I have an iPad Air with its camera. I need to show a color of my shoes to a person who will dye shoelaces for them. So I need to reliably capture and transfer this color somehow. I assume that simple photo is not enough because of monitor color profiles and calibration differences. I thought of using some "colorimeter" app for iPad and extracting a color hex code. Is it a feasible idea? What else can I do in my case without professional equipment?

  • \$\begingroup\$ If you're talking about shoes, why not use the actual shoe polish that matches your shoes instead of using photography? Then, there's no guessing. \$\endgroup\$
    – Stan
    Jan 9, 2018 at 6:45
  • \$\begingroup\$ Send the shoes to the person so they can choose an appropriate color for the shoelaces. \$\endgroup\$
    – xiota
    Feb 27, 2019 at 19:17

4 Answers 4


iPads [in fact most mobile devices] tend to be a bit 'over contrasty' unless you actually calibrate the screen with a hardware colorimeter; which probably means that even on your own screen the colour will be wrong. It will also vary depending on backlight brightness & surrounding lighting conditions.

Sending that value to someone else, who also has a non-calibrated & potentially over-contrasty screen, in an uncontrolled backlight/ambient light environment, just multiplies the potential for error.

Your only real viable solution, so you both know you are seeing exactly the same thing would be to print varying samples of the colour until you can clearly see, in good light [cloudy daylight may be the closest you can both get to being the same value], that it is a true match... then post it, snail-mail.

Alternatively, both of you would need either professional Pantone swatches... or at a push, find a paint swatch at a local DIY shop, if you both can source the same paint manufacturer locally [or, again, post it].

  • 7
    \$\begingroup\$ Using paint swatches from national chains is a good alternative to ponying up for a pantone book, good thinking. \$\endgroup\$ Jan 7, 2018 at 18:18
  • 1
    \$\begingroup\$ There is an affordable Munsell project that only costs $75. amazon.com/gp/product/1632270129 munsellcolourscienceforpainters.com/AnAffordableMunsellBook/… Comparing to standardized color patches like this is a classic and simple way to remotely describe color to a vendor. Of course you both need a copy of the patches and they have to be viewed under the same light. \$\endgroup\$
    – Chandler
    Jan 8, 2018 at 4:45
  • \$\begingroup\$ As above comment by Chandler states, "Of course you both need a copy of the patches and they have to be viewed under the same light." Using the same light source is the most important detail of any other in colour matching. Without this, all you have working for you is pure dumb luck. +1 for this critical detail. Nice work @Chandler. \$\endgroup\$
    – Stan
    Jan 8, 2018 at 17:52
  • \$\begingroup\$ What do you mean by "good" light? What is your definition of "good' light? Please correct this vague detail. \$\endgroup\$
    – Stan
    Jan 8, 2018 at 17:59
  • 1
    \$\begingroup\$ @Chandler your "affordable" Munsell probably costs 75x more than the shoelaces. \$\endgroup\$ Jan 11, 2018 at 21:27

Here's the important thing: "color" is a sensation that emerges in the mind, not a property of the physical universe.

We experience color when light hits the receptors in our eyes. We have different receptors which are more and less sensitive to different wavelengths, and the brain takes that information and a whole bunch of other context and the result is the sensation of color.

This is very different from the sense of smell, where there are some 10,000 different receptors in the nose, each directly linked to a particular molecule.

With vision, the perception of color is highly dependent on the environment. Consider this optical illusion:

two dogs

The two dogs are actually exactly the same in terms of red-green-blue pixel values, but we perceive them as colored differently!

And that's not even considering other factors. The color of the light itself is a big deal, both in terms of white balance (see What is the meaning of "white balance"? and Does Auto White-Balance Really Work? How?) and spectrum covered (see How does the colour of ambient lighting affect colour rendition? and What white balance settings do I need to capture the cast of a coloured streetlight?).

Today's digital cameras use a reasonable (but far from perfect!) approximation of the data-gathering part of all of this, with photo-receptors with filters roughly like the sensitivity of the receptors in our eyes. And given an external reading (or manual input), they can adjust for white balance — and can even make a reasonable guess in many scenes. But they can't account for missing spectrum, and they just plain can't do any of the fancy processing that you are doing constantly every walking moment.

