Monitor Calibration

[-default]

Well, it finally happened.  In spite of my contention that hardware devices are not necessary for calibration, in the name of research, I plunked down $160 for a Color Eye One 2. I'll have more to say about this device later - so far I'm reasonably well impressed with the software, and the device seems to work well.

In the process of checking out all the various options the software offers, I've come up with a test image that I'd be interested in what it looks like to others in Photoshop.  The image consists of several gradients.  Do you see any obvious off-color areas in any of the gradients?  In particular, is there a purple tint to the center of the blue-white gradient?  If you own Curvemeister, use the stand-alone hue clock to look at the image, and see if the pointer is midway between blue and magenta.

Most browsers will not show the sort of color You'll need to look at the image in Photoshop to see this effect.

Mike

[mikemeister_admin]

Mike,
  I checked out the test image on my computer. I do not see a purple tint in the centre of the blue-white gradient. I used the hue clock and it the pointerdoes not go to the centre between the magenta and blue. I also get the same readings on photoshop as I get with the browser, also no noticeable off-colour areas.
  I will be interested in your thoughts about the device.
                  Barry

[ggroess]

Here are some screen shots the Black dots are the mouse location for the Hue Clock sample

Greg

[mdavis]

Are you saying that the purple tint is after a monitor calibration?  Also, are you using an LCD or CRT?  My Viewsonic A90f+ (a fairly cheap CRT monitor) shows the bands nicely.  I check calibration weekly using only Adobe Gamma.

Again, as a fan of color by the numbers, if it looks good and the numbers confirm balance, then it prints as it looks.  Have you tried printing the chart to see if you still see a tint?

[-default]

Greg - thanks.  Your screen shots do show a very pure blue going out to the screen.  By comparison, on my notebook that I'm having trouble with, the hue clock is about halfway between blue and magenta.

I've made some progress on this issue, and will be summarizing it soon.  Briefly, the problem relates to screen angle.  In the room where I normally use my notebook, I have it angled to avoid a ceiling light reflection, and this throws off the perception of deep blues.

Mike - the purple banding only shows up in photoshop, not when printing the image or viewing it outside of photoshop.  The monitor profile is rather extreme and I'll post a summary of what I found out as soon as I've sorted it all out.

Mike

[mikemeister_admin]

Mike,
with interesting reading this site about monitor Calibration.
In the past time I have spend hours, hours, I think in the totality weeks about calibrating my 2 monitors and making them equal.
I have reading much about  calibration and I have tried every solution.
But no , I never succeed and was not satisfied with one monitor , and much more also not to try 2 monitors equal.
So, explain to me and others why you think it is not necessary to buy a calibrator.
You know , I have in the meantime Spyder2Pro and this works satisfying on both computer-screens.

---

In the past I have complained about that the hue-clock gives not the precise direction.
I see that Greg Groess have the right hueclock pointing  the right way.
I have tried this also, and indeed the Hue clock points exactly to B, when I look to test1.jpg  in the Browser.
But downloading gradient6.png and importing in Photoshop elements 4.0 gives other results.

I put the pictures here!
Do you know what could be the course of the differences.?

frits

P.S .first picture is looking the picture in the Browser.
Second picture is of PSE 4.0

[-default]

Mike,
with interesting reading this site about monitor Calibration.
In the past time I have spend hours, hours, I think in the totality weeks about calibrating my 2 monitors and making them equal.
I have reading much about  calibration and I have tried every solution.
But no , I never succeed and was not satisfied with one monitor , and much more also not to try 2 monitors equal.
So, explain to me and others why you think it is not necessary to buy a calibrator.
You know , I have in the meantime Spyder2Pro and this works satisfying on both computer-screens.

If you can get a good neutral gray ramp from your monitor, I believe this is satisfactory for most applications.  That said, I have no problem with using a device to calibrate. 

But many people will use a calibration device, and have no frame of reference to determine whether it has worked correctly or not.  Your example of getting two monitors to match is a case in point - given that they do not match, one of the two must be wrong.  Yet if you had only one of the two monitors, you would probably assume that it was correctly calibrated.

In the past I have complained about that the hue-clock gives not the precise direction.
I see that Greg Groess have the right hueclock pointing  the right way.
I have tried this also, and indeed the Hue clock points exactly to B, when I look to test1.jpg  in the Browser.
But downloading gradient6.png and importing in Photoshop elements 4.0 gives other results.

