I was just reading this post over on CML, and found the distinction of how different RAW formats actually quite useful, and worth keeping in mind.
While we commonly understand the principle of RAW vs. ProRes as the ability to influence certain parameters (e.g. WB, highlight roll-off, exposure) that otherwise get baked into the file, we don’t tend to go beyond that.
It appears understanding in a simplistic way goes back to the days when RAW first surfaced in stills cameras and was generally just a dump of the original camera data. Because of the size of the data in video, that’s actually not the case, but there are further distinctions Phil goes into, that may alter your expectation of how much of the image you still have latitude over.
RAW at some point in this industry mean raw data off the digital sensor, in the beginning uncompressed RAW data. Over time and somehow, it transformed into RAW-like data with implementations like: processed RAW data, compressed RAW not processed, partially demosiaced RAW data, a variety of compressed RAW types in a single camera, and even RAW-like and a variety of compressed types of RAW data as well as processed in a single camera.
