Hi. I have a few doubts. I’m using a Nikon D90. When I inspect my photos using the camera’s LCD, they look different -based on how the color space option was set. The manual says one mode is more vivid than the other one. Idk if my camera is bugged but Adobe RGB produces JPEGs with lower saturation (vs sRGB) in replay mode. In the "Nikon D90 Review", colors look more intense using Adobe RGB -according to my web browser.
https://www.dpreview.com/reviews/nikond90/24
So, what is the "truth"? I’m interested in JPEG/RAW (I shoot both) + Nikon software. I ask this because, depending on how I set my monitor profile (I don’t have a custom made one but I can pay someone to make one for me), the RAW file looks different. What is true, what is false? My eyes deceive me, my own computer deceives me. If I choose sRGB in Windows, photos look more saturated when processing RAW (regardless of the color space setting in Nikon’s software. Adobe/sRGB look the same). But If I choose Adobe RGB (in Windows), photos look duller (I like that. And then sRGB/Adobe RGB look both the same).
More doubts/general stuff. Correct me if i’m wrong:
-There is a input space for every camera. Custom ones can be made. Input spaces are huge.
-The option "assign color profile" does not change color RGB numbers, but images can look altered.
-The "convert" option changes color numbers but it aims for the same colors.
Now, part of the confusion comes from the fact that I was playing with options in the operating system/viewing-editing programs, etc. So the ultimate question is: will a monitor profile (made for my own screen) solve my issues? Thanks.
Bonus question: when processing is done and one saves a photo (not RAW) with a color space. Is the program (Photoshop, etc) actually changing the color numbers or is is just a tag?
There are a few issues here.
Let's start with ColorSpace.
A digital image is fundamentally a big grid of squares (we call them pixels). Each square has the number of a crayon assigned to it. Think of a ColorSpace as a box of 16,777,216 numbered crayons.
Different ColorSpaces have different assortments of crayons. If you are working in 8 bot RGB, all ColorSpaces have the same number of Crayons (however, in some wide gamut color spaces, not all the crayons are available).
The important concept is that the ColorSpace specifies the colors that are allowed to appear in your image.
sRGB is the most common ColorSpace. That vast majority of images you see on the web are in sRGB. Adobe RGB increases the color difference between adjacent crayons, and fits in a few more greens. If you want your image to contain a neon green, then you need to use a ColorSpace that includes neon Green.
Keep in mind that if all of the colors in your image fit into sRGB, the image should look the same in wider gamut ColorSpaces. However, if proper color management isn't in place, the colors are more likely to look wrong if you use a wider gamut ColorSpace (hence the warning in the manual).
The next issue is the colors in your raw file.
As it turns out, raw files don't record full color at any of the pixels. Each pixel records only brightness. The pixels are behind a checkerboard pattern of colored filters. Half the pixels are behind green filters, 1/4 are behind blue filters, and 1/4 are behind red filters (this is called a Bayer Pattern Filter).
You need software to guess at what the colors should have been at each pixel. Typically this is done by looking the luminance of the pixel, and the luminance of nearby pixels that are behind other color filters. While it sounds like this can't possibly do a good job, it turns out that it works quite well. The algorithms do a very good job of guessing colors that look pleasing to the human eye. However, these are guesses. Different raw processors can yield different results.
In fact, most manufacturers strive to achieve pleasing color over accurate color. If you are shooting landscapes the camera may accentuate greens. if you are shooting portraits, it may work to get good looking skin tones. A common selling point of many brands is that the colors from that brand "look better", than colors from another brand. Note that the claim is
not "more accurate".
Using the manufacturer's software is the best way to get something that matches the camera's jPEG conversion.
.
My advice is that you should work in sRGB. If AdobeRGB images look like they have muted colors, that suggests that your workflow is not properly color managed. In the long term, you can look into calibrating your monitor and your workflow. In the short term, I would suggest working in sRGB.
I would then suggest using the manufacturer's software for processing your raw files. Start with default setting, which should match the JPEGs produced by the camera. You can then make adjustments to control the images.
.
In terms of "accurate colors", that's a big can of worms. First of all, in the real world objects have texture. Texture is visible in the way the object looks different as you move (your eyes get slightly different views, and may see slightly different reflections off the surface). You can not match this with a typical digital image. You need to freeze a moment in time and capture the right reflections to make the viewer think the object has texture. In the extreme case, consider photographing a mirror. The mirror looks like whatever is being reflected in it. The appearance changes as the viewer moves. That doesn't happen in a photograph. Texture is in many ways like lots of tiny mirrors angled differently.
.
The next issue with color is that digital devices simulate color. In the real world an object might be emitting a single wavelength of yellow light. To the human eye that looks yellow. However, the typical digital display cannot emit yellow wavelengths of light. Your monitor emits red and green wavelengths mixed together, and the eye sees that as being the same as a single yellow wavelength. So while digital images can certainly look to the the human eye to be similar to the original color, they are unlikely to produce the same spectral response.
.
Of course the colors we see depends on the lighting. Consider shooting a photo of a neutral grey card outside at noon, and then again at sunset. At sunset, the light coming from the card will contain more yellow and less blue. Is it more "accurate" to record the card as neutral gray, or with the color cast of "magic hour"? What about shooting indoors under tungsten light? Should you record the card as neutral gray, or with the yellowish light of the tungsten bulb? The answer is that you can make a case that any of these options are "accurate". Which you use is usually an artistic, not a technical question.
.
Unless you are in a specific niche where accurate color is key, I wouldn't worry about absolute color accuracy. Go with pleasing color for your art. Let "magic hour" influence the tone of your portraits. Nudge your colors a little cooler or warmer to give emotion to the images. Adjust the color saturation up or down as you see fit. Create images that people will love.
