On Sun, 7 Mar 2004 Michael Shaffer wrote > Tony writes ... > > On Sat, 6 Mar 2004 Michael Shaffer" wrote: > > > ....... the dynamic range for negatives is said to be > > > greater than for chromes, ... e.g., 11-13 f/stops versus 9-10. > > > > Michael > > > > Maybe veering slightly off-topic, but with respect, can you > > give any references for this. I've also seen people stating > > 14 stops for col neg on this list. I always thought transparency > >could record 5 stops (32:1), B/W neg just over 7 (160:1), and > > col neg 8+ (256+, more whenpre-flashed). Maybe > > I'm confusing dynamic range with something else! > > If you visit tha Scantips wwwpages, (e.g., see > http://www.scantips.com/basics14.html ...), > who has exposed his scanning tips to the film scanning newsgroup for > years, ... even he is likely to confuse terms ... "optical density", "dynamic > range", ... "dynamic range", "Photographic range". His numbers are as > well more conservative than mine, and less than yours. He also explains the > connection between OD and scanning bit depth better than I did, but I > still haven't anywhere seen a good analogy which allows everyone to realize the > relationship. Definately including me. > If you google-search for ["dynamic range" film], you'll get many hits, > including those who speak of theoretical limits of film, ... e.g., see: > http://www.ltlimagery.com/dynamic_range.html
Thanks Michael for the sites - they are both useful, and I understand them, but I'm still confused! I can see how the figure for 'dynamic range' of film is derived - difference between max and min densities in 0.3 steps. The problem is this doesn't indicate what range of light values in the scene the film can record, I think this is my point. The opposite seems true in fact (higher contrast film like tranny, with higher max density and 'dynamic range', compresses the reproduction of the scene's tonal range far more than say col neg, which will compress less and record more). And I'm discounting for now any flare in the system, which has a big effect. Sensor and film response are both analogue, but one is linear (discounting noise for now!) and one is not. Then the former derives digital values over say a 16 bit range. Then it will often give you less than that in the file. Sorry about the rambly post, I'm thinking on the keyboard, and may have answered my own question. Best wishes Tony Riley (the Kendal one). =============================================================== GO TO http://www.prodig.org for ~ GUIDELINES ~ un/SUBSCRIBING ~ ITEMS for SALE
