For context, I am trying to generate quantifiable colour measurements to include in linear models, to see whether variables affect colours of my subjects. I am using subjects that are mainly dark with small brighter-coloured dots.
The subjects of my pictures aren’t bright enough, so while the image in general is well-exposed, the subject isn’t. Can I adjust the exposure/contrast/highlights of the raw file in photoshop first and save it as a .TIF file, then generate a Multispectral image in MICA by importing it as a “Linear Image” and measure the visible normalised RBG (which I would then use as response variables in my models)? Will this adjustment of the exposure/contrast/highlights affect my results? Is this even necessary? Does the selecting of grey standards automatically correct for problems with exposure in the image?
Thanks in advance for your help!
Without more context I can’t quite see what you mean by “quantifiable colour measurements to include in linear models, to see whether variables affect colours of my subjects”. It sounds interesting!
I would recommend not meddling with the images in Photoshop but reading them into MICA as they are, preferably as RAW files. The grey values you select will enable MICA to normalise the intensity response. As long as nothing in your image is overexposed that should be fine. The resulting greyscale image (or any RGB reconstructions thereof) can be adjusted to aid viewing by changing the brightness of the image in ImageJ. Just be careful not to save any brightness changes to the image, just change the settings for viewing. If your subjects are generally dark I would recommend taking the images as bright as you can without overexposing any of the bright parts in your image (e.g. the white or grey standard). The problems of having a large luminance contrast (dynamic range) in images are discussed here. I hope this helps. Please also check general guides on how to achieve good quality image data.