
GretagMacbeth. I remember my quotes from their director, and your
references to "most professionals", whoever they may be, and "almost
everyone who has tried it". I recall posting links to various
measurements, and in fact measuring my own CRT, and your never
responding once with any sort of quantitative assessment.

As far as I can tell, you don't know much about color science, don't
know the meaning of the terms you throw around, and would rather
follow "conventional wisdom" than actually find out the truth of the
matter.

 

[usenet]

[Followups set to rec.photo.digital only - if anyone else wishes to
respond to this troll, please snip out the extra groups as well.]

Kibo informs me that westin*[email protected] (Stephen H.
Westin) stated that:

Papers. From several sources.
Bull.

With actual measurements.

Taken under typical office lighting, as I discovered when I read the
paper you'd been quoting. Nobody doing serious colour/photographic work
operates under office lighting.

 

[usenet]

[Followups set to rec.photo.digital only]

Kibo informs me that (e-mail address removed) (Ron W. Silvas) stated
that:

Hi, Mr. Myers.

It might be time for me to exit from this branch of the
discussion, as I am admittedly not very knowledgeable about the
areas being discussed (dynamic ranges, contrast ratios,
hardware behavioral differences between LCD and CRT
technologies, etc.). =)

My earlier comment . . .


. . . was a jump to a conclusion on my part–

But a correct one, in most cases.

I should have asked
Mr. Westin if he could explain to me his comment about the
different dynamic ranges before I commented on that. Totally
my mistake.

Mr Westin is mistaken in his comments anyway.

o 30-bit color's >1000 intensity levels per channel *is*
finally enough.
o 30-bit color display is desirable.
o 30-bit color display on common 24-bit hardware would be a
nice bonus, and . . .

It works. On CRTs and LCDs, alike. No special hardware
requirements. No special OS support needed.

Indeed. Did you know that the Matrox Parhelia cards already have true 30
bit resolution in hardware?
<http://www.matrox.com/mga/workstation/digital_design/products/parhelia/256mb.cfm>

 

[Bob Myers]

I've referenced the thread that gives these numbers.

Right...and I perused that thread as far as I needed to,
to see that it was long on opinion and short on objective
data. Since I already have data from any number of sources,
including direct personal experience in the test lab, which
confirm my assertions regarding display gamut, it's not
my job to dig through that pile in search of the evidence that
YOU need to support YOUR assertion. I (and Stephen,
for that matter, at least as I understand his comments to
date - but I'll let him speak for himself) am asking that
you supply exactly six pairs of numbers - giving the
measured coordinates of the three primaries for
representative samples of the two technologies. I've already
provided a set that I believe is very representative; if
you don't have contradictory data, then I guess we'll just
have to take the matter as settled, won't we?

It's in
Stephen's own "research paper". You trimmed that reference,
and now are claiming one haven't been given.

No, I have made no such claim. I have said that YOU
have given no values which support YOUR assertion,
here. The truth of that statement is evident from a simple
review of the current thread. Which would then lead one
to believe that you are, in fact,

either
lazy, stupid or (most likely) just another pathetic troll

Look, I learned a long time ago there no accounting for
intelligence. If you believe LCDs have an equal or greater color
gamut to CRTs, then by all means use them.

Not only do I believe this - at least, under the commonly-
accepted meaning of the term "gamut" (God only knows
what you mean by that term) - but I have (and have presented
here) data which demonstrates this. It is up to you to show
contradictory evidence, or to retract your claim. It's rather
funny (and more than a little pathetic) that you're trying to
label the only people to actually present data here as "lazy"
when you yourself are unwilling to do so. What's the matter,
is it just too difficult to type in those twelve values?


Bob M.

 

[Bob Myers]

(comp.sys.ibm.pc.hardware.video added back in; please leave it,
as this is the group in which I normally see these discussions, and
this is quite a bit more relevant to a discussion of electronic display
technology - nominally the field covered in that group - than in
digital photography per se.)

LCDs greatly inferior to CRTs WRT to black levels & shadow detail. I
don't know of any LCD on the market that can match even a generic CRT in
that area.

