Ad
Ad
Ad
Pages: « 1 2 [3]   Bottom of Page
Print
Author Topic: Seems like you can make a really excellent sRGB mode on NEC PA  (Read 11696 times)
davewolfs
Jr. Member
**
Offline Offline

Posts: 97


« Reply #40 on: July 11, 2010, 09:04:39 AM »
ReplyReply

Quote from: LarryBaum
Actually I guess you are correct. The 2690^2 version does appear to let you calibrate the white point of the sRGB mode, the original version did not.

It still only has 1D 12bit LUTs and not 14bit 3D though so hah!

(without a 3D LUT green often looks like orange and sometimes white switches with black   )

I think one of the most interesting features if I understand it correctly.  Is that I can calibrate to my printer profile.  Wouldn't that eliminate the need for soft proofing?
Logged
Mark D Segal
Contributor
Sr. Member
*
Offline Offline

Posts: 6767


WWW
« Reply #41 on: July 11, 2010, 09:11:09 AM »
ReplyReply

Quote from: davewolfs
I think one of the most interesting features if I understand it correctly.  Is that I can calibrate to my printer profile.  Wouldn't that eliminate the need for soft proofing?

Post #3 describes the status of this process.
Logged

Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....." http://www.luminous-landscape.com/reviews/film/scanning_workflows_with_silverfast_8.shtml
Czornyj
Sr. Member
****
Online Online

Posts: 1304



WWW
« Reply #42 on: July 11, 2010, 04:48:14 PM »
ReplyReply

Quote from: LarryBaum
Actually I guess you are correct. The 2690^2 version does appear to let you calibrate the white point of the sRGB mode, the original version did not.

It still only has 1D 12bit LUTs and not 14bit 3D though so hah!

(without a 3D LUT green often looks like orange and sometimes white switches with black   )

Without 3D LUT you can only calibrate RGB gain and TRC - but you can't precisely calibrate chromatic coordinates of displays colorants. A 3D LUT display can simulate virtually any color space that's covered by its gamut, and considering that it uses 14(16)-10bit path one can expect, that the simulation might be better than standard simulation done by application CMM, that can only use 8bit.
« Last Edit: July 11, 2010, 04:53:19 PM by Czornyj » Logged

Marcin Kałuża
WombatHorror
Sr. Member
****
Offline Offline

Posts: 296


« Reply #43 on: July 15, 2010, 04:35:06 PM »
ReplyReply

Now I'm not entirely sure how the old 90 series did it but it seems like the 3D Lut means that the PA series can be calibrated using few data points.
For instance you can bring up multiprofiler and go into custom mode edit and then using SVII measure the primaries and white point and then you are set. Just dial in chromaticity coordinates in MP until the SVII patch reading for the primary matches reference, do the same for white balance and then just tell it what gamma you want and what brightness and you are set. The 3D Lut handles the interplay of everything at once and from those simple inputs (plus a few factory values) delivers the proper secondaries with proper luminances and the entire gray-scale and everything else.

Seriously in literally just two minutes I got the below (see next post) results for sRGB primary and gamma 2.2 (for blu-ray/TV) mode in multiprofiler.
I managed to use an i1Pro for this so I could get some sort of extra sanity check and see how the results read out and be able to look at how saturation tracks too, etc.

reference is:
R .640,.330
G .300,.600
B .150,.060
W .3127,.3290

to get a little closer to that (according to i1pro at least) I entered for the coordinates:
R .640,.333 (read back as .640,.300)
G .300,.596 (read back as .300,.600)
B .150,.056 (read back as .152 there is no way to get it better the display doesn't quite make .150 no matter what,.060)
W .319,.332

you can see that the i1pro didn't suggest much change other than for Wx, Gy and By and even then only Wx was more than .005 change

for reference when I used the NEC puck I had to enter:
R .642,.333
G .300,.597
B .150,.056
W .319,.334

so you can see that the i1Pro and the NEC puck (USED IN SVII only, outside of SVII it reads WAYYYY differently than the i1Pro or DTP94b pucks) were in remarkable agreement, and aside from WB both were in reasonably close agreement with factory settings.

