Have you tried portable paint matchers for finding the perfect color match?

[Stephen Tashi]

I like to hear from forum members who have used portable devices to determine which paint matches the color of a given surface.

I notice there are "apps" for smart phones that purportedly do this. e.g. https://www.sherwin-williams.com/homeowners/color/try-on-colors/colorsnap-mobile
Do these require a calibration process? Or are the cameras in phones so standardized that no calibration is needed?

There are also devices that attach to phones.
e.g. https://www.amazon.com/dp/B01KKEMIF0/?tag=pfamazon01-20

Lately, I've needed to find paints to match previously painted concrete block walls. For that purpose, it would be handy to have a something that works off a picture (getting the overall color correct ) rather than something that must be put right on top of a surface.

I'd prefer a standalone device to a smartphone app.

 

[symbolipoint]

Paint match using portable machine? Matching of paint may depend on viewing with different light sources. How does the machine know what amounts of colored materials are present?

 

[Stephen Tashi]

Paint match using portable machine? Matching of paint may depend on viewing with different light sources. How does the machine know what amounts of colored materials are present?

Those objections also apply to non-portable machines. Perhaps, in theory, paint matching machines can't work, but I'm asking about practical experience !

 

[NTL2009]

Those objections also apply to non-portable machines. Perhaps, in theory, paint matching machines can't work, but I'm asking about practical experience !

Well, the Color Muse you linked to gets good reviews from those users, it seems to work quite well, and is not expensive (~$60).

I don't see the issue of light sources, the device is placed on the surface, it appears that it blocks ambient light and uses its own light source, which could be a known color profile, and/or calibrated out at some point.

... it would be handy to have a something that works off a picture (getting the overall color correct ) rather than something that must be put right on top of a surface. ...

Now you have introduced the error of the camera lighting and printing process from that wall to the picture. I would not expect anything better than a very rough color match going that way. Just compare two different printed pictures of a known color wall, taken at different times of day with different cameras. You will likely see a wide variation.

 

[Stephen Tashi]

Now you have introduced the error of the camera lighting and printing process from that wall to the picture.

Printing process? I want to be modern and use the electronic version of the photo.

 

[NTL2009]

Printing process? I want to be modern and use the electronic version of the photo.

OK, the concepts and problems are the same. Lighting, camera, and display variation instead of print variation.

People who do serious photo editing go to great lengths to use high quality monitors and then they calibrate them (with devices that are probably similar to that Color Muse).

If you are thinking of using pictures (which I don't think will work very well at all), do you really need the color match system to be portable? Couldn't the pictures come to the device (physically, or electronically)? Actually, electronically would be better. You can get the RGB levels from the file, and skip the distortions in the display device, and subsequent re-reading.

 

[Stephen Tashi]

If you are thinking of using pictures (which I don't think will work very well at all), do you really need the color match system to be portable? Couldn't the pictures come to the device (physically, or electronically)? Actually, electronically would be better.

Yes, that is an excellent idea. A program for a desktop computer would be convenient. Photos could be emailed.

You can get the RGB levels from the file, and skip the distortions in the display device, and subsequent re-reading.

Next, one must use the RGB data to select particular paints. My guess is that Gimp and other programs for desktops can display the RGB data from pixels in a photo. What's needed are the algorithms that proceed from that data to picking paint from a particular brand - Sherwin-Wiliams, Benjamin Moore, Glidden etc.

I supposed a determined person could get color sample cards from paint manufacturers and photograph them to create a database. However, I'm not that determined! I'm assuming people have already done this task since smart phone apps have been created.

There are all sorts of reasons why matching paint from photos shouldn't work and I don't expect it to work perfectly. I'd be happy if it worked well enough to make the first quart of paint bought as an experiment be close to the desired result.

Edit: I found this chart from Sherwin-Williams https://images.sherwin-williams.com/content_images/sw-pdf-sherwin-williams-color.pdf but I haven't found documentation for it yet. Perhaps its RGB values on a 0 to 255 scale as used in computer graphics.

 

[NTL2009]

Yes, that is an excellent idea. A program for a desktop computer would be convenient. Photos could be emailed.Next, one must use the RGB data to select particular paints. My guess is that Gimp and other programs for desktops can display the RGB data from pixels in a photo. What's needed are the algorithms that proceed from that data to picking paint from a particular brand - Sherwin-Wiliams, Benjamin Moore, Glidden etc.

I supposed a determined person could get color sample cards from paint manufacturers and photograph them to create a database. ...

Yes, GIMP does this, and you can use a 'select tool' to select an area, and GIMP will provide the average RGB for that area (better than a single pixel in most cases). It might even provide min/max values, I don't recall.

However, I'm not that determined! I'm assuming people have already done this task since smart phone apps have been created.

There are all sorts of reasons why matching paint from photos shouldn't work and I don't expect it to work perfectly. I'd be happy if it worked well enough to make the first quart of paint bought as an experiment be close to the desired result.

Hmmmm, I wonder if that RGB data for Pantone and/or fro m paint manufacturers is available? I do recall some of them published the tint names and amounts for each color. I found that helpful, because subtle shades are difficult to distinguish, but I recall looking at different 'off-whites', and noting that one had some yellow or green, and I was thinking, no, I want it more bluish. By going by the content, rather than how it looked in one specific lighting, I think I got a better idea how it would look in different light throughout the day/night.

If googling doesn't go anywhere, I do know a graphic artist that I could ask.

 

[NTL2009]

Edit: I found this chart from Sherwin-Williams https://images.sherwin-williams.com/content_images/sw-pdf-sherwin-williams-color.pdf but I haven't found documentation for it yet. Perhaps its RGB values on a 0 to 255 scale as used in computer graphics.

They sure look like the RGB values I'd expect.

 

[f95toli]

There are all sorts of reasons why matching paint from photos shouldn't work and I don't expect it to work perfectly. I'd be happy if it worked well enough to make the first quart of paint bought as an experiment be close to the desired result.

That should be doable. However, different pain formulations can look very different even when the colour is nominally the same due to differences in e.g. reflectivity and texture. I suspect this is why matching even with professional equipment rarely works well with e.g. white(ish) colours where even small variations in reflectivity makes a difference.
We've spent the past 4 year doing up our house and have tried paint matching a few times (in a couple cases because we needed a specific type of paint) and so far we haven't had any success. That said, my wife is extremely good at picking up even very slight variations (due to her job) so it is possible that you will have more luck.