How Necessary is a Wide Gamut Display?

[Bob Rockefeller]

While certainly more is better, how necessary is it to use a wide gamut display to get professional level results? I read about many pros using Apple's 30" Cinema Display which is an excellent display, but not close to what is commonly referred to as wide gamut. The NEC PA series of monitors is claimed to be around 98% of the Adobe RGB gamut and is expensive enough for me to believe that.

What do other professionals use?

[T_om]

How necessary is it for you to see the colors you are wanting to reproduce?

I use a HP LP3065 and it is a fine monitor.  IPS panel and (very important) able to be turned down in luminance enough to get reasonable density correlations between screen and print, which some monitors are not able to do.  Although it is not quite as accurate as the new "Wide Gamut" monitors, it is accurate enough for me (weddings and portraits). 

Tom

[Scott O.]

Nice but not necessary.  Until recently, nobody had one and I saw many fine prints being made...

[digitaldog]

As said, not necessary but nice. If I had an option to use a standard sRGB like display that was high bit, had great software to calibrate and profile it (like the NEC SpectraView II 2490) or an extended gamut display that didn’t have these capabilities, I’d far prefer the 2490.

The Apple 30” is nothing at all special.

[ErikKaffehr]

Updated:

I stand corrected regarding 10 bit wide display pipeline. Also I obviously meant display pipeline and not image processing pipeline.

Sorry for bad info!
Erik
--------------------

Hi,

If your colors are inside the sRGB gamut it doesn't matter, if they don't it may not matter either. There will always be some colors that you cannot see on screen but can print and the other way around. Karl Lang has some reservations about using large gamut displays with 8-bit/color video as some tonal differences cannot be exactly resolved with 8 bits in wide gamut. Display Port is 10 bits, but I'm under the impression that the image processing pipeline in Photoshop is still 8 bit.

I use a somewhat simpler Eizo screen with near Adobe RGB gamut, and there are some colors that are distinguishable using Adobe RGB but not with sRGB but they are few.

You could test it by using a good test image in Adobe RGB or even Prophoto RGB and print it on a good printer like an Epson. Convert the image to sRGB and print again with the same printer settings. I used Jack Flesher's image but I don't know where to find it. It's based on Bill Atkinson's image which is here:

http://homepage.mac.com/WebObjects/FileSharing.woa/wa/Profile_Test_Images.zip.zip?a=downloadFile&user=billatkinson&path=/Public/Profile%20Test%20Images.zip



Best regards
Erik

[Roscolo]

The NEC PA series of monitors is claimed to be around 98% of the Adobe RGB gamut and is expensive enough for me to believe that.

What do other professionals use?

I think it is very necessary if you need to do much color critical work. Necessary as in "major time and money saver." And you don't have to spend a lot of money. I went with the NEC P221W with SVII Pro. I paid only $585 and it comes with everything you need to calibrate and is VERY accurate. Pays for itself in the first day or two of use with the time you save not having to run prints a third or fourth time.

Prior to that I did get really good results from a reliable Samsung CRT I calibrated with an old Spyder 2, but the as the CRT aged it eventually got to the point I couldn't calibrate it anymore. Love the NEC P221W.

[Mulis Pictus]

If your colors are inside the sRGB gamut it doesn't matter, if they don't it may not matter either. There will always be some colors that you cannot see on screen but can print and the other way around. Karl Lang has some reservations about using large gamut displays with 8-bit/color video as some tonal differences cannot be exactly resolved with 8 bits in wide gamut. Display Port is 10 bits, but I'm under the impression that the image processing pipeline in Photoshop is still 8 bit.

That's fortunately not true anymore and one can have 10bits output with few cards. CS4 and CS5 can render 10bits per color channel thru OpenGL See http://www.amd.com/us/Documents/48108-B_ATI_FirePro_Adobe_10-Bit_FAQ_R5_Final.pdf, some Nvidia workstation cards can do that too. Strictly speaking the image processing pipeline in Photoshop is wider and you can edit images in 16bits or 32bits per color channel, but I guess you meant the displaying pipeline.

It works nicely here with EIZO cg243w/ATI firepro v3800/Photoshop CS4,CS5 - with the color matching enabled (without it it would be useless).