Unfortunately, no one seems to be able to describe precisely
what these "limitations of LCD backlighting" ARE. Nor do these
factors ("black level" and "shadow detail") have anything to do
with color gamut, which was the original point of disagreement.

However, since you've brought these items up, let's dispose of
them as well.

It is absolutely correct that the LCD panels as typically used in
current monitor products provide a higher luminance at the
"black" level than a properly-adjusted CRT. However, this
is not necessarily indicative of an inherent inability of the LCD
to match the CRT's performance in this area. As they have
been driven by today's marketing expectations, LCD monitor
panels also offer a very significantly higher white luminance
(as much as 3-4X) over typical CRT monitor products.
Since the LCD does operate as a "light valve" - and it is
impossible to completely block the light from the backlight
- it is also correct to say that the LCD will, in practical terms,
never provide an absolutely zero-luminance black. But then,
neither does the CRT when properly adjusted. A CRT
monitor's black level must be set very slightly above zero,
or else the CRT cutoff point would have to be constantly
adjusted (the black would tend to slip below blanking,
making the lower end of the luminance range very non-linear).
Typical real-world values for the black-level luminance will
be in the range of 0.05-0.5 cd/m^2 (most often toward the
upper end of this range), vs. somewhere around 1 cd/m^2 or
so for a typical monitor LCD. But since the LCD's white
these days is up around 350-400 cd/m^2 (vs. something
more like 100 for the CRT), simply cutting down the backlight
output (or, in a more brute-force approach, filtering the LCD)
would bring its black and white values down to something
very comparable to the CRT. That few products do this
says much more about the demand in the market than it does
anything at all about inherent differences in the technology.

In terms of the "shadow detail" (i.e., the ability to
discriminate "shades of gray" at the low end of the luminance
range, we get back to differences in the response
characteristics of the two technologies - which, if you'll
look back over this thread, was my point in the first
place.

Bob M.

 

[Bob Myers]

Taken under typical office lighting, as I discovered when I read the
paper you'd been quoting.

I won't say anything about the "paper you've been
quoting", but I will note that all measurements I've been
speaking of in this thread were made under both dark
ambient (per the procedures of the VESA Flat Panel
Display Measurement standard, version 2.0) and, for
comparison purposes under "office" conditions (as the
monitors I test are generally sold for use in office
environments). So I am both very familiar with the
differences in test results between the two environments,
and have extensive test data taken under both sets of
conditions.

Nobody doing serious colour/photographic work
operates under office lighting.

While I understand the perspective you're coming
from in this statement, it is not altogether correct. For
example, those who are doing work in color graphics
which are ultimately expected to be viewed by average
computer users in typical environments (say, web page
design) are well-advised to view their products not only
in their preferred low-ambient environment, but also
under the conditions that end user of the product will
be experiencing.

Bob M.

 

[Phil Green]

Well, YOU made the claim that a given CRT monitor
"outgamuts" an LCD. Stephen quite rightly asked you to
justify that claim by describing the gamut of the monitor in
question in quantitative terms. I'd also like to see that, and
am willing to bet that you can't provide numbers that support
that specific claim. Whether or not "any serious attempt at
color correcting... is an exercise in futility" is not relevant to
the claims you've made. (In other words, you may be right
in that broad statement, at least for most current monitors,
but not for the reasons you stated.)

So let's see your numbers. End of story.

Bob M.

The colour gamut depends on the colour filters as well as the backlight.
Just to add some data to the mix: I recently measured a prototype EIZO
CG LCD (which was claimed to have an AdobeRGB gamut), with a Minolta
CS1000, in our lab. Set up to a nominal 200cd/m2 and D65 white point,
room blacked out:
x y Luminance (cd/m2)
White 0.3141 0.3133 119.6
Red 0.643 0.3328 38.42
Green 0.2058 0.7147 72.49
Blue 0.1464 0.0655 10.89
Black 0.3043 0.34 0.63

These data agree closely with the target AdobeRGB (1998) primaries, and thus
the gamut is considerably larger than a CRT using EBU or similar
phosphors as primaries.

