Grey calibration for coins scan

[Fab117]

Hi,

I'd like to scan my coins collection (Canon MX925).

In order to get the most correct grey level, I include 4 standards during my scan.

 

There in a significant drift when I check my standards vs the one created with Affinity

 

Would someone know how to perform coins colour calibration based on my 4 standards?

Thanks

 

NB: Attached my initial scan

Fab

Planche_100_a_123A_originale_tiff_niveau_de_gris.tif

Planche_100_a_123A.afphoto

Edited October 4, 2020 by Fab117
Added afphoto file

[lacerto]

(...)

[Fab117]

Hi,

First of all, a big thanks for your interest to my case.

I'll start from scrap.

I would like to scan all my collector coins.

I did some trials with my scanner (Canon MX925)

Someone suggested to correct the colour coin which drifted from original

He suggested to perform a grey calibration

=> I looked for 4 grey samples. Then using various Android app, I "defined" their RGB parameters.

On my scanner glass I put the coins and the 4 grey samples.

Then on my scanner application, I put resolution at 600 ppm and tested export in png, jpg and tiff. Visually, I had the feeling that the tiff picture was better than the 2 other one's => I decided to use this format (may be wrong).

In Affinity Photo, when in separate layer, I add samples of what my 4 references should look like (the 4 rectangles I added in another layer), there is a significant difference

=> my question is: On my scanned picture, is there a possibility to correct the grey kevels in order to match to reality (based on my 4 "standards")?

Regards,

 

Fab

[firstdefence]

Just out of curiosity I used a colour balance filter on the top left grey, I created a square swatch shape and filled it with the corresponding RGB colour 190, 185, 178 and adjusted the colour balance sliders to get the swatch to fade into the grey card sample, the arrow points to the square swatch. Using the eye dropper after using the Colour Balance Tool shows RGB 191, 184, 179-180 as you move the eyedropper about.

The problem with this is although it gets the other grey samples close to their RGB numbers it cannot make them accurately read the RGB values, they are close "ball-park” values, the other issue is if you use the colour panel eyedropper you will see the values change as you move it about showing pixel colour variations, using the colour picker tool allows you to take larger average samples to average out the variations in pixel colours

  

[Fab117]

Thanks for you interest, but unfortunately, not clear for me.

Was really hopping that with Affinity Photo, I would be able to correct the scan if I add standards.

Have a nice day

 

Fab

[thomaso]

5 hours ago, Fab117 said:

Was really hopping that with Affinity Photo, I would be able to correct the scan if I add standards.

I wonder how you got the "original" RGB values for your 4 gray "standards" – their simulated values (red numbers) seem to drift not all in the same but each in a slightly different color direction compared to their "original" values. Note the placed "standards" aren't single colors but have a fine structure which varies in tiny areas, this might result in imperfect measurements, in particular the bottom right gray shows quite colorful elements (print grid?) in close view which might disturb when trying to measure a clear gray.

Maybe you can use instead a neutral gray background for the entire scan? (~$10 for 8'' x 10'')

In your scan I would expect a White Balance Adjustment to be sufficient for color shift correction. You could try using its picker to achieve a neutral white or gray.