[narikin]

Interesting information about 10 bit display card. thanks.

the only problem is that we don't need the ultra power of these cards, but do need the 10 bits, so are stuck with beefing up everything (PSU, power leads, case cooling, etc) not to mention having a noisier unit (high power fan on card and PSU) to get 10bit pipeline on these hungry Pro cards.  Unless you know of a lower priced quieter open GL card now, for mid level users?

I am using the NEC 30" 3090WQxi, and it has been flawless since the day I got it. the Adobe RGB space is instantly apparent, the moment you turn it on, Greens especially. Hopefully they will upgrade it to 10 bits soon, (NEC do have 10bit approved monitors on ATI's list) as I can't live without 30" anymore!  The only other hope is the Eizo ColorEdge CG301W, which is very pricey at $5000+


PS1: Yes the Apple 30" display is poor and very behind the times. don't bother.
PS2: Karl Lang words against >8bits, must be the oldest piece of display (mis)information on the web. Ignore it now in 2010.

[narikin]

The NEC MultiSync PA271W seems a very good monitor, with almost 30" resolution, 10bit pipeline, Adobe RGB, IPS screen, and a street price of $1350 - terrific value for what it. It is also a very new product - May 2010.

[Rocco Penny]

ok just paying attn

I'm not sure the current required to run a Firepro card is greater than the current to run a mid-priced card like say a 5770 or a 465.
Taking my time this time for an image rendering solution, in a desktop & monitors,
with energy efficiency as a high criteria,
I've studied the wattage draw for an entire powerful computer & monitors.
If you choose properly matched components for an image dedicated tower on an I7 platform the power consumption between the firepro card and another mid-priced card is negligible.
In fact I'm gonna go on a limb here and say any high-end card will draw more than a firepro 3800 and ATI won't say you'll get 10 bit screen output unless a firepro is used.
Plus they're cheaper than most high end gak
I'm tiring of wide gamut,
it's fun for the kind of crazy colorful work I, as a beginner, like to do.
My brightness is turned down to 25% so I'm not sure the wide part of wide gamut is what I see on a cheaper 8 bit screen.
If i have this correct,
8 bit 16.7 million colors are actually augmented by the wide gamut.
So what you see in native resolution may be quite different than what you see when you're looking at the same live scene.
Even calibrated I couldn't make my prints match my screens.
I started doing calibration by eye, in the wee hours with a coke can against a green velvet backdrop.
Then just seeing something over and over gets you pretty good at calibrating.
The wide gamut screens aren't compatible with every calibration device.
If it were me and I had the budget, I'd skip a least expensive option, and jump in at the NEC 10 bit 27 inch level making sure I had the firepro to back it up
10 bits, no brainer

[digitaldog]

That's fortunately not true anymore and one can have 10bits output with few cards. CS4 and CS5 can render 10bits per color channel thru OpenGL See http://www.amd.com/us/Documents/48108-B_ATI_FirePro_Adobe_10-Bit_FAQ_R5_Final.pdf, some Nvidia workstation cards can do that too.

Now we just need at least one more OS to support the full 10-bit path (my understanding is that’s not yet the case on Mac).

[Sven W]

Now we just need at least one more OS to support the full 10-bit path (my understanding is that’s not yet the case on Mac).

It seems that Pete Myers use a Mac Pro in his review:
http://www.luminous-landscape.com/reviews/accessories/10bit.shtml

/Sven

[digitaldog]

It seems that Pete Myers use a Mac Pro in his review:
http://www.luminous-landscape.com/reviews/accessories/10bit.shtml

So how is he empirically determining the path is fully high bit, in and out? While my sources deep inside the “industry” must remain anonymous, his credentials are impressive enough to me that when he mentioned that we are still not fully high bit through the entire path, I take that seriously. But if there is a way to determine the path really is high bit all the way, I’m all ears.

[Mulis Pictus]

the only problem is that we don't need the ultra power of these cards, but do need the 10 bits, so are stuck with beefing up everything (PSU, power leads, case cooling, etc) not to mention having a noisier unit (high power fan on card and PSU) to get 10bit pipeline on these hungry Pro cards.  Unless you know of a lower priced quieter open GL card now, for mid level users?

