OK, thanks for all the answers. Some reading ahead 
I have used the X-Rite i1 first version with DisplayCAL and also the ColorHug on several monitors and laptop screens for a few years. The i1 had a cheaper version and a more expensive one that only differed in the software supplied (for Mac and Windows), but dispcalGUI and Argyll CMS would supply the tools of the Pro version of the software with the lower priced i1 package.
Both tools worked well and really improved some monitors and all laptop screens. Iāve used an $1800 Eizo and now an ASUS ProArt 279 for the last several years. Those havenāt needed as much profile adjustment, and the ASUS has a hardware switch for profile toggles that includes sRGB and Adobe RGB with very tiny Delta E (<2). So I havenāt felt the need to calibrate as much. (The hardware toggle is very handy when youāre editing RAW in Adobe RGB and want to check output in sRGB.)
I canāt find it now, but I believe I saw a reference stating that Richard Hughes is/was (?) involved in writing display drivers or something involving color management for Red Hat, i.e. he knows what heās doing.
Not really adding nothing new, but just my 2 cents. Iāve been using my Pantone Huey with either displayCal or the Gnome Color Management with no issues for years.
Has anyone used ColorVision Spyder2PRO on Linux? Will it work ? Is it a good device for a non-pro photographer? I am asking because I saw it for 75 maple flavoured dollars in a local store.
I spend those tooā¦what store? (in case I ever get enough free time to get color managed)
Not sure. I just saw it on calgary.kijiji.ca
1 Like
I have Spyder 4 Pro and I am Linux user (distros: Kubuntu 14.04 and now 16.04). DisplayCAL is the software of my choice. Everything has worked smoothly, but the calibration process is very slow. Depending from settings, it has taken 1,5-3 hours to complete. It seems to be an ok device for ordinary photographers.
Spyder 2 Pro is quite old and I have no idea how good it is and how well it will work on Linux.
Have you considered open source Color Hug? http://www.hughski.com/
DisplayCal has a knob for setting the speed (# of patches, time per patch). You can set it to 15 minutes, you can set it to several hours.
AFAIK Spyder Pro doesnāt work with LED TFT Monitors. I wouldnāt buy such an old device.
Edit: Besides that, the Spyder 2 was released 12 years ago. I wouldnāt bet money, that this device is working accurate.
I wish there was a āmiddle groundā option for those that donāt want to spend the money on a hardware colorimeter. If I could use a reference color card to visually match colors in some program that would then make an icc profile for me, that would be ideal. Iād gladly pay up to like $50 for the calibrated color card, and I could certainly live with the results being more subjective to than with a colorimeter, since theyād be much better than nothing (which is how I currently roll). My understanding is that there was an old program called āMonicaā that let you do something like this, but itās been long defunct abandonware.
You can buy a colorimeter for 50$ on eBay, new, e.g. Pantone Huey, or bid for a used one for 10$. The āProā in the āPantone Huey Proā model is identical to the non-pro in hardware, and only differs in software, which we donāt care about as weāre using DisplayCal.
@Morgan_Hardwood Wow, I didnāt know they went for that cheap! I looked them up on eBay, and there are a lot fir very cheap! It pays to know the right thing to search for, I supposeā¦ Guess I know what my next photography purchase is!
Well, I just bought a used Pantone Huey off of the FleaBay for the grand sum of $15, including shipping. Now to discover how to use DisplayCal!
DisplayCal is pretty automatic.
1 Like
Good to know! Iām hoping all this will help me to be less nervous next time I send stuff off to be printedā¦
with a caveat, at least in my experience. I have a ColorHug2 (bought because itās both affordable and because I wanted to support free/open solutions), and I had to struggle quite a bit with DisplayCal to get good results. The defaults were producing quite bad output to my eyes.
1 Like
What did you struggle with?
Wait half an hour after turning your screen on, let it warm up. Run the latest DisplayCal. Select āOffice & Webā, in the Calibration tab set whitepoint to a color temperature of 6500K and set a custom white level of 120cd/mĀ² (if your monitor can go that bright - you can measure to find out).
If you select the āPhotoā preset it will set the whitepoint to 5000K which will make your screen look too warm under normal conditions unless youāre working with a D50 light booth.
Leave everything else at default.
You can optionally enable perceptual mapping by clicking on the gear icon next to āProfile typeā in the āProfilingā tab, but WARNING! Some programs, for example all versions of Geeqie except for git, will use the perceptual intent if the profile supports it without giving you an option to not do that, and since perceptual intent shifts all colors your images may look whack. My advice is to keep perceptual mapping disabled until you both need it and know that your programs work correctly.
Calibrate and profile. If things look good, then in the Profiling tab you can increase the patch count of the test chart from 175 to something like 1148. The time it takes will change from 7 minutes to 41 minutes and you will end up with a good quality profile.
5 Likes
red color casts in the generated profile. They are sort of a known problem, especially for version 1 of the colorhug. I have colorhug2 but found it to be still the case. After some digging, it seems that the issue is due to an inappropriate (for my screens, at least) factory calibration matrix. DisplayCal has now a built-in āhackā to work around this, by setting the instrument mode to āAutoā (only available for colorhugs). Then it gives decent results, but you are forced to use the white point and the white levels that are ānativeā to the display. Another thing that I found (again, by trial and error, and judging only with my eyes) is that for poor quality screens (like that of my old laptop) using a āsimplerā matrix-based profile gives more consistent results than a supposedly more accurate LUT-based one. Again, this is not the default.
But as I said above, YMMV quite a lot with a different calibration device.