Exerpt from "Shooting Digital"
----------------------------
RAW is referred to as the negative of a digital image while TIFF or JPEG files are derived, much like a print from a negative in traditional photography.
Not all RAW data is alike. Raw files are generated differently from manufacturer to manufacture. Saving RAW data doesn't automatically translate into the best quality image but does gaurantee flexibility because all of the image data produced by the sensor is kept intact.
Hence, shooting in TIFF is not the same as in RAW because TIFF has already been processed by the camera and certain sensor information will be dicarded after applied to a TIFF image.
As our skills increase and RAW data imaging software improves, we can always go back to the original data and try again. Like negatives, RAW data files should be saved and archived carefully.
These are the steps typically requried to process a RAW image (not necessarily in order).
1. The patchwork of monochrome data is reassembled into red, green and blue channels.
2. De-mosaicing = interpolating gaps between pixels.
3. Sharpening and noise filtering are applied. An optical blur filter is used in front of the sensor to prevent aliasing due to high frequency patterns which cause the moire patterns and color fringing.
4. White balance is determined and applied.
5. Color space is expanded from the 12 bits per pixel generated by most sensors to 16 bits of color per pixel. At the end of the process, images are TYPICALLY saved in either 16 bits per color channel or 8 bits per channel. The more bits per channel, the more color information is retained.
My comment:
--------------
When you mentioned 24 and 48 bits, I was surprised. I am assuming that a 24-bit referred by your utility is probably an 8-bit TIFF (8 bit/channel, 3 channels -> 24 bits) and a 48-bit is probaly a 16-bit TIFF ( 16 bit/channel, 3 channels -> 48 bits). It's somewhat strange that your utility describes it as a 24 and 48-bit instead of a more conventional term, 8 and 16-bit.
A 16-bit TIFF image retains more color information but has limited filter processing in Photoshop. Luckily, "unsharp mask", the most useful filter in Photoshop, is available for a 16-bit TIFF. However, layers are not availble for 16-bit images. A higher bit, a longer it takes for Photoshop to apply signal processing to the image.
I would recomend using a RAW converter utility provided by your camera manufacturer.
I love shooting in RAW on my Canon 10D. That's because I can change several settings available on my camera using their converter utility such as white balance, exposure compensation, contrast, color compensation, color tone, sharpness and color space.
I found color space and white balance the most useful to me. I can select Adobe RGB or sRGB before conversion. White balance is very neat because I usually shoot using "auto white balance" and I can later change it to more appropriate or desired white balance (or any color temperature I want). I can set to a higher temperature to counter underwater blue.
However, although my little Canon Powershot S50 will do RAW also , most of those settings are NOT available because either a sensor in Powershot won't save such information in a RAW file but rather apply it right away or the lack of it.
It very much depends on a type of your camera (which dictates how flexible RAW files will be) and converter utility. Different camera manufacturers store their (custom) settings on RAW files in which only their utility can read. The custom data will aid their utility in acheiving the best image quality.
Hope that helps.
---------------------
I attached 2 images of my dog yawning on the deck.
1) The first jpeg is derived from the original RAW image taken with Canon 10D using the "daylight" white balance setting.
2) The second jpeg is derived from the same RAW file but using the "shade" white balance setting.
