A method to better calibrate an LCD monitor with eye-one (i1)

I am happy to report I have discovered a method for a more accurate calibration for the LCD screen (monitors, displays), using Gretag Macbeth (now an X-rite company) i1-Match series of color management solutions.

I haven't written it up yet in a nice way, but the idea is that the reference file with 49 RGB patches, while good enough for CRT calibration and a calibration with the profile creation that uses a conversion matrix, the data acquired is not enough for the LUT table method that is preferable for LCD screens.

There is a way to hack i1-Match into using reference data that contains either 99 patches (Gretag Macbeth LCD ref file) or even more if you desire. All is needed is a couple of copy paste moves and a two file renames.

Will publish it later, stay tuned.

Colorimeter measurement supports a claim of the iPhone yellow screen fix with the 5A347 release.

There were reports regarding the iPhone 3G screen being too yellow.
It looked absolutely fine on the iPhone I got.

Later still many have claimed [AppleInsider] that the restore to the 5A347 release from 5A345 "fixes" the "problem".
So, I used my trusty Gretag Macbeth i1 spectrometer to see for myself.

First I would measure the white and gray fields on the iPhone with 5A345 set at the medium brightness with the auto brightness adjustment disabled, then I performed the restore to the 5A347, using the iTunes and repeated the measurements under the same conditions.

To begin with the original measurement showed the white point set to 5500K, which is awesome, and as many of you know is a standard for the many industries as far as the color management is concerned. CRI of 75 with white point brightness of 246 lux.

After the software restore the numbers were as such: 5755K, CRI = 74 and the brightness of the wight point at 241 lux.

For those of you unfamiliar with the color temperature concept, 5500K is yellower then 5755K.
The color curve showed no change in the blue part of the spectrum but some (very minimal) tweaking in the green and red, significantly more so in the green. This goes well with the whole change in the yellowishness theory.


So here you go, real, hardware measured data.

Subjectively though, I could not see any difference and I found the color rendition to be perfect on the original release anyway.

Measurement methodology was such: two measurements (averaged) for each brightness, 8 measurements total, with a calibration of the eye-one colorimeter after each sampling.

Questions? Comments?

Practical side to the AdobeRGB versus ProPhoto RGB discussion.

After reading the mentioned article I have decided to find all the truth about it myself. I used an image I took last year in San Juan, Puerto Rico as a test dummy.

to purchase this print please click here

I converted this image from a CRW file format as a 16-bit ProPhoto RGB image using the Adobe Photoshop Camera Raw. Then I saved it as a PSD file and resaved as a copy after a conversion to AdobeRGB using the Adobe ACE with relative colormetric and a black point compensation.

After opening both 16-bit files inside the ProPhotoRGB color space and pasting the AdobeRGB tagged file as a layer into the ProPhotoRGB tagged image I switched a blending mode of the pasted layer to the Difference mode, thus getting the following image.

click to see a Flickr! of this image

It is granted that the original image does not represent an ideal test subject as it can't contain all of the colors involved. It is, however a real world image, and it does show the actuall application implication of using different profiles when converting RAW images and chosing a workspace.

Now, about the real world application.
The ProPhotoRGB version of the image and AdobeRGB version of the image were both printed on the Canon i9900 printer, using a custom profile for the paper built with the Eye-One Match 3 from GretagMacbeth hardware and software solution. This is how the custom printer/paper combination profile compares to the AdobeRGB colorspace.

Shaded shape is AdobeRGB, mesh is a printer profile.

As you can see many of the colors fall out of the range of the AdobeRGB color space. Yet all of this colors are contained within the ProPhotoRGB as it is a much larger color space.

After letting the prints dry I have evaluated them visually under the 5000K light and under the available daylight. The prints were very similar, but the ProPhotoRGB print had slightly "punchier" orange in the color of the building. The AdobeRGB print had more magenta in it and was further from the image rendition on the calibrated CRT screen then the ProPhotoRGB print.

Interestingly enough, when both versions were viewed on the screen there was virtually no difference in the color appearance. This is probably due to the limited capabilities of the CRT color reproduction.

As an added bonus I have printed an 8-bit image version but could not see a quality difference from the 16-bit version with a naked eye. This, however, might be due to the nature of the image subject. When using the difference blending mode procedure described above as well as boost in saturation and brightness of the resulting image the minor difference was descovered. Here is the map of it.

please, click for a Flickr! of the image

So, here you have it. There is an undeniable destinction in theory, that, while having a practical realization, not absolutely critical as far as ink-jet printer concerned. Also, color space proved to be more affecting the final output then the 8-bit versus 16-bit difference.

I will follow up with the instrumental measurment analysis of the different prints of the test image in the near future.

What a difference the green ink makes.

So. It was a time for me to change a green ink tank. I decided to do a little experiment, to see how much of a difference in color fidelity would printing without a green ink tank make. Granted, I am not sure that the printers driver was aware of the fact that the green ink tank was absent.

The difference is obvious. Look closely at the lemon tree leaves and a fruit.

With the green ink tank.

Without the green ink tank.