Reposted when I discovered that Shel had launched a new, more relevant thread. _________________________________________________________________
> -----Original Message----- > From: Rob Studdert [mailto:[EMAIL PROTECTED] > (snip) > > My question back to you is; do you think that we ever that concerned > about burnt specular reflections in film based images? And just to be > specific I'm not talking about generally avoidable nasties like flash > problems in portraits but metal or bright white objects reflecting > full sun. Maybe a question for a new post and image? > ---------------------------------------------------------------------------- -------------------------- In the "film age" it was only ever diffuse highlights that were expected to hold some detail. Specular highlights are little mirror images of the sun, or the prevailing light source, and no sane photographer ever expected them to carry detail. The current obsession of some observers to decry any instance of 255, 255, 255 is an example of a little knowledge being a dangerous thing. Blown out specular highlights can only be avoided by deeply underexposing the remainder of the picture contents, which IMO is even less acceptable. The solution to the problem, if you see it as a problem, is the same now with digital as it always has been using film; shoot only under diffuse light where specular highlights are absent. As an aside, a few of us old farts might remember the introduction of Fuji Velvia. It was touted as a film meant primarily for graphic reproduction processes, and carried a little neutral density in the base to limit D-min. This was for good drum-scans that could accommodate overexposed specular highlights. BTW this also explains the misconception that Velvia is slower than its published speed, the results are MEANT to be darker than slides from other films that were made for projection. Because we often work with ink or pigment images we need to think in terms of halftoned images where the entire gamut of 000, 000, 000 to 255, 255, 255 is unsupportable by the media. The extreme example of this is a newspaper type halftone image that never closes up to 100 % black or opens up to 100% white, otherwise the image would show posterisation at the extremes, and because the ink coverage could break up on the printing plate. Inkjet printers are also halftone devices, although the dots are mostly too small for the unaided eye to see, and where the conversion to halftone happens within the print engine and is unrecognized by most users. The remedy to excessive output range is as simple as pulling the output levels back to what the media WILL support. A little experimentation and some trial & error will reveal what figures are suitable. e.g. IIRC newsprint halftone is 95% black to 5% white. I've found that when black is raised to 005, 005, 005 and white is reduced to 250, 250, 250 they still look black and white, but the detail close to the extremes becomes more visible. Of course YMMV. regards, Anthony Farr
