- This topic has 9 replies, 2 voices, and was last updated 1 year, 5 months ago by
.
-
Posts
-
Windows 11, Calibrite Colorchecker Display Pro, Philips 27M1N5500ZA Monitor, Lightroom 14.1.
I have been struggling with calibrating my monitor for some while. I have nearly got there but there is still a slight issue. I get my printing done by a well thought of professional print company and as such their calibration should be correct. However when I get prints back they do not quite match the monitor. The blues seem to have a purple cast. I have played around with the HSL settings in Lightroom (Classic) to get the image a better match with the print and the adjustments I would need to make are detailed below. This is obviously the wrong way round and to get a good match my calibration must be out by this amount. I am wondering if my Display pro needs a correction added when I calibrate. I have tried all the ones available from the DisplayCal menu, but none seem to sort the issue. I don’t have access to a spectrometer and wouldn’t know what to do with it if I had. I am flying by the seat on my pants a bit, when it comes to calibration so I am looking for advice.
Saturation
Red – 20
Blue 0
Purple +10
Hue
Red -20
Blue + 20
Purple -30
Luminance
Red +10
Blue +5
Aqua +10
Monitor Dimmed by 2/3 stop to 1 stop (Dimmer 30)
Calibrite Display Pro HL on Amazon Calibrite Display SL on Amazon
Disclosure: As an Amazon Associate I earn from qualifying purchases.The problem is 99% chance in the light source you use to evaluate your print, unlikely to be d50 normalized light.
Since its Spectral power distribution do not match D50 although it may be close (multicolored led, Violet led + RGBphsophor) or not close at all (WLED) some colors won’t match.Also as a general rule (that may need a manual tweak): monitor whitepoint = lightsource color coordinates + manual tweak due to surondings or “ambient light reading” limitations on your HW. monitor brightness = lightsource lux at paper level divided by Pi number + manual tweak (etc).
Once you meet this, to get a match:
a) recommended, buy better light source or live with these errors.
b) if you wish to take HSL 6 axis approach to get a match WITH THAT LIGHT SOURCE… which may not be a good idea:
-dowload a colorcheker 24 and get a physical color checker 24.
-on digital colorchecker, convert to monitor colorspace in PS, this means NOT to leave it non color managed or RGB monitor, this means “convert to”.
-apply HSL layer till colorchecker matches.
-add (by enlarging canvas) 6 patches 255 RGBCMY and at least 10 black to white.
-measure 24 + 16 patches with spotread and write manually a TI3 file. I do not thing that you could automate dispread through Photoshop output…. hence spot read and read manually. Remember that spotread must use your display colorimeter correction.
-with those measurements ina Ti3 use colprof ARgyllCMS app to create a icc porfile that describes this new behavior in your monitor
-use LUT3D crator and make a LUT3D, source this new profile, destination old you custom monitor profile.
-on OS settings keep your old profile
-use DWMLUT on MS Windows to load that HSL 6 axis layer globally, WITHOUT assigning to OS settings the new porfile made from 24+16 patches.Anyway, option A) should be your recommended action.
Many thanks Vincent.
I understand your point about viewing conditions. I have used downloaded standard test prints and viewed the print under various viewing conditions including daylight and 6500k lighting in home office. The deviation on the blues seem consistent across all lighting. I could construct a lighting booth to view the print but it is the monitor cast that is the problem for me.
The second problem is that I am a bit of a numpty when it comes to the technical side and I don’t reall understand your suggestions. I could probably figure out “monitor whitepoint = lightsource color coordinates etc.” after a week or two’s sleepless nights, but option b) blew my tiny mind. I am not sure if the solution is going to be too technical for me?
D50 & others is a syntethic illuminant made from several meaurements, hence you cannot use daylight to compare. Same applies to an unknown led or CCFL lamp which color even if it is exactly 6500K intersection with planclian locud or daylight won’t have the same spectral power distribution (SPD) as standard illuminants.
To appraise a printed copy against a display you need a “near” normalized light source, from the very expensive normlicht for pro enviroments to teh small Grafilite from Calibrite and between all the far east miltiled/violetd led technology.
It’s a requirement.Ragarding option b) your HSL laver depends onc image colorspace app, so if you do not have an advance monitor with such controls in HW, per HW calibartion slot, it’s a way to simulate what you did in LR system wide. On pro monitors it easy, it’s in their price, on lesser monitor and thanks to DWMLUT you have to do manually all the steps. Also since this procedure is not recomended because it’s only valid for your far from ideal light source, nodoby automated it for these lesser monitor.
Anyway… after reading your last message I remembered something. Some low cost LED monitors have an SPD in blue that produces a very saturated blue pushed towards violet zone. Although you may think by seeing a 2D gamut plot sRGB blue falls inside, it does not if you plot the true 3D volume of each colorspace. In these displays sometimes there is a spike in 3D with sRGB blue that falls outside the native gamut monitor, even if native gamut blue is farther away in 2D coords. You can use DIsplayCAL profile info window to generate Lab* 3D view.
