Finally, all the pieces for ubiquitous digital photography are nearly here. Great D-SLRs, pretty good digital point and shoot equivalents, stunning print quality, and plenty of computer horsepower. Digital photography is set to take off with a bang. But where do we go from here? The market has been obsessed with counting megapixels for years, to the point where all of us are tired of answering the "How many megapixels is that?" question about our cameras--and explaining why the D1 at 2.7MP can still take a picture that can appear on the cover of a national magazine or the wall of an art gallery. Now that nearly everyone can afford a 4+MP camera if they want one, what's next?
The sad fact is that the Megapixel wars are likely to continue long after they should be put to bed--just as the car horsepower wars, the nuclear megaton wars and the computer megahertz wars all continued long after they had outlived their usefulness. It is easy to become obsessed with a statistic. Unfortunately the result is that other features and benefits get short-changed. What are the real issues facing digital photography? And what are vendors doing about it?
The number one issue for digital is overall ease of use. Film has decades of infrastructure development behind it. For a couple dollars you can get good quality images in minutes nearly anywhere in the world--with no investment in computers, software or training. You can do it while you're shopping and even with your children in tow. The magical mini-lab machine corrects dozens of possible problems and Voila!--the family memories are safely preserved.
Microsoft and Apple have taken steps to solve this problem for active computer users. Windows XP and iPhoto both offer image solutions. As you'd expect, Apple's is more complete, but Microsoft's offers a platform which can be used with your existing photo providers. But frankly, according to the latest PMA data, neither is getting large market share. Possibly because they don't really provide an end to end system from the time you open the box containing your new digital camera until you have your finished images.
Kodak has done some really good work here, since they understand more than anyone about photo consumers. But they represent only a fraction of the total D-Cam market and have only limited influence over the computer portion of the user experience. As long as buyers are buying based on megapixels and not ease of use vendors will continue to shirk this topic and cameras will wind up on closet shelves.
At the high end, Nikon has the right idea by promoting their solution as a "Total Imaging System". While we all know that some of the pieces aren't quite as magical as they may appear in the brochure, this is the goal we all want to see achieved. With film, the F5, TTL flash, studio strobes, and fast lenses were a killer combination. They were a finely tuned system which produced world-class repeatable results. Even with the best equipment digital hasn't reached that level. The need to set white balance manually and the lack of true TTL flash are two of the largest technical issues. At the low end, the sluggishness of operating a digital point and shoot compared to a film point and shoot also contributes to the low usage rates of digital.
The emphasis on pixel count ignores some real issues of image quality. First, color correction is still a nagging problem for digital images. Better and more accurate color fidelity right from the camera would be a huge boon to all users. Part of the problem is that the sensors on digital cameras don't perceive light the way our eyes do. More work is needed on better sensors and better firmware to process the sensor data.
Two companies doing interesting work in this area are Pixim and Foveon. Pixim is using an "intelligent" CMOS sensor to greatly extend the dynamic range of digital image captures. Foveon has tackled an even larger problem, by attempting to revolutionize the sensor structure itself. Foveon sensors capture all three colors at every pixel. This doesn't sound dramatic unless you realize that traditional CCD and CMOS sensor designs actually throw away 2/3 of the light reaching them, and in fact only record one color at each pixel. So those fancy megapixels we're all buying are really only 1 color each. 2/3 of your image is created inside the camera!
The sample prints from the Foveon sensor are stunning. If production SD-9 cameras from Sigma using the Foveon X3 chip can follow through on this promise Foveon will have a good chance to get one of the larger camera manufacturers to take notice and perhaps take a chance on an X3-based design. CMOS sensors, such as the X3 and the chips in some of the other new D-Cams also hold the promise of allowing true TTL flash and other technology advances that have been difficult with the less flexible CCD sensor architecture.
Photography is an art and a profession that has already spanned centuries, and individual images are expected to span decades. Silver-halide prints can last for over a century and properly stored slides can often be used for over 40 years. But computer image formats come and go. Photographers face a tough choice. JPEG is the most common standard, but it is far from archival since it does lose some information during compression. And image attributes are non-standard--most vendor specific information is hidden away in subtle and proprietary tags.
Photoshop and TIFF formats are lossless, but are very space intensive ways to store images. They also don't correctly represent vendor specific information. Then there are Raw files. These are arcane representations of the raw sensor data as captured by the camera. They are each unique and proprietary to the camera vendor and often to the specific model of camera. Vendors are promoting them as "archival", probably realizing that if photographers use these formats they will be locked into the camera manufacturer for life. Imagine having to pay a tax or buy a product from Kodak or Fuji every time you wanted to project a slide or print a negative!
Even Microsoft has fallen for this marketing speak. Despite the fact that Windows XP had to be patched to avoid destroying Raw files and that Windows XP now ignores them completely, Microsoft's web site proclaims, "A camera’s RAW format is the truest digital negative it can muster, since it contains the full range of tone and color information captured by the camera. Preserving this version of your photos in an image archive is important." Nowhere does it mention that if one of the camera companies--not known for their long term commitment to software products--abandons the format your images will be stranded.
Photo equipment vendors owe it to photographers to work together to create a truly useable open image format that would meet both the quality needs served by Raw files along with the requirement for openness and truly archival longevity. Kodak has done some stunning work in this area. Their ERI format is both compact and high quality. If it was opened up to the industry it could be the basis for a truly wonderful and lasting solution. Failing that, then JPEG2000 at least offers a range of quality and size options and can support more flexible meta-data than JPEG. It too could be the basis of a future format.