Back when the first really good DSLRs were starting to become available, the vast majority of photos still were being taken on film. So when you picked up a digital camera for evaluation, you naturally compared it to film. The quality of film was the bar. The new digital images not only had to at least equal film in sharpness and resolution, but they had to look like film, meaning an image had to look like a “photograph,” as we knew it, that had been scanned and displayed on a computer screen.
Because of this expectation, camera manufacturers had to start learning things about the art and science of image processing that they never had to worry about before—things Kodak had been thinking about for over 100 years. Raw digital image data had to be processed into RGB before it even could be viewed on the camera’s LCD!
Being in control of the image processing for the first time not only was a new responsibility, but it also represented an awesome new opportunity. For perhaps the very first time, camera makers could truly begin to differentiate themselves with the actual look of the image their cameras produced. And it wasn’t very difficult to figure out what people were going to want in this regard. By then, we had over 60 years for “the look” of color film to be refined and etched into our minds, first with Kodachrome and Ektachrome setting the standard, and later with Fujichrome upping the ante with even more saturated color.
So camera manufacturers had a very well-defined look to aim for when setting out to tune their in-camera processing. The result of that processing was an RGB image written directly to the camera card in JPEG or TIFF format. If the camera didn’t allow you to save the raw exposure data, what we now think of as the “digital negative” simply evaporated. Even today, most smartphones and point-and-shoot cameras still work this way. Letting the photographer have access to the original raw, unprocessed image data was a pretty radical idea. And, at first, I’m not even sure photographers knew what they would do with raw files, either. There weren’t any consumer-level tools available for processing camera raw files, so what would they do with them?