Tutorial for Argyll and dispcalGUI

Thanks, I am glad to hear, for me they are a lifesaver too.

Hi, I do not have a graphic printer or a spectrometer, but know that for each type of paper or ink type you need a different profile.
For Argyll, look http://www.lassini.com/blog/?s=argyll&search=Search

For Canon + Photoshop
http://www.youtube.com/watch?v=FQDxI4Xrmzk
http://www.colourphil.co.uk/photoshop-cs6-print.html

Pictus wrote:

An updated tutorial, now it is easier...

-Download and install Argyll + dispcalGUI
-I installed argyll to c:\argyll and dispcalGUI to c:\Argyll\dispcalGUI\
(better *not* to use version number for directory name)
-For some instruments Argyll driver must be installed, check here

Here a video showing how to install Argyll driver for Spyder 3 in Windows 7

-dispcalGUI should run as administrator (Windows vista/7/8)
Click with the right mouse button over C:\Argyll\dispcalGUI\dispcalGUI.exe go to
Properties>Compatibility> and enable “Run this Program as an administrator”


-First time you run dispcalGUI.exe, go to Menu>File>Locate Argyll CMS executables
point to c:\Argyll\bin\ (remember, I used NO version in directory name)


-Go to menu>Options and enable "Show advanced calibration options"
set dispcalGUI like this



Click in “Calibrate & profile”

-Put the calibration device over the spot and click in "Start measurement"


-Click in “Whitepoint/White level”


Wait 5 seconds and Click in “Start measurement”
Using the monitor RGB channels, try to make 6500 with the smallest delta(all 3 colored bars with the same size) and adjust the monitor brightness to +- 120 to 160 cd/m2, as I work into a dim environment I prefer 120 cd/m2, when you have the desired values, click in “Stop measurement” and “Continue on to calibration”
OBS, notebook users probably will not have the option to change the RGB channels, so just adjust the brightness. For the contrast most of the time is best to leave at the factory default.
When the process finish, install the profile...


To verify the result, go to menu Tools>Verify profile, open the verify.ti1 and save with any name.
Can also use:
http://www.photofriday.com/calibrate.php
http://www.drycreekphoto.com/Learn/Calibration/monitor_sensitivity.html
http://www.lagom.nl/lcd-test/gradient.php

Consider reading the dispcalGUI documentation
Also look at the review sites like http://www.tftcentral.co.uk/reviews.htm
or http://www.prad.de/en/monitore/reviews.html
to see how they calibrated their units for optimum results.

Here some interesting links about Tips on choosing a monitor and Best Monitor For Photo Editing 2012

Nice manual! Great job Pictus.

For Spyder4 or  i1DisplayPro/ColorMunki Display go to menu>tools>Import colormeter corrections>AUTO, then select the proper one in the Correction, also can try to use the ones from dispcalGUI database, just click in the small Earth icon.

(For the AUTO to work, first you must install the default calibration software from Datacolor or X-rite)



Glad to help.

When buying a used model, be warned that some pucks got organic filters that age with time, the DTP-94 and i1Display Pro/Munki are supposed to have glass filter, the Spyder 4 I do not know.

Argyll is the BEST, but it is not capable to access the PRO monitors internal LUT, so for PRO monitors basICColor(for NEC) or ColorEyes(for Eizo) are the best options, with non PRO monitors basICColor is not much forgiving of their flaws, but ColorEyes is.

The result from X-rite software is ok, but the software is toooo intrusive and may create some problems with Windows, the current Datacolor software for the Pro/Elite models is ok too, but Argyll+dispcalGUI is the king of the hill.

What monitor do you plan to calibrate?

