The default structure that we use for our color photographs (JPEGs, TIFFs, PNGs, etc.) is three channel RGB. A pixel in a photo then has a red channel value, a green channel value and a blue channel value. They're digital so we're talking about numbers. The range of numbers we use is typically from 0 to 255 (2^8). So a green pixel in your photo might be R = 20, G = 170, B = 25. Those three color values mix to make green (see 1st illustration below).

However that's not enough yet to identify the color. An analogy would be that many of the different languages in use around the world have adopted the same alphabetic symbols. I'm writing in English and I just used the conjunction "and." You know what I meant. BUT in a different language context I could use "and" like this: En ny and i dammen. And in Danish is a water fowl.

So with only the three RGB values we don't really know the specific color. In the 2nd illustration below I hope you see two different greens. Both have the exact same RGB values so that based only on those numbers there's only one color there. Just like with the language context for "and" those RGB values require a color space context in order to have meaning. We maintain and use different color spaces for different reasons. We don't have a one size fits all option and we're not going to do that. So when we create a digital photo and want to specify the colors in the photo we have to use the correct color space -- color management.

All digital cameras without exception assign a color space when they create a final photo. The color space is embedded in the photo file as an ICC profile tag. All raw processing software assigns that color space tag when a photo is exported in an RGB format. All digital photos must have a color space tag embedded in the file if the colors in the photo are to be interpreted and displayed correctly. Color managed software (almost all photo processing and display software) looks for and correctly applies that color space tag that should be in the photo. Any software that doesn't do that is doing something very wrong.

The photo comparison below is an egregious example and I set it up to make the point. For most people they won't notice the change because it'll be less extreme. I deliberately looked for a photo with lots of intense greens as they are especially effected. I opened the photo in PS (left side) and then opened the exact same photo in Photopea and made no changes but used Photopea to export a new JPEG. The difference you see is that Photopea not only ignored but in fact removed the photo's ICC color space tag and so left the colors in the photo undefined.

As a supplement to my earlier comment on this post, I thought I’d send along a photo of my current rig. What you see is an Olympus OM-D-E1 Mark ii, with a LUMIX 30mm macro lens, and then the Nikon ES-2 slide and negative copier. There is no other “glass” in what you see, only some step down rings. The ES-2 comes with capability for some limited forward and backward adjustment, to adjust the distance from the front of the lens to the face of the slide to be copied. I came to this rig by trial end error at my local camera store, who sold me the ES-2 earlier, and more recently the LUMIX macro lens. In a post of a couple of months ago, I showed you my original “rig” which used the Olympus PEN-F camera with an Olympus zoom lens and a front-end macro lens by Hoya. This new rig pictured here, eliminates unnecessary glass, uses a fixed focal length macro lens, and a very capable sensor that can capture even very very underexposed slides. Good luck with your own approach, whether it be a scanner (which I tried on many slides, but then went back and re-did a number that were not so good), or something like this, which is really an optical system using the features of the camera (auto white balance to correct color, auto focus, and HDR if needed). JimR