I think you will see professional prepress graphics users increasingly
adopting LCDs as price/performance improves.

In the above measurements, the screen was displaying a set of mixed
colour patches so the black point includes some flare light within the
face plate. As an aside, claims regarding dynamic range depend heavily
on the instrumentation used to measure the black - relatively few
instruments can get down even to 1 cd/m2.

--
Phil Green
Colour Imaging Group
London College of Communication
Elephant and Castle, London SE1 6SB
Tel: +44 020 7514 6759 Fax: +44 020 7514 6772
http://www.digitalcolour.org

 

[Ron W. Silvas]

Hi, imagenoir person. =)

But a correct one, in most cases.
. . .
Mr Westin is mistaken in his comments anyway.

Yeah, I really wouldn't know--This project was mostly concerned
with color depth extension.

Indeed. Did you know that the Matrox Parhelia cards already
have true 30 bit resolution in hardware?

Yes, thank you for making sure, though. I actually have a
Matrox P650, which is like the baby brother of the Parhelia--
It has their 30-bit "GigaColor", also. But, unfortunately . . .
o I don't believe there's currently much OS or software
support for 30-bit color. (Outside of Windows's
DirectX 8+, anyway?)
o No 30-bit color on nearly all of today's LCDs, as far as
I can tell.

But with the synthesized 30-bit color that this project dealt
with, none of that really matters, right? =)

Thank you for checking about the Matrox deal, though,
- Ron

 

[Rick]

The colour gamut depends on the colour filters as well as the backlight.
Just to add some data to the mix: I recently measured a prototype EIZO
CG LCD (which was claimed to have an AdobeRGB gamut), with a Minolta
CS1000, in our lab. Set up to a nominal 200cd/m2 and D65 white point,
room blacked out:
x y Luminance (cd/m2)
White 0.3141 0.3133 119.6
Red 0.643 0.3328 38.42
Green 0.2058 0.7147 72.49
Blue 0.1464 0.0655 10.89
Black 0.3043 0.34 0.63

These data agree closely with the target AdobeRGB (1998) primaries, and thus
the gamut is considerably larger than a CRT using EBU or similar
phosphors as primaries.

I think you will see professional prepress graphics users increasingly
adopting LCDs as price/performance improves.

At $2800 for the 21" model, there's a lot of "improvement"
to go. Personally I'd rather have FOUR 22" Mitsu 2070's
for the same money.

Rick

 

[Ron W. Silvas]

Hi, Rick.

No, I haven't written any sort of print driver.

And, like you, I suspect that visualizing in color depths beyond what
your target output device is capable of producing might not be the best
idea, but . . .

Paper is by no means the only target medium out there these days,
right? The video game and multimedia industry produces content for
display on computer monitors and televisions, the motion picture
industry--including their animation and special effects studios--
produce content for film and 30-bit digital projection, and I believe
medical imaging usually requires greater than 24-bit color fidelity (or
greater than 8-bit grayscale, etc.).

You know, the more I thought about this, I don't see any problem
at all with visualizing work destined for 24-bit output in 30-bit
color. This seems like a totally healthy workflow, to me:
o Capture/Create and Work in High Fidelity Color (48-bit, etc.)
o Display in 30-bit Color
o Appropriately Convert (probably via dithering) to 24-bit
Target for Output

You don't need print driver support for that--Photoshop will
optionally dither when converting from 16-bit channels to 8-bit.
I'm not sure if it will automatically do this when a 48-bit image
is sent to print--I suspect--but, in either case, you can still
get your dithered 24-bit output.