Colormunki only works with X-rite software, Argyll and HCFR(Argyll)

You can get a Spyder 4 Express and use with Argyll+dispcalGUI.

http://www.luminous-landscape.com/forum/index.php?topic=53825.msg475402#msg475402

http://www.gamutprints.com/2012/04/i1display-pro-vs-spyder-4-and-spyder-4-quick-review/

http://nativedigital.co.uk/site/2013/04/i1display-pro-vs-spyder4elite/

The vantage of ColorMunki Display is that it is more sensitive, so it does a better reading from the darker parts and the i1Display Pro is like the ColorMunki but 5x more fast, it is a beast! For a Pro monitor get a PRo device, for a common monitor the Spyder 4 Express + Argyll+dispcalGUI will be fine.

Next time before buy a monitor, check here .

To install Unsigned Drivers in Windows 8 you must do a trick.

Thanks for this!
I have successfully used your instructions.
I am using Windows7, Xrite Colormunki Photo and a Dell 2209WA.
I had to go through a lot of hoops at first to disable Xright services, and get the Argyll driver associated with the Colormunki Photo, but it works good.
Oh look, new dispcalgui is out, 1.7.1.6

Anywho, I have noticed that the gamut of the Argyll profile is somewhat smaller than the Xrite ones.
I am using Gamutvision for the comparisions.
Has anyone else noticed this?

buggz wrote:

Thanks for this!
I have successfully used your instructions.
I am using Windows7, Xrite Colormunki Photo and a Dell 2209WA.
I had to go through a lot of hoops at first to disable Xright services, and get the Argyll driver associated with the Colormunki Photo, but it works good.
Oh look, new dispcalgui is out, 1.7.1.6

You are welcome, the X-rite software result is good, but it is a bloated software with services and etc...

Anywho, I have noticed that the gamut of the Argyll profile is somewhat smaller than the Xrite ones.
I am using Gamutvision for the comparisions.
Has anyone else noticed this?

I use ICCView and do not remember the differences last time I checked, but even if it is a bit smaller for me still the best result.

Pictus wrote:

For Spyder4 or i1DisplayPro/ColorMunki Display go to menu>tools>Import colormeter corrections>AUTO, then select the proper one in the Correction, also can try to use the ones from dispcalGUI database, just click in the small Earth icon.

(For the AUTO to work, first you must install the default calibration software from Datacolor or X-rite)

I think that loading .edr file called "RG_Phosphor_family_25Jul12.edr" might be the best option for i1 Display Pro (i1D3) when using with the new Dell wide gamut GB-LED monitors. It's confusing since the file name doesn't mention that one of monitor types in that file is actually GB-LED type.

I was wondering why Argill+DispcalGUI profiles seem to list primary coordinates for R,G,B in adpated D50 coordinates (as per ICC standards, which seems to me to be a bothersome relic since most stuff is D65 these days) but seems to list the white point in D65 and not adapted. When I look at profiles from NEC Spectraview or from Dell the white point also seems to be given in adapted coordinates. Is this a bug in argyll/dispcal or is it supposed to be like that for some reason?

I also notice that argyll+dispcal profiles don't bother to give a Chromatic Adaptation Matrix 'chad' element. I thought that the ICC standard said that unless a 'chad' is present that programs are supposed to use an older form of Chromatic Adaptation that is not as suitable for display technology as Bradford is. Now maybe most or all programs simply use Bradford anyway, figuring display's are the name of the game today, but could this potentially lead some programs to end up using an inferior chromatic adaptation??

I notice that NEC SV II gives a matrix in the 'chad' very similar to Bradford D65->D50, but a trace different, probably altered a trace because of the measured white point not exactly matching D65.

