We need to start out at the point you press the shutter to see what the camera does. The diagram below summarises how the camera handles the two options.
At the point of exposure, the camera is working with two sets of data. The pixels from the sensor that comprise the image data, and the camera settings (White balance, exposure, …) when the shutter was pressed.
If you’ve selected for the camera to produce a JPG file, the image data is processed according to the camera settings to produce a finished JPG file at the time of exposure.
If you’ve selected for the camera to produce a RAW file, the image data is retained in its original form, and the camera settings are retained within the RAW file as a ’sidecar’. The RAW file also contains a small processed JPG file that you display on the back of the camera when you do a preview.
Some cameras (eg Nikon and Canon) give you the option to have the camera produce both a RAW and a JPG at the time of exposure. That’s a very convenient option if you’ve want files you can deal with quickly and have the RAW files for more refined post-processing later.
So theory lesson over, what are the practical advantages and disadvantages of the two file formats?
JPG files have the following characteristics:
They’re smaller. On my Nikon I can get around 600 JPGS on a 4GB card against about 200 RAWs.
They’re quicker. They come off the flashcard as finished files. If I’m happy with them as is, that’s the job finished. I can upload to the web or email to a magazine immediately.
They’re harder to edit. Whilst automatic white balance on cameras is pretty good these days, in my experience more than half of the shots I take benefit from a slight tickle. Of course this can be done in Photoshop, but it’s not easy.
Editing a JPG file is highly destructive. Every time you press ‘Save’ in Photoshop (or any other image editing application), you’re destroying more of your picture. Why? A JPG file is a compressed file. So every time it’s opened, it’s uncompressed. And every time it’s saved, its compressed again. And every compression destroys more of the original.
RAW files have the following characteristics:
They’re bigger. Every single pixel of the original image is retained as what amounts to a digital negative. So they take up more space on the card in the camera, on the disk in your PC, and they take up more backup storage space. (You do back your pictures up, don’t you?)
They can be slower. Even if you accept a RAW file ‘as is’ out of the camera, you still need to open the RAW file and save it in whatever format you want. Websites, magazines, photo libraries won’t accept RAWs.
They’re dramatically easier to edit. To process a RAW file, you open it with a software application. Photoshop includes Adobe Camera Raw, a very good tool to process RAW files. Nikon offers Capture NX to process Nikon RAW files (.nef). These RAW file processors are very good at making colour balance corrections, getting exposures right, going some way to recovering lost highlights, opening up deep shadows, increasing vibrancy or saturation. (Don’t expect miracles, if you’ve got a lousy RAW to begin with, there’s limits to what they can achieve).
They’re completely non-destructive. Your original pixels are preserved. The processing settings can be changed to improve a picture, whilst the original image data stays intact. Edit as much as you want, you can always discard your changes and go back to the point of exposure. And when you’ve finished editing, you can save as as .PSD, .TIF or (if you must) a .JPG for whatevr it is you’re going to do with the finished picture.
This raises another interesting point. As your skill with image processing software improves, so you can get more out of your pictures. Recently I’ve been going back through stuff taken as recently as three years ago. I’ve been able to revisit some of the frames I discarded then as being too difficult to get anything out of. As my skill has improved in the intervening years, now I can. I call it mining old RAWs, and its a fascinating and rewarding way of spending a wet Sunday afternoon.
Somebody made the point to me that he’d done an experiment to see if he could see the difference between the two file formats. So he shot a picture as a JPG, and then as a RAW. When he compared the two, he couldn’t see the difference.
Think about it. He wouldn’t, because the RAW picture he’d be looking at would have been processed according to the camera settings at the time of exposure. He’d only see the benefit when he opens the picture in a RAW file processing application, and started to correct what the camera did as the shutter clunked. Then the benefit becomes apparent.
There is a risk you need to guard against. All RAW files are proprietary to the brand of camera you’re using. A Canon RAW converter won’t open a Nikon RAW file. The good news is Adobe Camera Raw will open pretty much anything.
That means when you think in archival terms, if you want to use RAW files as the basis, you’ll need to maintain a copy of the converter. And if you think this is a theoretical risk, it isn’t.
Nikon’s early DSLRs (like my D100) used Nikon Capture to process Nikon RAW files. Nikon Capture was a product of the days of Microsoft Windows XP. Guess what? When I upgraded my PC to Windows Vista, I couldn’t run Nikon Capture anymore. Not supported, and not going to be, ever. Nikon want you to buy Capture NX, and that’s the way they force you to do so. Archival storage of digital picture files is going to open a host of questions as the years pass.
If you want high quality pictures that get the most out of your camera and computer, there’s only one way to go. The RAW way.