Does that seem right?
- Ron

 

[Bob Myers]

The colour gamut depends on the colour filters as well as the backlight.
Just to add some data to the mix: I recently measured a prototype EIZO
CG LCD (which was claimed to have an AdobeRGB gamut), with a Minolta
CS1000, in our lab. Set up to a nominal 200cd/m2 and D65 white point,
room blacked out:
x y Luminance (cd/m2)
White 0.3141 0.3133 119.6
Red 0.643 0.3328 38.42
Green 0.2058 0.7147 72.49
Blue 0.1464 0.0655 10.89
Black 0.3043 0.34 0.63

Thank you very much for the numbers, Phil; at least
SOMEONE else out there has actually measured these
things. I haven't tested that particular product, but these
values are certainly in line with what I would expect from
a current high-end product (which is, of course, exactly
what the oft-referenced "professional graphics" sort
of user would actually be using). That's a pretty
impressive green. by the way!

Still no data from the "CRTs 'outgamut' LCDs"
crowd? Sigh....

I think you will see professional prepress graphics users increasingly
adopting LCDs as price/performance improves.

Agreed; in fact, I've recently been working very closely
with a number of customers of this category, and I
would say that the above chageover is already well on
the way.

In the above measurements, the screen was displaying a set of mixed
colour patches so the black point includes some flare light within the
face plate. As an aside, claims regarding dynamic range depend heavily
on the instrumentation used to measure the black - relatively few
instruments can get down even to 1 cd/m2.

Also agreed, strongly! In fact, the difficulty of properly
measuring "black" is why just about all published
contrast, etc., claims are to be taken with an enormous
grain of salt as the default position (i.e. until you know
something about the measurement procedures and
equipment used).

Bob M.

 

[Bob Myers]

At $2800 for the 21" model, there's a lot of "improvement"
to go. Personally I'd rather have FOUR 22" Mitsu 2070's
for the same money.

And let me guess - you believe there's something inherently
pricey in the PANEL that Eizo uses in this monitor that
accounts for this pricing?

Bob M.

 

[Phil Green]

Thank you very much for the numbers, Phil; at least
SOMEONE else out there has actually measured these
things. I haven't tested that particular product, but these
values are certainly in line with what I would expect from
a current high-end product (which is, of course, exactly
what the oft-referenced "professional graphics" sort
of user would actually be using). That's a pretty
impressive green. by the way!

It certainly looks impressive. They chose these primaries at the design
stage, based on the fact that a large proportion of people in publishing
and printing use them in defining a calibrated RGB editing space (since
the blue-green region encompasses the gamut of offset litho printing,
unlike the EBU-type primaries).

Without the restriction of rare-earth phosphors, other 'primaries'
should also be possible.

--
Phil Green
Colour Imaging Group
London College of Communication
Elephant and Castle, London SE1 6SB
Tel: +44 020 7514 6759 Fax: +44 020 7514 6772
http://www.digitalcolour.org

 

[chrisv]

Also agreed, strongly! In fact, the difficulty of properly
measuring "black" is why just about all published
contrast, etc., claims are to be taken with an enormous
grain of salt as the default position (i.e. until you know
something about the measurement procedures and
equipment used).

It seems, then, that it would be better to simply calculate the black
level, based on the known properties of the LCD and it's backlight.

 

[Stephen H. Westin]

It certainly looks impressive. They chose these primaries at the
design stage, based on the fact that a large proportion of people in
publishing and printing use them in defining a calibrated RGB editing
space (since the blue-green region encompasses the gamut of offset
litho printing, unlike the EBU-type primaries).

Without the restriction of rare-earth phosphors, other 'primaries'
should also be possible.

At SIGGRAPH, NEC/Mitsubishi was showing a prototype LCD with
tricolor LED backlighting. Not only is the white point adjustable,
but the primaries approach the spectrum locus!

<snip>

 

[Stephen H. Westin]

And let me guess - you believe there's something inherently
pricey in the PANEL that Eizo uses in this monitor that
accounts for this pricing?

Well, it's another change in subject. Color gamut turned into
dynamic renge, and now it's price. Rick apparently doesn't know
what a top-quality CRT costs. And presumably the extra cost
goes into ameliorating the deficits inherent in a CRT: geometric
accuracy, long-term stability, etc.

My impression is that at the high end, an LCD will beat a CRT.
But cheap LCD's seem to be much worse than cheap CRT's.