David,

So let me get this straight: I need to duplicate your measurements but you won't duplicate mine. Moreover, you don't even comment on why my measurements and my conclusions would be in error in a meaningful way. You even haven't said one single word about the gamma issue that I have brought up.

This is not going to be a one-way street. I'm very willing to duplicate your methods, but only if you duplicate mine.

In order to duplicate your methods I am going to need a little more help and information from you. What kind of hardware and software do I need and what kind of tests and calculations do I need to perform? Can you refer me to a website or book to learn more on this method?

Frans Waterlander
pixographer/printer

Frans,

The gamma issue is not pertinent to this because you yourself stated that you are using the 2.2 gamma setting that I am and is virtually standard. And gamma does not affect color content just lightness and darkness relative to scale.

Besides it is a test procedure almost identical to what is requited to linearize black and white reproduction, which I have done on numerous occasions.

>>This is not going to be a one-way street. I'm very willing to duplicate your methods, but only if you duplicate mine.<<

What is this some kind of child's playground game. This is not a contest, and if you think it is, game is over before it starts.

All that is required is a colorimeter that has software support to measure and read out on-screen color in LAB/RGB number values. I use a ColorVision Spyder2 Pro and its latest version software. The target is a TIIF file of an ICC IT-8 color target. I am not sure if a file can be downloaded from the ICC web site, but a Google search should provide a source readily. The procedure is simple, open the IT-8 file in Photoshop so it is in Adobe RGB (1998) color space and read the color patch values on screen and record the numbers for each patch coordinate. Repeat the process for both a 5000K and a 6500K monitor setting and profile.

David,

This is not a game. I want to understand why your findings are so different from mine. I'm not sure what you mean by LAB/RGB number values. I have a ColorVision Spyder with OptiCAL software and in colorimeter mode it measures the following data:
- CIE XYZ,
- CIE xyY and
- CIE Luv.
Which one(s) should I use?

Frans Waterlander
pixographer/printer

Frans,

CIE values are different modes of LAB values, which can be directly translated to RGB values. The translation can be done directly via Photoshop's Info and Color dialogues. It really does not matter which is used, because the indicator is differences or lack of difference. Personally I prefer to work with RGB values because it is a familiar "language" with number values I can relate to real colors.

I have received a number of messages from Forum participants that the tone of this "conversation" has taken a turn that is not helpful or constructive; indeed, many find it distracting and less to do with sharing information and more personal than we intend this to be. I am not looking for any apologies or change in opinions, but a more objective discussion of the issues here that will be more helpful to others than what has happened here.

David,

My late response is due to a 6 week RV trip to the east coast. If my memory serves me well, you found measurement values to be within about 1% when you compared results between a monitor calibration of 5000K and 6500K and you suggested I also run those tests. I did and that's not what I found when I measured 15 color patches of an IT8 target on-screen with a ColorVision colorimeter. When recalibrating the monitor from 5000K to 6500K, the R values decreased an average of 10%, the G values increased an average of 2% and the B values increased an average of 28%, consistent with my observation that the monitor image colors shift significantly towards the blue when changing the monitor from 5000K to 6500K, a behavior that, in my opinion, many digital darkroom users will confirm.

So this is yet another observation that tells me that Photoshop does NOT keep colors virually constant when making a significant change to the monitor color temperature as you maintain. Let me recap these three observations:
1) I see a sifnificant shift towards the blue when changing the monitor color temperature from 5000K to 6500K
2) I measure about 5000K color temperature of a grayscale image when the monitor is calibrated to 5000K and about 6500K when calibrated to 6500K
3) there is a significant shift towards the blue when measuring 15 color patches of an IT8 target with a colorimeter when changing the monitor calibration from 5000K to 6500K

You said before "I have already specified precisely what tests must be done, and that I have done, to determine precisely what a working space profile in Photoshop does in terms of what is displayed as a color image on a calibrated and profiled display. So why don't you just do it so you have something more than opinion to go on?" So now that there are three observations to go on, I would appreciate your comments.

Frans Waterlander
pixographer/printer

Frans,

That was not the test I described, so of course you obtained results that validate your pre-determined conclusion.

David,

Since we are still - through private emails - trying to understand and validate what each of us has been measuring, could you please hold off on your judgment?

Frans Waterlander
pixographer/printer

David,

It is quite obvious from our private email exchanges that you cannot explain how exactly your measurement method works or how it can possibly be as accurate as my method of direcly measuring Photoshop images on-screen. It is also very clear that you are not willing to engage in a constructive dialogue on this and other issues, so I won't bother you any longer.

Frans Waterlander
pixographer/printer

It's been quite a while since this thread started, and it seems to have veered significantly... however, I wanted to come back to the original subject for a minute...

i just moved to a hurricane zone, and have finally decided to get serious about scanning in my negatives. I have purchased a Nikon Supercoolscan 4000 (and i'm sure I'll be asking questions about that later)

I've managed to scrape together enough funds for a sub $1000 monitor, and am looking for the best value in a 23/24 inch. Has anything supplanted the Samsung 244T? I see it regularly available for under $900, and all the reviews I read loved the quality, but hated the initial $1500 price tag.

I'll be hooking this up to a powerpook g4 with an ATI 9700 card (128 MB). i have the gretag macbeth Eye one display calibrator (the original, I think a new one came out about a month after i bought it...)

While I obviously need a monitor that can be accurately 9and easily!) color calibrated, I would like something that can handle video and games as well. the samsung seems to be a great Jack of all Trades.

I would love to hear any advice, or alternate recommendations.

Thanks!

I follow this market closely, and there are new offerings coming out every so often, but in terms of size and format with your requirements to display video and play games and at under $1000 I have not seen an LCD display that also effectively supports digital photography that competes with the Samsung Syncmaster 244T.

David -

Since I am about to purchase a new monitor this has been a most interesting thread. The Dell 2407 was considered but reading your comments and others on the web, the 244T seems worth the extra dollars.

A question relating to graphics cards. In addition to photography I do a lot of video editing. I am running a 2.40 gigahertz Intel Pentium 4 with 1536 Megabytes installed memory. However I know my current Nvidia GeForce FX 5200 graphics card is probably not going to work with the 244T. Do you or anyone out there have any recommendations for a graphics card that will do the job but without spending hundreds of dollars?

If your video card is one with just a 9-pin D-sub analog connector it is essential to replace it with a video card that has a DVI digital connector. LCD displays function at a great disadvantage, if at all, with an analog connection.

To obtain the most effective 2D performance supporting good Photoshop response an ATI Radeon 9600 is good and the Pro 256MB VRAM model is best.