Beyond all that, different materials react differently to different lighting and to different angles of viewing. Color in print is inherently different from color on a screen, and even if we're considering reflected light, color on paper is different from color on, say, cloth. People often have the misconception that an RGB hexcode represents a true color in some way. But, really, it just can't.

Generally, even with our carefully calibrated camera profiles and monitors and workflows, what we hope for as photographers is to get a result which matches our artistic vision ­— being perfectly true to the human perception of the world (our literal vision) is out of reach. And that's when making prints or images for display on our own hardware. When making images which will be displayed on other people's computers or printed on other people's printers, allbetsareoff.

All of this is why Pantone makes a lot of money on something which seems simple. They produce standards which can be used to reproduce similar-enough colors in a variety of media. So, one approach would be to get a Pantone sample which matches your shoes, and send the number of that. (And make sure you do the matching in similar lighting ­— daylight is going to be easiest, even though daylight too varies by location and time.) However, this is out of the budgetary reach of most individual humans. Another approach would be to find something else which matches the color closely and to send that.

  • \$\begingroup\$ "Another approach would be to find something else which matches he color closely and to send that." …Important Detail: Also send the light you use along with the sample. The only way it can work is if both individuals examine the same sample under the same illumination. \$\endgroup\$
    – Stan
    Jan 9, 2018 at 6:11
  • \$\begingroup\$ One would expect that a hexcode of, eg #FF8040 could literally be interpreted as "dial a red light of a standardized wavelength to 255 lumens, a green light to 128 lumens, and a blue light to 64 lumens, mix it all on a perfectly white surface, turn all other lights off, be perfectly done" ... no? \$\endgroup\$ Jan 28, 2020 at 9:42
  • \$\begingroup\$ @rackandboneman: there are two main problems: how much power do I need to apply to the green dot in my monitor to get it to emit 128 lumens, and what to do if the green dots on my monitor are a slightly different wavelength from the standard? It's not enough to have a standard, one must also be able to realize that standard, and realizing them is what the "monitor color profiles ... calibration ... professional equipment" mentioned in the question are for. The colors of reals objects are even more complex since they contain more than 3 wavelengths \$\endgroup\$
    – Max
    Jan 28, 2020 at 14:45
  • \$\begingroup\$ I was facetious. At the same time, I'm often amazed technology still has not got such simple accuracy down :) \$\endgroup\$ Jan 28, 2020 at 17:05
  • \$\begingroup\$ @rackandboneman there is indeed a standard called sRGB, which defines the color space that the (non-wide-gamut) monitors are supposed to work in by default. But in reality primary colors of each monitor differ, white point is often far from D65, and some deviation from sRGB gamma of 2.2 also exists. Moreover, the same value of e.g. G channel in a hexcode may result in somewhat different brightness of corresponding subpixel depending not only on the RB channels, but also on the surrounding pixels (e.g. making a small white dot on black screen brighter (in cd/m²) than full-white screen). \$\endgroup\$
    – Ruslan
    Jan 28, 2020 at 21:42

To scan a precise color at home is nearly impossible or very expensive.

A practical solution would be to go to a professional printing shop with the shoes. They usually have paper samples of Pantone color and place your shoes next to the samples, which are the closest visually. So you know which color you want. Then send the Pantone color number to the person. And the person-to-paint-your-shoelaces should match the dry paint color to the Pantone color on their side.

Alternatively if you know which paint brand will be used you can go to a paint shop and match the dry color.


You have to correct the white balance of the picture. Include a pure white object in the image and change the white balance of the picture till the object is pure white in the image. There are many apps for this. E.g. snapseed.

  • 1
    \$\begingroup\$ This is not sufficient for an accurate colour-match on two disparate screens. If it were, X-Rite & Datacolor would have been out of business years ago... \$\endgroup\$
    – Tetsujin
    Jan 8, 2018 at 17:54
  • 1
    \$\begingroup\$ You don't need a pure white object. You need a neutral color object like the famous 18% gray card. And not even that if you have a object in your picture, where you know it's color exactly. \$\endgroup\$
    – Horitsu
    Jan 9, 2018 at 6:09

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.