As Rocco said, many of them are not that power hungry (the cheaper ones). v3800 requires no more than 43W and has quite small heat sink. It costs around $100.

[narikin]

I stand corrected, thanks. The ATI 100-505606 FirePro V4800 seems like a good mid point Pro card at $160, and the latest NEC monitors like the PA271W come with Adobe RGB and 10 bit support plus Display Port connectors.  I will certainly look into this combo for my next machine, which is only a few weeks away.  I will probably order a low noise machine as I like to work in near silence.

Luckily am on 64bit Windows 7, and a big fan of it, so the full 10 bit pipeline should not be a problem!

[tho_mas]

Karl Lang has some reservations about using large gamut displays with 8-bit/color video as some tonal differences cannot be exactly resolved with 8 bits in wide gamut.

what Mr. Lang was referring to is only true if image-value = target-value for the monitor. But in a color managed workflow this won't happen.
If you are viewing a 16bit wide gamut image in Photoshop, then Photoshop will convert the source profile into the monitor profile... and first this data is going to the display. Too, internal high bit processing of current monitors will re-calculate transitions on the display (this is why you effectively can see a change in transitions when you are swichting from 8bit to 16bit mode in Photoshop).

[digitaldog]

what Mr. Lang was referring to is only true if image-value = target-value for the monitor.

I’m not so sure. I suspect the famous post made by Karl several years ago is what is being referred to here. Here’s the original text from the Betterlight forum, you can all examine what may or may not have changed in technology since then:

Karl Lang wrote:

Greetings, fellow betterlight users. I lurk here and I try to keep my
mouth shut ;-) I can't spend too much time on this right now so
forgive me if I don't respond to questions quickly. For those of you
who don't know me I was the architect of the Sony Artisan, the Radius
PressView, ColorMatch, ProSense and many other products. I have worked
with display technology both CRT and LCD for the last 15 years.

Color accurate LCDs pose many problems. I will not argue the CRT vs LCD
debate. Suffice to say there are elements of a calibrated CRT that
still can't be matched by any LCD - available - and there are also
elements of LCD technology that exceed CRTs. We are improving things
at a rapid pace. I expect within 2-3 years to be able to finally feel
comfortable stating that we have an all around superior product in the
LCD space.

I am writing this email to attempt to dispel some myths and provide
some guidance for your LCD purchasing. You can't buy a good CRT any
more, the only ones left are of poor quality because the cost has been
reduced so much all the expensive quality components are not used
anymore. There was a reason that some CRTs cost 2-3K - the parts were
very expensive. Now the analog electronics use VLSI to reduce cost,
resulting in poor comparative quality.

1) A wide gamut LCD display is not a good thing for most (95%) of high
end users. The data that leaves your graphic card and travels over the
DVI cable is 8 bit per component. You can't change this. The OS, ICC
CMMs, the graphic card, the DVI spec, and Photoshop will all have to be
upgraded before this will change and that's going to take a while. What
does this mean to you? It means that when you send RGB data to a wide
gamut display the colorimetric distance between any two colors is much
larger. As an example, lets say you have two adjacent color patches one
is 230,240,200 and the patch next to it is 230,241,200. On a standard
LCD or CRT those two colors may be around .8 Delta E apart. On an Adobe
RGB display those colors might be 2 Delta E apart on an ECI RGB display
this could be as high as 4 delta E.

It's very nice to be able to display all kinds of saturated colors you
may never use in your photographs, however if the smallest visible
adjustment you can make to a skin tone is 4 delta E you will become
very frustrated very quickly.

2) More bits in the display does not fix this problem. 10 bit LUTs, 14
Bit 3D LUTs, 10 bit column drivers, time-domain bits, none of these
technologies will solve problem 1. Until the path from photoshop to the
pixel is at least 10 bits the whole way, I advise sticking to a display
with something close to ColorMatch or sRGB.

3) Unless the display has "TRUE 10 bit or greater 1D LUTs that are
8-10-10" user front panel controls for color temp, blacklevel and gamma
are useless for calibration and can in fact make things worse. An
8-10-8 3D LUT will not hurt things and can help achieve a fixed
contrast ratio which is a good thing.

