oregonfrank wrote:
Not sure my understanding of the difference is accurate. Is it true that an image in both JPEG and RAW consists of the same number of pixels in either format and can be printed equally as large with equal sharpness? And, that the major difference is RAW offers greater latitude for PP?
They both have the same number of pixels. The JPEG is created from the raw file which comes first. They can both be printed equally large with nearly equal sharpness -- raw can handle sharpness a little better.
That raw offers greater latitude for PP is correct and that is a big difference. Why that's the case needs to be understood. The JPEG by definition must be compressed and is limited to an 8 bit image. The compression used with JPEGs is what we call lossy as in loss. Data is lost in the compression method and unrecoverable. The 8 bit limit applied to the JPEG likewise fixes the amount of data the file can contain and compared with the raw file it's less.
The JPEG is a finished image with the processing work completed. In the specification for the JPEG algorithm the term archive is used to define the format -- JPEG is an archive format. What they mean is that JPEG was intended to save the final form of the image in a compressed state.
The JPEG algorithm is brilliant plain and simple. It can take an RGB image and successfully compress it by 80% such that viewing the image we can't see any difference. It's a technological triumph and has enabled our current sharing of data and images. Imagine if all the images on the internet (still and video) instantly increased in size by 85% -- the increased bandwidth demand would strangle the internet to death instantly.
Because JPEG saves/archives a processed image you can face problems if you don't like the processing that was applied to get the image to it's final form. Many people will further edit JPEG images. This can be more difficult and require more skill and time than processing raw data because it can be hard to "unbake a cake" as it were. It's easier and more direct to effect a processing result with raw data than to start with a different result applied and saved in a JPEG and then try and get the desired result with the JPEG that was already created differently.
I personally prefer to work from the raw data because not only can I achieve the results I want, I can do that so much easier and so much faster. I also prefer to limit my effort behind the camera to just capturing the subject I want. If shooting only JPEG and you want to avoid the increased difficulty that comes from editing a JPEG then you really need to devote attention at the time you're behind the camera to adjust the various JPEG parameters available in the camera software. I need only compose and nail exposure. The JPEG shooter needs to compose, nail exposure, consider WB, consider the camera's DR (lighting contrast) adjustments, and potentially tone, color and sharpness adjustments, etc. as they will all be baked into the camera JPEG.
Two final considerations: 1. The camera engineers must pay attention to how fast your camera can create and save a JPEG and then take the next photo. They must address the expectation that the user could want to take several photos in quick secession. If they devote more effort to creating the JPEG it will take time. They have the option to increase the power and memory of the camera but that increases cost. So there's always a trade off made in the design of every camera to cut a few corners and/or use a faster but not quite as good algorithm than what's available in a computer workstation. No free lunch.
2. The camera software has no idea what you photographed and what you want it to look like. Therefore it has no alternative but to process your image to the bell curve. The camera software must apply averaging methodologies to the processing it does and treat your photo with the expectation that it meets the standard average parameters. This process of applying averages is something we do to manage many aspects of our lives. It allows us to automate processes. The corn is an average height and the machine can harvest it etc. Applying the rule of a mean makes automation possible. The software in your camera is automated and although you can influence how it behaves you ultimately don't fully control it's automatic behavior. The mean will rule. Another word with the same root as mean is mediocre.