I think it is pretty clear that NEC really does apply a custom calibration to the i1D2 NEC puck and also provides highly customized tables for each of their displays for it since an off the shelf puck has something like an average 8dE and up to 18dE copy to copy variation according to one study and it seems like considering that it would be quite unusual that my NEC puck and i1Pro would be so close if they didn't custom calib each hardware unit plus provide a table for each monitor on top.

The NEC puck and the i1Pro actually provided more closely matching results both reading the primaries one monitor than the NEC puck reading the sRGB primaries of two different NEC PA241W monitors. Are the factory presets any better than the puck readings then??? perhaps, but I wonder a little now. Not sure about the native primary readings, maybe different there.




Of course keep in mind even with compensation on from spot to spot there will be some variation across the screen and all the measuring units have some error and so on so none of this is to the nth decimal point in reality. So the reality that things are never nearly so close as we think we ahve gotten them with out calibrations and charts and going for every little detail.




I definitely do see metamerism issue comparing wide gamut to sRGB displays though, they simply don't look the same. To my eye the wide gamuts with no correction actually appear to match real life (suing a color check chart) a little bit better, but who knows. Using the metamerism toggle makes white and quite a few shades look more similar between the two display types (while making certain other shades seem even more off) but they still don't look the same.

It is a tricky issue and a little troubling since it seems like images edited on a wide gamut in regular mode won't quite look the same if you switch to an sRGB-gamut monitor, at least for some people and possibly the degree of difference varies a lot from person to person. A tricky business.

In some ways I wonder if it not the sRGB gamuts that are more troubled though since I could swear that a real-life CC chart more closely matched the look on the wide gamut although it didn't exactly look the same in either case.

I wonder what a reference CRT looks more like, a wide gamut CCFL LCD, wide gamut+metamerism toggle CCFL LCD, CCFL sRGB LCD, LED sRGB LCD or RGB LED sRGB LCD? hah.
Some say the whole tri-stim color management system is basically a failure when it comes to getting different display technologies to deliver identical looking results to the eye.

Oddly the srGB monitor calib with DTP looked closer to the NEC+i1orNEC puck+metamerism than the i1pro used on both and the i1pro sRGB seemed perhaps even a bit farther away from a real life color checker chart comparison. (The NEC puck on the sRGB monitor using third party software made an awful looking calibration it made it look really red it sort of did what the i1pro did compared to the dtp only did that about 100x as much) Probably need a 1nm $$$$$$$$$$$$$$$$$ spectro to get a better clue, I think some display types can have very spiky spectrums that might throw off some colorimeters and even a 10nm resolution i1Pro.
« Last Edit: July 15, 2010, 05:03:45 PM by LarryBaum » Logged
WombatHorror
Sr. Member
****
Offline Offline

Posts: 296


« Reply #44 on: July 15, 2010, 04:50:37 PM »
ReplyReply

Since the i1Pro is supposed to be reasonable ok on some wide gamut monitors (although i'm sure it might still miss some sharp spectrum spikes) here is what I measured from the sRGB gamma 2.2 mode I programmed in multi-profiler (in all of about 2 minutes):


Tracking of primaries and secondaries saturation (on top) and saturation hue shift (on bottom) with respect to measured primary and secondaries:


Tracking of primaries and secondaries saturation (on top) and saturation hue shift (on bottom) with respect to theoretical exact sRGB specs:


Saturation luminance tracking of primaries:


Gray-scale tracking:



« Last Edit: July 15, 2010, 04:55:22 PM by LarryBaum » Logged
WombatHorror
Sr. Member
****
Offline Offline

Posts: 296


« Reply #45 on: July 15, 2010, 04:58:53 PM »
ReplyReply

continued....

Color temperature tracking:
 but whether the metamerism issue tosses it who knows?
« Last Edit: July 16, 2010, 02:47:45 AM by LarryBaum » Logged
WombatHorror
Sr. Member
****
Offline Offline

Posts: 296


« Reply #46 on: July 31, 2010, 01:21:02 AM »
ReplyReply

It seems like maybe the monitor drifts the first few weeks until the factory pre-set white point is closer to D65?
perhaps they start it a bit off so that it tends to settle in closer for the long haul??

above I needed to set: W .319,.332 to get i1pro measured .313,.329 and now I need to set only W .315-.316,.329
granted the WP does vary across the screen but I tried to test areas all around the spot
in the end I also went with what seemed the most average patch to adjust
these were at compensation level 4

short time monitor and instrument drift and slight changes spot to spot on top of inherent instrument systematic error and uncertainty do make it tricky to every really figure everything out
Logged
Steve Weldon
Sr. Member
****
Offline Offline

Posts: 1441



WWW
« Reply #47 on: July 31, 2010, 03:44:35 AM »
ReplyReply

Quote from: LarryBaum
It seems like maybe the monitor drifts the first few weeks until the factor
short time monitor and instrument drift and slight changes spot to spot on top of inherent instrument systematic error and uncertainty do make it tricky to every really figure everything out
My 2690uxi2's certainly drifted significantly until about 500-600 hours of use.. now they've settled down.

