From: "James B. Davis" Bob writes: > >Such a basic piece of common sense yet after almost 5-y (?) of > >consumer digital imaging something that completely evades most users. > > > >Display-screens, where most people are viewing their images these > >days, work in pixels, period. > Ya, and amazingly, I don't see any grain on my monitor at 100 or so > dpi... nor should you, given the viewing distance, the video cards rendering tricks and the fact that you're used to seeing a screen made up of lots of little dots. > I still think you all would be amazed at this one image I had blown up > to 10 by 15 inches on the Fuji Frontier. The actual crop from camera > was less than 1000 pixels on the long side. that's 66 pixels per inch. Admittedly full colour dots unlike an injet printer, but still 66 pixels per inch. Let me tell you what you're looking at when you view that print - the nice people who made the print for you resized it first to 300 pixels per inch, which they must have done since the frontiers *only* print at 300 ppi, and so some algorithym be it PS, GF, Irfanview or something else was used to add pixels to your image. Lets see, you said 15 inches was the long side, about 1000 pixels so that means the short side would have neen 667 pixels or less. effectively your image was 666,666 pixels in size. the nice people ADDED 12,833,3334 pixels to your image (!) effectively they kindly stuck in 'guessed' pixels amounting to data 20x what your image contained. that's a lot of guess work, but unsharpening should make it all look quite acceptable. >No grain or pixels are > visible on the print. At viewing distance it looks like it was taken > with a 4 by 5 film camera :-) sticking point here for me. the digital revolution has been fought mostly on the grounds of 'resolution'. an unfair fight since a lot of the data 'recorded' in digital images is guessed or interpolated by software in the camera. The specifications of the CCD's show that in some instances, the actual parts of the sensor that sees light is around 25% of the sensor - 1/4 of the sensor is seeing light, the rest of the imager is made up. On top of this, there's no way a 1/2 frame image can *ever* produce an image that *looks* lifke a 4x5, simply because of the size of the light capturing area. there is a 'look', a dimentionality of larger formats that simply cannot be reproduced by smaller formats. The classic example of this was an image presented in many of the older photo magazines and books showing a die photographed with a succession of increasing formats. It was a pic that made little sense to a lot of people looking at it analytically, but it revealed the difference dramatically. basically a 35mm sensor or smaller could and would only reveal the upper surface of the die when photographed from above. Increasing to 6x6 one could see the edges of the die, almost as though one were seeing around the sides spatially. 4x5 revealed clearly the dots on the upper surface *and* the sides!~ almost a 3D effect was reproduced on a 2D media! although it's almost intangible, difficult thing to see in other less analytical photos the effect remains there, and a dimentional quality is rendered that smaller formats absolutely cannot reproduce. This contributes to the 'look' a lot more than resolution does. karl >
