Point and shoot cameras are getting better and better. I've owned several and still use them. So what's the difference between an "8 megapixel" camera like the Canon Pro1 for $600 and an 8- megapixel $4000 "professional" camera (body only) such as the Canon 1DMarkII? There are a lot of more subtle differences, but the major one is that all pixels are not created equal.

Its the size and quality of the pixel, not just the number of pixels, which determines image quality. Here is a good graphic which shows the size of some professional camera sensor pixels (consumer cameras use *much* smaller sensors). Professional digital SLR cameras have huge sensor arrays, and since silicon is more or less sold by the square millimeter, bigger sensors cost more. It's an area rule, and also tends to be somewhat logarithmic, so a 20 x 36 mm sensor costs many times what a 5 x 4 mm sensor does. Small sensors are much more subject to noise. Point and shoot cameras get noticably noisy at anything above ISO200, and many are not capable of a professional-grade image at ISO400. Even at lower ISO settings, the signal to noise ratio of point and shoots is many times lower than professional cameras. Consumers generally believe that the more megapixels, the better the image, but this is absolutely not true. All things being equal (which, admittedly, they seldom are) if you have two sensors with the same overall dimensions, the one with fewer pixels will have better image quality per pixel. This is one reason why cameras such as the Canon G3 are more highly regarded than newer models such as the G5. So the camera makers can pump out new cameras with a larger number of smaller, noisier pixels, on a smaller silicon sensor (which costs them less to make) and charge you more since it has more megapixels! If you now feel confused you are not alone.

Most consumer cameras capture an image in JPEG format, which is a highly-compressed (using lossy compression algorithms), restricted color space (typically sRGB), which uses 8 bits of data per pixel. This is great for saving a snapshot in the smallest space, but terrible for best image quality.

The best image quality is obtained by shooting in what is called 'RAW mode'. RAW images use 12 bits per pixel for 16 times the dynamic range of an 8-bit image. I use ProPhoto RGB color space for all image capture and manipulation. More on this below.

Finally, the lens quality of point and shoot cameras is a far cry from that of pro quality lenses such as the Canon L series, which start at several hundred dollars each and run up to thousands of dollars. Here's some information on a Canon 17-40 f/4 L lens which has a discount price of $680, and notice that this circular polarizer (a special filter which screws onto the front of the lens) sells for $175.

Still, just buying a bagful of expensive equipment doesn't make you an experienced photographer any more than buying a grand piano makes you a concert pianist. This isn't a perfect analogy since anyone can get lucky with a camera, but it's much harder to get lucky with a piano.

All professional DLSR cameras support a 'RAW" image format. In the case of the Canon DSLRs, these raw files have 12 bits of resolution per color per pixel. This means each pixel's red, green, and blue colors have 4096 levels of intensity. Even the best quality JPEG images have only 8 bits of intensity - only 256 levels. This is 16 times less range than the RAW files! This is the reason I shoot all professional jobs in RAW mode. RAW images allow much more headroom for processing after the image is captured. There is a downside to RAW. Image files are much larger, and they require more processing to convert them to a usable or printable form.

JPEGs are compressed into 8 bits of color per pixel, so they cannot contain the same level of detail as a RAW file. JPEGs are fine for use on the web, and in small prints. They are much more convenient. A typical 6.3 megapixel high quality JPEG file will be up to 3 megabytes in size. JPEGs are easily emailed or transferred over the internet. A CDROM will hold about 300 high resolution JPEG images.

TIF files can be either 8- or 16- bits per pixel, with lossless compression. A typical 16-bit 6.3 megapixel TIF file will be 20-40 megabytes in size. 8-bit TIFs are about half that size. TIFs are not practical for email and take a long time to transfer over the internet. A CDROM will hold about 20 high resolution TIF files.

The human eye's color gamut still exceeds the best film, digital sensors, monitors, and reproduction methods. Our brains process colors such that a "white" object looks white to us regardless of the color of ambient illumination (sunlight, candlelight, tungsten or fluorescent light, etc). The human eye, photographic film and digital sensors record light very differently from each other, using very different sensing techniques. Getting a photograph to look the way we want it to (the term 'acceptable' is perhaps more meaningful than 'right' in this context) requires some understanding of these differences, as well as the photographic process being used.

There is a huge amount of misinformation about color space, calibration, and so forth. Much of this misinformation stems from the complexity of color management, and from now-obsolete rules of thumb left over from the "old days" of drum scans of photographic prints. In the early days of computers and the internet, color space was very limited (early PC graphics cards displayed only 16 basic colors...) and therefore many early pragmatic standards map very poorly to the photographic work of the current time.

Professional digital cameras such as the Canon 1D series are capable of capturing more colors than can be displayed on the best current computer monitors. Epson professional printers are capable of printing more colors than a monitor can display.

A "color space" is a mathematical way to describe the way colors are represented. The color space used to capture, manipulate, and output digital images is absolutely critical to obtaining the best quality images. Color spaces exist in order to deal with the various limited gamuts of image capture, display, and reproduction technologies.

White balance deals with the issue of correcting or compensating for the color of lighting (it's color temperature, expressed in kelvin) so that white objects appear white, and neutral gray and other colors appear accurate, when viewed under defined conditions.

In order to trust what you see on a monitor, what you print on a printer, and so forth, all of these devices must be calibrated and use appropriate color profiles. This is all part of color management.

My standard professional prints are produced on Epson "professional" papers using Epson wide-format professional printers and Ultrachrome inks. All of these will outlast typical wet chemical prints from film, with a rated lifetime of 100-200 years (depending on how you store them).

"Giclee" is probably derived from the French "gicler" which means "to squirt". Giclee prints are produced using seven or more archival inks, on archival photo media (generally paper- or cotton- based). Done properly, Giclee prints have a lifetime of 100+ years, and have much better color accuracy and gamut than other means of reproduction. Epson Ultrachrome inks use pigment (rather than dye) to create colors. Currently, pigment inks have a much longer lifetime without fading or color shift, compared to dye based inks.

There are many internet references to giclee, incuding this one at gicleeprint.net.

If you have special requirements or questions please contact me.