Considering the physical location of the back lights the screens will always vary from one location on the screen to the next.. which brings us back to square one.. no matter which puck, screen, etc, if you get it close enough to match your prints in a closed system its about as much as you can expect.. and the sRGB emulation mode helps keep us "close" but not perfect across the web.

There's always going to be some variance and nothing makes this more clear than my (2) LCD2690uxi2's side by side, one has 1000 more hours than the other (ordered at the same time, delivered 5 weeks apart), one day the white point visually matches, the next day there's a slight difference.  When there's a difference which monitor is correct?  Why does it change?  Of course those are rhetorical questions but it just goes to show when dealing with the finer aspects of white points there's no perfect system out there.  Though.. I'm still tempted to pick up a few of those 21.5" RGB LED backlit NEC's that are now at rock bottom prices and give them a try out of pure curiosity..
Logged

----------------------------------------------
http://www.BangkokImages.com
WombatHorror
Sr. Member
****
Offline Offline

Posts: 296


« Reply #48 on: August 01, 2010, 01:17:40 AM »
ReplyReply

Quote from: Steve Weldon
My 2690uxi2's certainly drifted significantly until about 500-600 hours of use.. now they've settled down.

Considering the physical location of the back lights the screens will always vary from one location on the screen to the next.. which brings us back to square one.. no matter which puck, screen, etc, if you get it close enough to match your prints in a closed system its about as much as you can expect.. and the sRGB emulation mode helps keep us "close" but not perfect across the web.

There's always going to be some variance and nothing makes this more clear than my (2) LCD2690uxi2's side by side, one has 1000 more hours than the other (ordered at the same time, delivered 5 weeks apart), one day the white point visually matches, the next day there's a slight difference.  When there's a difference which monitor is correct?  Why does it change?  Of course those are rhetorical questions but it just goes to show when dealing with the finer aspects of white points there's no perfect system out there.  Though.. I'm still tempted to pick up a few of those 21.5" RGB LED backlit NEC's that are now at rock bottom prices and give them a try out of pure curiosity..

and toss metamerism in too boot and it's hopeless


wide gamut LCD displays just don't like standard gamut LCD displays to me, with or without metamerism comp/10 degree comp and neither looks the same to me as a real life color checker chart (although the wide gamut monitors actually look, to my eyes, closer to the real life colors than the standard ones not sure if that would be the same for most of if for some reason my eyes see a more natural color out of wide gamut monitors than standard)



but more seriously i really wonder if they realize that the screen will drift a lot the first 200-400 hours and purposely undershot the factory white point calibration. It's at 260 hours now. It will be interesting to see if it settles in perfectly at 400 and then stays there for a while.... (not that it matters so much if you have a puck i guess but I wonder if the reviews that all also got it too cool to start should've waited a while to measure first)
Logged
RobWalstrom
Newbie
*
Offline Offline

Posts: 12


« Reply #49 on: September 01, 2010, 02:49:30 PM »
ReplyReply

Hey Larry,

I just got my PA271W and tried to use the sRGB mode for non-color managed applications. The sRGB preset seems so dark at 80 cd/m. I believe this is part of the sRGB standard, but is it so low because the standard is from the days of CRTs?
Logged
Czornyj
Sr. Member
****
Online Online

Posts: 1304



WWW
« Reply #50 on: September 02, 2010, 02:27:33 AM »
ReplyReply

The sRGB preset seems so dark at 80 cd/m. I believe this is part of the sRGB standard, but is it so low because the standard is from the days of CRTs?
Of course.
Logged

Marcin Kałuża
Pages: « 1 2 [3]   Top of Page
Print
Jump to:  

Ad
Ad
Ad