So sRGB blue is out of gamut and the edge towards cyan may fall OoG too.Of course you issue may be just a wrong whitepoint with an overall blue cast by wrong or missing colorimeter correction or even that you are an outlier for “standard observer”.
But even if one of these were happening, you still need a good light source for print to screen appraisal.
Also I forgort one of the most important things…. you need to softproof using the profile of your print lab for the specific product you order. There are a lot colors in sRGB and AdobeRGB that you cannot print. Softproof simulated this.
Since this is the first step, always, I assumed that you did… but you may not. And if printer lab does not provide their softproof profiles for each of their products, this is a very good reason to do not work with them.
Thanks once again Vincent.
You are correct I have the paper profiles supplied by the printers and I “convert to profile” before softproofing and sending them to print. I don’t fully understand your point about 3d gamut plots so I will have to give this more thought (if my brain can take it).
I feel I should clarify my situation. I am not a professional photographer, but I am a memberr of a local club. We do have competitions but I don’t know or cannot specify the viewing conditions either for the judges or displaying in our hall. Most of the time I view my images in daylight. I don’t have a viewing booth and therefore believe that 6500k is the right lighting for me to view the images. I have tried D50 but for me the monitor is too red and a bit dark. I have tried to research this and most advice for non professionals is D65. One of the best explanations on a website said “However, if you are looking at an image on monitor by itself without reference to a print, whites under D50 calibration will look yellow and dingy. Unlike a paper print, our eyes never fully adapt to see the white from the monitor as pure white. This is a quirk of our visual system and how our eyes interpret reflected versus emitted light. In contrast, with a D65 whitepoint, our eyes can fully adjust and whites on the monitor will seem truly white. That’s why D65 is recommended for photographers – most of the time we are editing our images without reference to a print in an adjacent light booth with controlled lighting.” This seems to fit my processing position. I may have to live with option a) in your previous reply.
I have re-read (and re-read) your previous posts and I think I am beginning to understand option b) (I have ordered a Datacolor Spyder Checkr 24 as this seems to provide a short term fix by creating a preset to counter any anomolies. ) If I am correct option b) is a way to create a corrected profile for the monitor. Is that right? My problem is I don’t understand how to use the Argyll code stuff and I am a bit lost on some of the acronyms you have used. Hopefully I will figure this all out.
The proper way to do this is not to buy a colorchecker. It’s to buy a GrafiLite2 or at least the affordable E26/E27 bulb from Yuji (NormLite™ CRI 98 D50/5000K Standard Illuminant LED Bulb (ISO3664:2000) for Art & Studio, ~$70 for 2 pieces).
Any tweak is pointless without those devices or better product bc your visual reference (the printed copy) likely to be wrong.Human visual system will adapt to D50 display after some minutes. Also as I wrote before from starting point (match white & brightness) you can apply whatever whitepoint tweak you may need. For example on those yujis you may need WP around D55-D58, excluding paritular paper color cast.
Also nodoby converts to softproof profile. You deliver on certain standard colorspace (lab may have its preferences about one or leave you as long as image is tagged, with embeded profile) and softproof on each particular lab product (matte, poster, album, ine art product… whatever product you want).
So it looks like every step on the whole process may be wrong, from your enviroment to evaluate prints to lab itself.Vincent apologies if it sounds like I am arguing against your advice, but as you are aware the internet is good for providing advice (tons of it! ) and, whilst making mistake along the way, I am trying to find a route that I can understand and suits my situation. I have limited budget and may need to make a viewing booth/get a bulb/grafilite etc. but that won’t be for a while. I still believe that, as my viewing conditions are in daylight and my monitors are almost there with that ( save minor the blue/red issue), D65 suits my daylight viewing. Prints processed and viewed on the monitor in D50 and just look odd. Maybe my eyes are what’s wrong?
With regard to the “covert to profile” I have spent some weeks discussing this with the printers and their advice is to edit the image and when I am happy convert to profile. I suppose my original message was misleading as I edit, then soft proof using the print profile and then convert to profile before sending to the printers as per their instructions.
I do appreciate your comment about the Color Checkr as this (Inderstand) will only create a preset as a quick fix and not solve the underlying issue. But it was a budgetary decision. What I really want/need to do (I think!) is find a way to create a calibration correction with the knowledge and equipment I have available. This may not be something I can acheive!
On low budget you do not need to make a booth, just use an existing dimmable flexo desk lamp that admits E26/E27. That’s the point of yuji bulbs.
Also this is the point of the small GrafiLite2Thanks for a speedy response.
I do use a desk lamp with a 6500k daylight bulb for viewing when I can’t be bothered to walk out side (or the weathers grim!) This allows me to compare with the screen with cautious use of a bit of card to prevent glare affecting the screen. I will have a look again after christmas at using 5000k if I have a bulb kicking around.
In the meantime, does your option b) solve the issue of creating a calibration correction if I am able to decipher and learn to use the Argyll code to make a monitor profile?
