Monitor Calibration

[Abitconfused]

I recently calibrated my monitor using Spyder4Elite (and the latest software update). The results showed that my monitor was displaying 73% of the sRGB colour space. This I did not like. So I loaded an earlier .icm profile and ran Spyder4Elite's colour accuracy test again and got 98% of the sRGB colour space. How can my monitor simultaneously be at both 73% and 98% of a color space?

Also... The Delta-E rating for yellow is 9 and seems to be bad (see photo). What does this mean exactly?
Could not upload photo

[ST1]

Ed have you checked what your monitor manufacturer states for its sRGB performance.
If the second (older profile) you used was derived from your Spyder4Elite then I suspect that the profile you created that only gets 73% sRGB must have errors in it.


Sent from my iPhone using Tapatalk

[Manfred M]

Unfortunately, I don't know the intricacies of the Datacolor (Spyder) hardware and software, as I have been using the x-Rite tools for years (i1 and ColorMunki). The fact that it is reporting less than 100% sRGB is quite strange as virtually every modern computer screen should easily display that colour space.

A few thoughts as to what might be going on...

1. These devices do two things, they perform a calibration function - brightness, contrast as well as a profiling function, i.e matching / correcting your specific output to a known "correct" value. The process is not 100% reliable and if the sensor is not sitting properly on the computer screen's surface, faulty readings can occur, especially if there is a fairly bright light falling on the screen's surface. Re-running the calibration / profiling operation will usually produce a correct result.

2. Certain computer screens can only be calibrated with specific hardware - the some of the Dell Ultrasharp screens must be calibrated / profiled with the i1 using customized Dell software (the lookup table (LUT) cannot be accessed by other hardware / software solutions). Some of the higher end Eizo and ViewSonic screens also ship with custom software.

3. Some lower end screens using TN (twisted nematic) technology do not natively display true 8-bit colours, but rather are native 6-bits and use dithering (i.e. rapidly pulsing the pixels) to emulate 8-bit colour. I could certainly see how this might fool the calibration / profiling hardware and software.

[Abitconfused]

Thank you. ASUS mentions 97% of the sRGB colour space. So something is consistently in error with the Spyder but what concerns me even more is the Delta-E rating for yellow (see photo) which seems odd.

[Manfred M]

Is this a laptop? Touchscreen?


The reason I ask is that the colour reproduction of my ASUS laptop is terrible. The unit tended to the blue (i.e. the complement of yellow), so the blue really needed to be turned down to get the yellow to look right. I rework all my images after I edit on it (and yes, the screen has been profiled and calibrated).

[Abitconfused]

Its the latest in desktop with Nvidia 4GB video card. But, yes, there are many monitors out there that should be kicked over the back fence into the alley. This is why Apple did/is doing so well versa the PC. We need a Steve Jobs for the PC world...well town.

[Manfred M]

This is why Apple did/is doing so well versa the PC. We need a Steve Jobs for the PC world...well town.

So far as I understand it, Apple just announced it is getting out of the standalone computer screen business and is recommending LG branded screens.

If you are paying Apple prices, there are a number of screens out there (Eizo, ViewSonic, NEC and one specific Benq screen) screens that are considered superior to the current crop of Apple screens. Unfortunately, if you buy an iMac, you are stuck with the Apple screen.

[Abitconfused]

Yes, I think I will next invest in a ViewSonic.
I discovered that the Delta-E scale is from 1-100 with 1 to 2 representing color differences the eye is unable to detect and 100 being maximum brightness...luminance? "What does Delta E number mean? The higher the Delta E (ΔE), the further away the colour is from the true hue, using CIELAB. Perfect colour has a Delta E of zero, although this cannot be detected using the human eye. The minimal detectable difference is between 1-2.5 Delta E. Without a fully-realised colour management system, it is very unlikely that monitor calibration can achieve this level of performance for primary and secondary colours. Most virtual displays have a full complement of gray scale settings, so for white, calibration can often meet this standard. If a Delta E number is less than 1 between two colours that are not touching, it is barely perceivable by the average human observer. A Delta E between 3 and 6 is usually considered an acceptable number in commercial reproduction, but the colour difference may be perceived by printing and graphic professionals. (Note: Human vision is more sensitive to colour differences if two colours actually touch each other)."

[Ken MT]

After reading that I am a bit confused?

[Shadowman]

I'm slightly familiar with that instrument, I use ColorMunki and what you should be probably considering is the changing light in the room you have your monitor. This link is for the next version of spyder and I wonder if your monitor/colorimeter is resetting based on light at that particular moment. One thing I would like to do is to have my monitor in a room where light is constant day and night, my ColorMunki monitor calibrator will change my monitor settings based on the lighting measured throughout the day. So far I haven't had any issues with any prints, however I will say that images processed on another laptop positioned in another room look totally different on the calibrated monitor, however when I print the image; using Photoshop Elements set so that that software manages colors, my prints come out meeting my expectations.

You might ask "why am I processing images on another monitor which isn't calibrated by the ColorMunki, the processing monitor is fairly new and I haven't seen any issues yet with color management. I do plan to calibrate the processing monitor in the future.

http://spyder.datacolor.com/portfoli.../spyder5elite/

[tao2]

Hi Ed,

before ye do any calibration, ye should make sure that both the monitor and the graphics card are set at factory default. Otherwise , ye will get skewed results.

[dje]

Hi Ed

There's something fishy here!

It sounds like either your monitor is playing up or your Spyder is faulty. I have a Spyder 5 PRO (not the Elite so I can't measure Delta E's). I think you should be getting much better consistency than 73%-98% of the sRGB coverage. Have you redone the calibration two or three times?

I assume you are looking at a screen like this



Towards the top right there is a box where you can select past profiles to analyse. If I select several of these in turn, I get very consistent results. It's probably a silly question but you haven't inadvertently selected the NTSC comparison for your latest test have you ?

I'd keep an eye on it over a period of time (say a few weeks).

Dave

[Abitconfused]

My Spyder now consistently shows 73% but earlier .icm prove it is showing 98% sRGB. I think it is broken.

[Manfred M]

My Spyder now consistently shows 73% but earlier .icm prove it is showing 98% sRGB. I think it is broken.

Have you considered that while your earlier profile might have shown 98% sRGB, it might have also suggested that while the gamut of the screen was quite compliant with the sRGB spec, the actual colours were off? In fact I am suspicious of the 98% reading. Most modern screens are 100% sRGB compliant and the better Adobe RGB screens are 99% and better compliant with that colour space.