One weird thing is that the profiles from Dell for my UP2414Q have a matrix in the 'chad' that is also mostly like the Bradford BUT elements in the spots where I put B's are like 30-50% higher than expected:

xBx

Bxx

xxx

Do you think that is a straight out bug? Some special extra tweak they apply knowing that the i1D3 maybe doesn't quite measure the screen properly?? Or maybe it happens because of this next oddity:

I notice that when I tell it to target (for an sRGB+gamma 2.2 emulation) standard sRGB primary locations, for instance .640 for the x coordinate (in xyY) for red, that x coordinate for red that is stored in the profile, after chromatic adaptation using either the inverse of the embedded 'chad' or straight up Bradford D50->D65 ends up giving me values like .650 or a touch more even instead of like .640. The y coordinate is close to expected and the xy for green are close to expected and the xy for white point are close to expected (using Bradford instead of the embedded chad the white point even turns out to register super exact to D65, in general the coordinates all come out the smae to somehwat farther off using the embedded chad matrix with the two odd elements).

Do you have any clue why it might try to set x for red at more like .650 than to .640?

What is realllly weird is that if I do a gamma+matrix calibration and turn BPC off for profiling and then look inthat argyll profile that the x for Red after adaptiation turns into the expected .640! But if I keep BPC option turned on or use 'Curves+matrix" instead of "Gamma+Matrix" I get the same old story where the x coord of Red turns into measured as .650 instead of .640.

I'm not sure if Dell calibration solution has some bug that targets the x-coord for red wrong or if there is some rhyme or reason to this whole .650 when .640 was requested thing.

bronxbombers4 wrote:

I was wondering why Argill+DispcalGUI profiles seem to list primary coordinates for R,G,B in adpated D50 coordinates (as per ICC standards, which seems to me to be a bothersome relic since most stuff is D65 these days) but seems to list the white point in D65 and not adapted. When I look at profiles from NEC Spectraview or from Dell the white point also seems to be given in adapted coordinates. Is this a bug in argyll/dispcal or is it supposed to be like that for some reason?

I also notice that argyll+dispcal profiles don't bother to give a Chromatic Adaptation Matrix 'chad' element. I thought that the ICC standard said that unless a 'chad' is present that programs are supposed to use an older form of Chromatic Adaptation that is not as suitable for display technology as Bradford is. Now maybe most or all programs simply use Bradford anyway, figuring display's are the name of the game today, but could this potentially lead some programs to end up using an inferior chromatic adaptation??

I notice that NEC SV II gives a matrix in the 'chad' very similar to Bradford D65->D50, but a trace different, probably altered a trace because of the measured white point not exactly matching D65.

One weird thing is that the profiles from Dell for my UP2414Q have a matrix in the 'chad' that is also mostly like the Bradford BUT elements in the spots where I put B's are like 30-50% higher than expected:

xBx

Bxx

xxx

Do you think that is a straight out bug? Some special extra tweak they apply knowing that the i1D3 maybe doesn't quite measure the screen properly?? Or maybe it happens because of this next oddity:

I notice that when I tell it to target (for an sRGB+gamma 2.2 emulation) standard sRGB primary locations, for instance .640 for the x coordinate (in xyY) for red, that x coordinate for red that is stored in the profile, after chromatic adaptation using either the inverse of the embedded 'chad' or straight up Bradford D50->D65 ends up giving me values like .650 or a touch more even instead of like .640. The y coordinate is close to expected and the xy for green are close to expected and the xy for white point are close to expected (using Bradford instead of the embedded chad the white point even turns out to register super exact to D65, in general the coordinates all come out the smae to somehwat farther off using the embedded chad matrix with the two odd elements).

Do you have any clue why it might try to set x for red at more like .650 than to .640?

What is realllly weird is that if I do a gamma+matrix calibration and turn BPC off for profiling and then look inthat argyll profile that the x for Red after adaptiation turns into the expected .640! But if I keep BPC option turned on or use 'Curves+matrix" instead of "Gamma+Matrix" I get the same old story where the x coord of Red turns into measured as .650 instead of .640.

I'm not sure if Dell calibration solution has some bug that targets the x-coord for red wrong or if there is some rhyme or reason to this whole .650 when .640 was requested thing.

Monitors and graphics cards wear. So, in years the right color is off considerably and the calibration software needs to compensate for this.

So, the settings for a new set up will be different then for a 10 years old computer.