Info panel shows wrong LAB values

[stevenmiller]

2.4, windows 11

Hi, when using the crosshairs and looking at the LAB values at the info panel the values are wrong (without converting to LAB). Also the range of the AOpponent- and the BOpponent-Channel does not go from -128 to 127 (it's much smaller).

Only when converting the image (via Document>Convert Format/ICC Profile) to LAB, the info panel shows correct LAB values and the range is correct.

[Hangman]

Hi @stevenmiller,

The values you see in the Info Panel are displayed as percentages rather than 8-bit which I believe is a bug possibly related to AF-2424.

[stevenmiller]

@Hangman thx for the info. But finally I figured it out, the values in the info panel are not shown in percentages. And there is no bug.

The reason is that LAB/16 has a bigger color gamut than RGB/8. Any image that has RGB/8 format cannot have the full range of LAB values. It is just as simple as that.

U can simply check as follows:
1. Draw a rectangle (vector shape).
2. Color panel: switch to sliders, choose LAB, set values (L,A,B)=(50,-128,0). That colors the rectangle green.
3. Convert back and forth between LAB/16 and RGB/8  (via Document> Convert Format) and observe the info panel.
4. Observe LAB values in the info panel: In LAB/16 format it shows (L,A,B)=(50,-128,0)   vs  in RGB/8 format it shows (L,A,B)=(56,-43,9).
5. Use online color space converter to check that this is indeed the result of the color space conversion.
    e.g. use https://www.easyrgb.com/en/convert.php#inputFORM

 

[Hangman]

Hi @stevenmiller,

You're absolutely right... I'd been seeing what looked like some odd LAB values in the info panel even working with a LAB document and have equally been testing reaching the same conclusion...

 

[stevenmiller]

Hi, @Hangman, true, you get values even working with a LAB document for the same reason. If the original image was created in (or converted to) RGB/8 the color gamut is restricted to RGB/8. Converting the document from RGB/8 to LAB/16 document cannot change that or restore any lost information.

1. Using the above example you can use LAB/16 format and create the (L,A,B)=(50,-128,0) colored rectangle with vector shape tool. Then export this as an image as .tiff file with LAB/16 color profile.
2. When opening the .tiff file, the info panel will show the correct (L,A,B)=(50,-128,0) values. But once you convert it to RGB/8 and then converting back to LAB/16 the info panel will always show (L,A,B)=(56,-43,9), simply because the conversion to RGB/8 will lose the color information past recovery.

What happens during the conversion is: (L,A,B)=(50,-128,0) gets mapped to (R,G,B)=(0,155,116) from the bigger LAB gamut to RGB gamut - this RGB representation of the original (L,A,B)=(50,-128,0) value triple is already an approximation (the original color info is lost). When converting back from RGB to LAB the triple (R,G,B)=(0,155,116) gets mapped to (L,A,B)=(56,-43,9) because the original info is lost for good.