I put the pictures here!
Do you know what could be the course of the differences.?

This is a limitation of the hue clock.  It relies on the RGB values from the display for its calculation.  Since Elements and Photoshop use the display profile to modify the RGB values, you will get different hue clock values when you examine an image from within either of these programs.  This will be particularly true when viewing saturated colors, as appear on the test gradient image.

I've thought about a solution, and probably the best one is to have the Hue Clock use the display profile, perhaps with an option to disable this feature.  But the big problem is that if the application is not color aware, then the Hue clock should probably not use the display profile.  To address that I could have the Hue clock determine whether Photoshop or Elements is running, and use or not use the display profile on that basis. 

In practice the amount of the change is so small that I wonder if this would be worth it.

Mike

[mikemeister_admin]

If you can get a good neutral gray ramp from your monitor, I believe this is satisfactory for most applications.  That said, I have no problem with using a device to calibrate. 

But many people will use a calibration device, and have no frame of reference to determine whether it has worked correctly or not.  Your example of getting two monitors to match is a case in point - given that they do not match, one of the two must be wrong.  Yet if you had only one of the two monitors, you would probably assume that it was correctly calibrated.

Mike,
that's to easy said. Only a neutral gray ramp. ? what about gamma and brightness of the screen.?
So, I thought it also in the early days, I can try the suggestions which to find on www.
But, I never get the two screens equal.
So, my reference is now Spyder2Pro. 
In my opinion it measures no color at all, but  light densities.
It is in essence a electronic device/chip what measures the light through a opal glass.
This is translated in densities or candelas.
The program generates indeed different colors and grays, but it all is measured an translated.
So also the gamma is generated.
The en-result is a profile specific for that videochart and screen.
After calibration you must not change anything on the videochart or screen.
A grey ramp is to easy, in my opinion.
So, it would be nice if anyone of this group has a special reference-chart, with which he can compare and measure this in  photoshop or PSE and together with curvemeister which gives exact ciphers.
So, a reference-chart you specify or what you can find on www what you declare as basic.
Then you measures each quadrant with curvemeister  and give us the measurements.
So, everyone without a spyder2pro or other calibration-device reaches about the same results with curvemeister,
when he does the same.
Is that a good idea?

Frits

[mikemeister_admin]

This is a limitation of the hue clock.  It relies on the RGB values from the display for its calculation.  Since Elements and Photoshop use the display profile to modify the RGB values, you will get different hue clock values when you examine an image from within either of these programs.  This will be particularly true when viewing saturated colors, as appear on the test gradient image.

I've thought about a solution, and probably the best one is to have the Hue Clock use the display profile, perhaps with an option to disable this feature.   But the big problem is that if the application is not color aware, then the Hue clock should probably not use the display profile.  To address that I could have the Hue clock determine whether Photoshop or Elements is running, and use or not use the display profile on that basis.   

In practice the amount of the change is so small that I wonder if this would be worth it.

Mike

Mike,

perhaps you are right.
But I have not investigate it, or the curvemeister prgr has the same deviation.
The Browser uses the screen profile and the hueclock measures right.
After loading photoshop or PSE , these prgrs uses a extra working profile., and translates the results to the screenprofile.
I could not change that in PSE. But I could say to PSE, "make no use of colormanagment".
But the results with the hue-clock are the same.
Strange that Photoshop and PSE give different results with the HUE clock.
The amount is Small, you said, but I think working with skincolors, it is important.

Frits

[ggroess]

Guys...not to beat a dead horse or anything but Monitor calibration is half of the problem.  It can look good on screen all day long but what do the prints look like? 

The issue seems to me to be double edged...make it look good on the monitor and then make the prints look good as well.  If you send prints out to a pro processor you will always get reasonable skin tones.  BUT...and this was the gotcha for me...you have to stick to a color profile supported by the processor.  I was lucky and the company I use supports both AdobeRGB and sRGB.  I have had some prints come back crappy but it is usually my problem and not the printer.

I also think that unless the color is severely off say 10 Yellow or 10 Cyan, most people would never know.  I have to agree with Fritz that the Gamma is more trouble than the color.  I find most of my problems fall into exposure compensation...the darn camera can see a wider range of exposure than film...go figure...

so for me...so far...I use a good greyscale for gamma adjustment, I check the whites against a piece of white photo paper and I watch my exposure...

But hey...thats just me....
Greg