Steve Kapsinow (bio)

SK: What standards do you see developing in digital cameras?

NK: Bascially I believe the smaller APS size sensor (with the notorious focal length multiplier) really is here to stay. These sensors are much less expensive and the quality is pretty extraordinary, they will be around more or less permanently. Larger ones are much more costly because little defects can kill a sensor, and the larger the sensor area the harder it is to get one that's defect free, so the price goes up very fast. So full-frame 35mm sensors will be much more expensive.

SK: But lenses are still manufactured based on 35mm standard?

NK: Well, all the independent manufacturers and some of the camera manufacturers now make lenses for the APS size only. The advantage is that in the longer focal lengths they can be made lighter because they are only covering the APS size. Plus there are enough super wide-angle lenses for that format that makes me absolutely certain it's here to stay.

SK: Will image print size quality be affected at all by the smaller sensors?

NK: What we're seeing is that the full-frame 35mm sensors are replacing medium format cameras and the APS sensors are replacing 35mm film cameras. The compact digital cameras are replacing the traditional point and shoot cameras. In fact, some of the compact cameras do very well with 8x12 prints. Their limitation is that they get very noisy at high speeds and they may have a limited dynamic range, and even that is improving.

SK: So for all practical purposes image quality at popular print sizes will soon be a non issue?

NK: Correct.

SK: Then what about pixels and pixel sizes? Is that important?

NK: About 2/3 of my Imatest customers are manufacturers in the camera phone industry. Of course camera phones will never be as good as real cameras, but they will be quite amazing. The fundamental limitation is that when you try and make pixels really small you get a lot of noise and diffraction.

SK: So is there an ideal pixel to sensor size ratio people should be looking for?

NK: Not an ideal ratio, but the more pixels the merrier in terms of sharpness, but the quality of pixels is important as well. When you try and make pixels smaller than 2 microns, you really run into trouble with noise. Three to 4 microns is a little better, but SLR cameras typically have pixels 6 to 9 microns. Larger pixels have a large dynamic range and low noise, so you get beautiful quality pictures.

SK: But then you need to increase the size of the sensor to get more of those larger pixels?

NK: That's right, but large sensors are very expensive. Kodak is making a 39 megapixel medium format sensor that costs I think $25,000 for the complete package.

SK: So what general recommendations can you make?

NK: If you're photographing for magazines, you'll want to use a large sensor, at least a full frame for studio work and many studio photographers use the medium format sensors which produce gorgeous results. But a lot of graphic artists can do great with compact cameras because they are very sharp and the dynamic range limitations may not be as important. If you're going to take an image and do a lot of artistic effects on it, dynamic range limitations may not bother you all that much.

SK: Even with the price of digital SLR cameras falling, people still wonder if they should invest in a film SLR or buy a compact digital camera?

NK: They shouldn't waste their time on film. You can learn so much through digital—really incredible quality and instant feedback. But the bottomline is it still comes back to the artist's eye. There will be lots of technically good but pictorally god-awful images you'll have to wade through. Those with the really good eyes will still stand through.

SK: So what improvements are we going to see in digital cameras?

NK: In the smaller cameras, sensors are pretty much approaching their physical limits. Pixels are getting about as small as they can be. Where we will see increases in quality is at the higher end. We'll also see a drop in price. Features, however, will be interesting. Dave Pogue of the New York Times writes that the pixel race will be ending and among the key features will be improvements in image stabilization and advances in workflow management for wireless transfer of images from camera to computer.

SK: So what other projects besides imatest are you working on?

NK: In a couple of weeks, I'll be introducing a new program called gamutvision as a module of imatest and then as its own program, which will help teach about color management, gamma mappings—what they really mean, color spaces and what differences they make. For instance, is Adobe RGB really much better than sRGB? Actually, it's a lot better. But what's cool about this program is that you can print a test pattern using all your profiles. Scan the image in and then run it through the program and you'll get enormous detail about how the printer responds to the colors that are being sent to it. And you can compare printers, profiles, papers, inks, setting.

SK: That sound very promising. We'll have to catch up with you in a few months once you're further along with it.