Only Mitsubishi/NEC displays with "GammaComp" have 8-10-8 3D LUTs at
this time. Some Samsung displays may have this I don't test many of
their panels as the performance in other areas has been lacking.

Only the Eizo 210, 220 and NEC2180WG have 8-10-10 paths. If you really
want to know... the path in the Eizo is "8-14bit3D-8-10bit1D-10" go
figure that one out ;-) The 2180WG has an actual 10 bit DVI interface
with a 10-10-10 path but nothing supports it so you can't use it yet -
but for $6500 your ready when it does ;-)

4) The testing methodology for the seybold report article was very
poor. It demonstrates the authors complete lack of understanding with
regards to LCD calibration. At some point I may write a full rebuttal.
As an example the fact that Apple's display has no controls other than
backlight is actually a very good thing for an 8-8-8 LCD if your going
to use calibration. Apple optimizes the factory LUTs so as to provide
the most individual colors. smooth greyscale and the least loss. Then
the calibration is done in the graphic card LUT. As these are all 8 bit
it's best if the user does not mess with the display LUTs at all.
Overall Lab to Lab Delta E of 23 patches is a very poor metric to
evaluate a display. It completely leaves out many areas of color space
(the tool they used is designed to make the colorimeter look good so
tuff patches are not included) contrast ratio, stability, aging,
greyscale performance and other important considerations.

Many people ask for my recommendations. I am not happy with anything we
have right now. That said I can evaluate what there is.

Price performance wise the great bargain is the NEC 1980SXI BK the
price/vs colorimetric performance of this display can't be beat. The
2180ux Is a great display at a reasonable but high end price.

In the mid-high wide screen I like the Apple and the SONY. Reject the
display if uniformity is bad and make sure whomever you buy it from
will exchange it.

The Eizo 210 is great if you can justify the current cost. Give it two
years and most high-end displays should perform at this level. 220 is a
great display but suffers from all the downfalls of any wide gamut
display.

There is no reason to buy the La Cie 321 it's just an NEC with their
label on it and an extra $400.

The Monaco Optix XR is the best colorimeter for LCDs at this time.

These are my personal opinions.

Karl Lang

[tho_mas]

I suspect the famous post made by Karl several years ago is what is being referred to here.

Yes, that's exactly the post I had in mind.
The passage about the smallest possible adjustemtents don't make sense in a color managed workflow … and in this particular case the values are also wrong:

RGB 230/240/200 & RGB 230/241/200 =

sRGB Lab 93|-8|18 & Lab 94|-9|19
Delta E = 1,73

AdobeRGB Lab 93|-10|18 & Lab 93|-11|19
Delta E = 1,41

ECI-RGB Lab 94|-7|15 & 94|-8|15
Delta E = 1,00

which also makes sense as ECI-RGB with its Gamma 1.8 resolves better in those bright tonal values.

[Mulis Pictus]

Yes, that's exactly the post I had in mind.
The passage about the smallest possible adjustemtents don't make sense in a color managed workflow … and in this particular case the values are also wrong:

RGB 230/240/200 & RGB 230/241/200 =

sRGB Lab 93|-8|18 & Lab 94|-9|19
Delta E = 1,73

AdobeRGB Lab 93|-10|18 & Lab 93|-11|19
Delta E = 1,41

ECI-RGB Lab 94|-7|15 & 94|-8|15
Delta E = 1,00

which also makes sense as ECI-RGB with its Gamma 1.8 resolves better in those bright tonal values.

Not sure I see your point. Why would you round the Lab values here?

Without rounding I get Delta E (CIE1976) 0.64, 0.72, 0.59 for sRGB, AdobeRGB and ECI-RGB for 230,240,200 and 230,241,200 which makes sense as AdobeRGB is bigger space than sRGB and higher part of theirs gamma curves are similar. Thus one would expect higher DE in AdobeRGB, indeed 0.72 > 0.64  in this case.

The part about 1.8 gamma is true even without rounding, not quite sure how Karl got the DE 4 there. Which looks like one might indeed get nicer light colored skin tones on 8bit wide gamut displays when they are hardware calibrated to 1.8 gamma.

[tho_mas]

Not sure I see your point. Why would you round the Lab values here?

just thought to refer to the PCS... either way - with or without rounding the statement about the smallest adjustments is not true.