iMac calibration woes, i1 display pro

NeilA1975

NeilA1975

Suspended / Banned
Messages
5,026
Name
Neil
Edit My Images
No
I've recently purchased my first imac, a 27 inch late 2013 i5 model which I'm really enjoying getting to grips with.
One aspect though which is frustrating me is the less than desirable calibration performance of the panel, which indicated by the i1 software suggests the white balance is still too warm (the RGB channels in the 2D graph shows the red channel dominating at all triplet levels, above the green and blue which are equal) even though 6500k is selected and is apparently achieved. I resorted to upping the CCT target to 7000 in order to achieve a better result, this brought all of the RGB channels much closer together and gives a better overall appearance even though the achieved result states 6900K!!

I must say I never had this type of issue with my Dell U2713HM which is probably a superior monitor, but the behaviour of the i1 software on the Mac seems to contradict itself in the context of the results.

Anyone else experience this type of problem?
 
Recently purchased a 27 iMac late 2013 model.
Using i1 profiler and an i1 D3 what settings should I apply?
I've already measured the native white point at around CCT 6900k, moving away from this towards 6500k skews my 2d graph. Native gamma also measures 2.3 not the desired 2.2. The lack of monitor controls especially the RGB channels leaves me with this dilemma.
Appreciate your input.

Thanks
Neil
 
NeilA1975 said:
Recently purchased a 27 iMac late 2013 model.
Using i1 profiler and an i1 D3 what settings should I apply?
I've already measured the native white point at around CCT 6900k, moving away from this towards 6500k skews my 2d graph. Native gamma also measures 2.3 not the desired 2.2. The lack of monitor controls especially the RGB channels leaves me with this dilemma.
Appreciate your input.

Thanks
Neil
Click to expand...

Daily bump - can anyone help??
 
I've used the i! on several macs without any problem. The latest being a late 2015 iMac with the P3 display. I did find though that a 2009 MacBook Pro had some problem getting all three in line, but I was using ColorEyes software in that case. I have a 2013 rMBP which does run the X-Rite software and that's fine. My 2015 iMac does show slight splitting of the RGB curves but that's about the thickness of each line. The 2009 machine is a little wider but nothing to write home about. All three on both machines are linear all the way to the top of the chart

Are you letting the software set the brightness level, and if so what have you set it to. Also what is the final verification page stating it has achieved as far reaching the aims you have requested
 
Chappers said:
I've used the i! on several macs without any problem. The latest being a late 2015 iMac with the P3 display. I did find though that a 2009 MacBook Pro had some problem getting all three in line, but I was using ColorEyes software in that case. I have a 2013 rMBP which does run the X-Rite software and that's fine. My 2015 iMac does show slight splitting of the RGB curves but that's about the thickness of each line. The 2009 machine is a little wider but nothing to write home about. All three on both machines are linear all the way to the top of the chart

Are you letting the software set the brightness level, and if so what have you set it to. Also what is the final verification page stating it has achieved as far reaching the aims you have requested
Click to expand...

Hi John,

Thanks for your post.

No, I set the brightness level myself (120cdm) which is pretty much the default setting of the monitor anyway. My concern is the level of correction the software is having to make to the graphics card LUT which seems quite high in order to achieve the CCT of 6500 and a gamma of 2.2.
I'll do a quick calibration now and post the results so you can see for yourself.

Cheers
Neil
 
Hi John,

Different approach this time - ran two calibrations, one at native, the other at D65.
Other settings, identical for both runs.

Target Luminance - 110 cdm
Gamma - Standard (2.2)
Contrast Ratio - Native

Chromatic Adaptation - Bradford
ICC Profile Version - Version 4
Profile Type - Table Based

Patch Set - Small

ADC ticked

Native result for the GPU LUT - CCT achieved D6903

Screen Shot 2016-04-15 at 15.33.10 by Neil Almond, on Flickr

D65 result - Achieved D6483

Screen Shot 2016-04-15 at 15.41.18 by Neil Almond, on Flickr

Judging by the above graph the issue in using the D65 calibration is that the colour channels for both blue and green are significantly reduced by around 1.4 million tones out the 16.7m available, nearly 10%!
This will cause posterisation and banding most likely. I think in this case i'll stick with the native WP and adjust accordingly!
 
Hi Neil, calibrating can do your head in can't it!

Neither results are far out, do you mainly put stuff on web or do you print a lot?

So you have told the software to aim for a 2.2 gamma and it is achieving 2.3? Or 2.3 is native and you are wondering if you should change to 2.2?

Out of interest what brightness are you calibrating to? The one thing you can change on a Mac, I would start at 100 luminance.

Your white point is slightly blue, I've learnt to go for quality of calibration (lower delta e values and good 2d graph) over bang on white point but it is a case of trying multiple calibrations at different settings to see what gives the best results.

I felt I'd only really nailed it when I got a whole variety of prints viewed under daylight that as near as possible due to tonal limitations bang on matched my monitor. If you have some prints then use them to compare, it is the best test, anything else by eye is guessing.

Il check my screen settings tonight and see if I remember anything useful.
 
