How many colors do you need ?

[Ink stained Fingers]

Profiling software typically has a function to create a page or more with color patches, let the user print those pages directly or save them for later use, such files are saved in .tiff format.
Do not ever try to save them into .jpg format e.g. after editing - adding comments etc or changing the size , the JPG un/compression algorithm will create for you lots of new colors you never dreamed about.

I cannot upload the original .tiff file here directly, about every image editing program has a hidden funtion - 'count unique colors' - this one displays 1101 colors - a few spots are double - probably some black spots.
Saving this file as a .jpg file and displaying it again will give you 177 453 unique colors - wow . The JPG arithmetic is pretty inventive here, most of the additonal colors are visible along edges - you typically can see those compression artifacts. The number of additional colors is related to the compression ratio, but in this case the number of colors is not compressed but expanded siginificantly.

 

[Roy Sletcher]

Very interesting comment which goes into areas I am interested in, but beyond my knowledge base.

Are you advising us of the error of using JPG files for critical colour? I think I was aware of this.

I am assuming your colour count coordinates were measured in the L*a*b* colour space which exceeds the gamut measurement accuracy of the RGB color model used in most of our inkjet printers.

Andrew Rodney of the Digitaldog has several informative postings and videos on this subject which I found valuable and informative. However, I lack the resources to follow up with my own measurements and tests. The colormunki just doesn't measure up (pun intended) and I can't justify the expense of a more sophisticated spectrophotometric tool.

rs

 

[Ink stained Fingers]

I am assuming your colour count coordinates were measured in the L*a*b* colour space

Actually I was not measuring anything yet at this point - I just created a target patch set and saved it as a .tif file from whitin i1Profiler. And I'm using this file to print it onto various papers with different printers as needed with Qimage. I'm typically adding a small comment field - printer - paper - driver settings - inks - at one edge of that image. You just should not save it as a .jpg file then because that would add 176 000 colors to your printout, most of them probably just a pixel wide but you wouldn't know where they are located. I'm just using the 'Count colors' function of a photo editor.

 

[Roy Sletcher]

I'm just using the 'Count colors' function of a photo editor.

If it is very basic colour editor it could be counting colour values which may not be actual visible colours.

I would like to get more involved in this field, especially with regard to producing better quality profiles than are possible with my entry level Colormunki. Unfortunately, Xrite's next model up is over $1,600 Canadian. A bit too rich for me, and I am not confident buying used versions on eBay because of my limited knowledge in the field. So much to learn!

rs

 

[Ink stained Fingers]

I got a new - unused i1Profiler Pubish packet via Ebay for 30% of the regular sales price including the newer i1Pro2 spectro. I was using an older Profile Maker package since recently.
i1Profiler created the 1107 colors patch field, no question about the correct number of colors, and .jpg would add 117 000 more - two programs are telling me that.
I have magnified the jpg image very much and cropped to a small part showing some of the
additional colors - pixels and lines along the edges

 

[turbguy]

To answer the original post question, on Tuesdays and Thursdays, 3.

But on Mondays, Wednesdays and Fridays, 4.

 

[The Hat]

On Saturday and Sundays, we just admire our own handiwork...