The most accurate, pixel for pixel image files are he .RAW format. Virtually all professional photographers utilize this format, but you do have to know what you're doing with it. Many computer systems won't even recognize a .RAW file as a file at all, so it won't even show up on a search. It, however, is just what the file extension says it is; it is the native image in its most raw state (not some proprietary file type). Most modern cameras convert the files to some other format before you can export them from the camera, but higher end cameras will now export the actual .RAW files directly. If you ever intend to do some more advanced manipulation of an image, such as layering of multiple images for a better final image, really need to be done using the .RAW format. It's not impossible with .JPG and other similar files, just harder. The other truly accurate image file type is the .TFF file. These are also pixel for pixel correct images. If accurate representation of all the pixels is important to you, these are the file types you should be looking at using.
Celestial images is one area where layering is critically important. You can, for example, eliminate all of the Earth's rotation from layering these types of images. Even if you have a high end Equatorial mount on your telescope or camera, there is still motion captured in the image. But if, for example, you layer several (sometimes even dozens) of the same image, you can eliminate almost all of the blur caused by the Earth's rotation. Off-topic a little, but I thought I'd just throw this in.
As for text files, the only problem with .TXT files is, it will eliminate all formatting which is where a lot of your proprietary file types come from. If a document is well written with good grammar, then formatting shouldn't matter too much, but for a lot of people removing the formatting can change the meaning of their document entirely in some cases.
edit - As for me personally, I generally use the .DOCX format because it's cross-compatible with so many other platforms. For images, I use the .RAW format and ultimately convert to generally .JPG with the finished image. If we want to get really confusing, then we should talk about video formats and video compression algorithms in particular! That can get pretty complex because they often involve more than one frame and cross-frame extrapolation.
Celestial images is one area where layering is critically important. You can, for example, eliminate all of the Earth's rotation from layering these types of images. Even if you have a high end Equatorial mount on your telescope or camera, there is still motion captured in the image. But if, for example, you layer several (sometimes even dozens) of the same image, you can eliminate almost all of the blur caused by the Earth's rotation. Off-topic a little, but I thought I'd just throw this in.
As for text files, the only problem with .TXT files is, it will eliminate all formatting which is where a lot of your proprietary file types come from. If a document is well written with good grammar, then formatting shouldn't matter too much, but for a lot of people removing the formatting can change the meaning of their document entirely in some cases.
edit - As for me personally, I generally use the .DOCX format because it's cross-compatible with so many other platforms. For images, I use the .RAW format and ultimately convert to generally .JPG with the finished image. If we want to get really confusing, then we should talk about video formats and video compression algorithms in particular! That can get pretty complex because they often involve more than one frame and cross-frame extrapolation.