Recently, you wrote: > From: Ardeshir Mehta <ardeshir@xxxxxxxxxxxx> > (heavily snipped for brevity) > > Question: Are enlargers obsolete? > > I was wondering whether the experts on our list have a clear answer to > this question. > > My problem is this. I sometimes LIKE to see the grain in my B&W > photos. [...] > > Now even the best of scanners can scan at only at 9600 dpi. Enlarge > such a scan 100 times, and you end up with a print resolution of 96 > dpi, or in other words, a print in which the grain looks fuzzy rather > than sharp! And most scanners people can afford to buy scan at only > 4800 dpi, which is even worse. [...] > In the remote past, I have been able to use an enlarger to get some > really nice grainy results, that is to say photos in which the grain > was clearly visible and resulted in a pleasing aesthetic effect. But I > don't think it is possible to do that with a scanned image - or is it? > Just as there are differences between analog and digital image capture, there are differences between analog and digital processing and printing methodologies. We sometimes disregard that optical enlargement of images is a 3-dimensional process that has a secondary variable due to the fact that both the film and print material have grain structures. Forgive my returning to such elementary aspects of photography, but I hope that it will bring to light some of the reasons that digital and optical methods should continue to coexist. The grain structures of film and print material are 3-dimensional as grains have thickness. The secondary variable is that the grains are not arranged in a flat matrix, but are suspended in a "thin pile" for lack of a better way to describe it at this early hour. The projection of light through this pile can create a very broad range of tonality. Optical prints are created by light projected through the film's pile onto another pile whose "resolution" automatically varies with print size, print material and processing methods. There is no single representation of the image on film, only a particular instance as a result of the process. As one enlarges optically, one is creating a print with a resolution that *increases* with size. In other words, there are more grains on the print material representing the same area of the film as the size of the print increases. Therefore, the shadows created by light projecting through the 3-dimensional grain pile are more accurately represented as print size increases. Of course, other factors involved in optical projection may act to diffuse the image on the print. For scanned film, one should consider the impact of artifiacts and quantization. One artifact that may limit the usefulness of a scan is grain aliasing, where the grain in the film has been emphasized and misrepresented as a 2-dimensional element of inaccurate density. There are ways to reduce this artifact, but at the cost of sharpness. Unlike optical enlarment, the creation of the 2 dimensional pixel matrix is the final arbiter of image detail (whether direct image capture via digital camera or by scanning film). Also, quantization restricts the tonal range that can be represented by a pixel. As print size increases, the tonality represented by the matrix can result in banding, where larger areas of the print have the same density value. As print size increases, there is no more information available about the tonality as there is in optical enlargement. I believe that it is the combination of artifacts and quantization that differentiates the results of digital vs. optical enlargement that you describe as lacking "pleasing aesthetic effect". My above opinions are derived more from practical experience than theory, so I'm sure that others with a better technical background can flesh out the details and/or correct some of the representations that I've made. Best regards, Neil