With thanks and apologies to Rick Sammon, Ive lifted and readily use a phrase I learned from him, and that is that RAW Rules
If youve seen the television show I do with my friend Bryan Weiss, or taken a private workshop with me, or done a workshop with me at The Newmarket Camera Club, you know how often I say this.
So for those who dont know or those confused by a load of bafflegab, what does this actually mean?
RAW, as the name implies means uncooked. The image that is captured is precisely what the sensor saw in all its glory or lack thereof. That little screen on the back of the camera does not show the RAW image, it shows a JPEG and while JPEG does have a real name, I refer to the process of making a JPEG as microwaving. You cook the image really quick and pretty much from the inside out. A cooked image can look great, but its been cooked by someone elses recipe and things get lost in the cooking.
When people start to shoot and edit in RAW, they invariably say it didnt look like this! This is flat and boring, and the colours dont pop, and I thought the detail was crisper and that there was more contrast and they go on and on and without support eventually say, screw this, Im going back to JPEG, because the pictures looked better. They arent wrong, those JPEGs did look better, but what if you wanted to cook the photo yourself.
Picture this. You want a hamburger. You want that hamburger to be cooked medium and charred on the outside. You want it to be juicy but not runny, maybe a tiny bit pink but not bleeding. You get your hamburger and its pretty darn good but it isnt what you really wanted and you know that if you had started with the raw ground beef, you would have gotten what you really wanted. Exactly correct.
Most higher end DSLRs capture in 14 bit RAW. Capture in JPEG and they drop to 8 bit. What does this mean? Consider the following table, where each column shows the number of variant tones at each level.
What does this mean? Fundamentally 8 bit images provide fewer tonal variants than 14 bit images. So when we look at the whites, we go from 8192 levels of white to 128 levels of white and down in the blacks we go from 256 levels of black to 4 levels of black.
And you may say, so what. Lets suppose that the image is not perfectly exposed. Its a bit underexposed, say by a stop and you need to brighten it up. There is a significantly lower amount of data to work with, in the example above, 64x less data to work with from a tonal perspective. Oh and did we mention that when you save as a JPEG at the default settings you are throwing away at least 30% of the pixels you captured? Yes this is done mathematically and for the most part it looks ok, but if you spent the money on a D800 with 36MP and save everything in large JPEG, you are getting at best 24MP out of the camera you spent so much for.
When you work with JPEGs you are working with less content. And every time you export as a JPEG, you lose some more. Work with the highest bit depth and the maximum amount of data through the entire edit process and make exporting as JPEG the VERY LAST THING YOU DO, not the first thing even before the image leaves the camera.
Now if you are shooting sports for a wire service or you are a photojournalist on deadline, your mileage may vary and you will need to do JPEGs, but thats not most of us. If all you want are snapshots, JPEGs are just fine there too, because we are less likely to spend time working in the digital darkroom on snapshots. But for your serious work, or when you are trying to grow as a photographer, RAW Rules.