NeilA1975 said:
Recently purchased a 27 iMac late 2013 model.
Using i1 profiler and an i1 D3 what settings should I apply?
I've already measured the native white point at around CCT 6900k, moving away from this towards 6500k skews my 2d graph. Native gamma also measures 2.3 not the desired 2.2. The lack of monitor controls especially the RGB channels leaves me with this dilemma.
Appreciate your input.

Thanks
Neil
Click to expand...


Capture.JPG

That's for a dimly lit room. Use 120CD/m2 if you have average room lighting
 
I prefer to edit in a dimly lit room so that offer light sources of differing colour temperatures stent affecting my vision or my screen, also seems to get the brightness close for most prints. For me at 120 prints come out dark, even 100 is too bright for some.
 
Craig_85 said:
I prefer to edit in a dimly lit room so that offer light sources of differing colour temperatures stent affecting my vision or my screen,.
Click to expand...

Use daylight balanced lighting.. that's the best way.
 
Craig_85 said:
Hi Neil, calibrating can do your head in can't it!

Neither results are far out, do you mainly put stuff on web or do you print a lot?

So you have told the software to aim for a 2.2 gamma and it is achieving 2.3? Or 2.3 is native and you are wondering if you should change to 2.2?

Out of interest what brightness are you calibrating to? The one thing you can change on a Mac, I would start at 100 luminance.

Your white point is slightly blue, I've learnt to go for quality of calibration (lower delta e values and good 2d graph) over bang on white point but it is a case of trying multiple calibrations at different settings to see what gives the best results.

I felt I'd only really nailed it when I got a whole variety of prints viewed under daylight that as near as possible due to tonal limitations bang on matched my monitor. If you have some prints then use them to compare, it is the best test, anything else by eye is guessing.

Il check my screen settings tonight and see if I remember anything useful.
Click to expand...

Hi Craig and thanks for replying.
Yes, indeed it can! I'm used to calibrating TV's and projectors which can be nearly as frustrating but have hardware controls so a lot easier!
Anyways, what I decided to do was to perform a clean install of El Capitan and start over again.
Results were slightly different, but I've decided to calibrate to the native white point of approx CCT 6900K, cdm of 110 with a gamma of 2.2.
The issue I had with calibrating to D65 meant that i would of lost approx 1.3m tones out of the 16.7m available given the LUT adjustments. This would likely cause banding and other artefacts so in the grand scheme of things, not really worth it.
Cheers for your input.
Neil
 
I have a 2012 iMac and tried to calibrate it and I could not get anything decent at all. I gave up in the end and sent the spyder unit back. After talking to a few people in the know they said that Imacs in the last 5 years or so are set up better than any calibrator can do. They don't need calibrating apparently.
 
AndyWest said:
After talking to a few people in the know they said that Imacs in the last 5 years or so are set up better than any calibrator can do. They don't need calibrating apparently.
Click to expand...

In one aspect they're right. That being as long as you're happy with whatever you're presented with then they don't need calibrating at all.
If on the other hand you need to work to a specific standard then it will require calibration with a meter and software to make those necessary adjustments (as I've illustrated), but even then the results of the iMac are poor compared to pro grade displays.
 
NeilA1975 said:
If on the other hand you need to work to a specific standard then it will require calibration with a meter and software to make those necessary adjustments (as I've illustrated), but even then the results of the iMac are poor compared to pro grade displays.
Click to expand...
True.
 
Neil

That's 6450K doesn't look like a bad calibration, not a good one , but I've seen worse , a lot worse.
Try using the the larger target and see what difference if any that makes. This calibration wont cause posterisation or any similar problems, it just means that the images will look a bit warm. I had a similar display results with my 2009 MBP with the blue channel not matching the red and green. But then I wasn't using it to match back prints

I wonder if there is any stray light getting into the puck. Try calibrating in the dark. I have found this give slightly better results in the past.
 
Last edited:
Neil

Tried to replicate your results. However doing a little research ( buried on the X-Rite web) the curves you see are not calibration curves, they are in fact the LUT for each channel. The calibration is the patch screen, where if calibration is achieved the appropriate patch is struck through.

So if you need the screen to be cooler you will get the blue LUT to be higher. This fooled me too as the previous calibration software I used showed you the calibration curve not the LUT curve. The Dell display was obviously nearer the aim you were after hence the near perfect curves on the LUT window.

I've learnt something new today

John C
 
Chappers said:
Neil

Tried to replicate your results. However doing a little research ( buried on the X-Rite web) the curves you see are not calibration curves, they are in fact the LUT for each channel. The calibration is the patch screen, where if calibration is achieved the appropriate patch is struck through.

So if you need the screen to be cooler you will get the blue LUT to be higher. This fooled me too as the previous calibration software I used showed you the calibration curve not the LUT curve. The Dell display was obviously nearer the aim you were after hence the near perfect curves on the LUT window.

I've learnt something new today

John C
Click to expand...

Yep, after a bit of research i also learnt that the 2d graph in the i1 profiler software is a representation of the LUT for the said CCT target and therefore the 2d graph of the 65k calibration makes sense after all!
I'm still of the opinion however that sticking with D65 is reducing the colour tones in blue and green significantly as the LUT illustrates - I think in this scenario ill stick with the native white point to be on the safe side.
 
You must log in or register to reply here.
Back
Top