Archived posting to the Leica Users Group, 2006/02/08
[Author Prev] [Author Next] [Thread Prev] [Thread Next] [Author Index] [Topic Index] [Home] [Search]Scott McLoughlin wrote: > Good info! My JPEG curve is the popular, updated version of > the old "White Wedding" curve. > > In a digital context, how would one "test" ISO. Excellent question. Film ISO is measured as defined by the International Standards Organization. I suggest that the following method would allow one to develop an equivalent "ISO" for digital sensors: The method that I've used to measure my own "ISO" has been to shoot a textured grey fabric with a meter at the recommended exposure and then each frame at a successive underexposure of one stop. Using the Zone system, Zone 0 ought be pure textureless black. consequently the proper "ISO" for a particular film/dev combo ought be when 5 stops of underexposure give a blank textureless negative. That is to say the exposed part and the unexposed edge have the same density (film base + fog). In digital one can likewise underexpose until the image becomes background noise. 5 stops up from this gives an "ISO" for which a Zone 0 is background noise (pure black in the digital world). By this method one could actually compare the DMR with a Canon sensor -- it may be that Leica is just being more conservative than Canon regarding the ISO ratings. Underexposing by more stops than is supported by the sensor is just like pushing a negative -- although you get a picture you loose shadow detail. Do you fault Kodak or Ilford for selling ASA 3200 films that really have a true ISO rating of 800-1000 (given the film base + fog measurement technique)? This might be the exact same situation as Canon allowing a higher "ISO" for the 20D than the Leica DMR. Has anyone measured this? Jonathan > > Tim Atherton wrote: > >> Bear in mind that the on camera histogram may be a little >> misleading - it is >> usually the histogram as applied to a to a JPEG which has had the >> camera's >> default conversions applied to it (including a fairly steep S curve) >> >> So what you see there has already dumped a lot of the information >> that would >> be in the RAW file - including highlight detail >> >> Add in - that rather like the old days you still need to tweak the >> iso of a >> digital camera (remember shooting slide film at 80 or 125 instead >> of 100? >> etc or rating 160 neg film at 125 or 100?) . A digital camera's >> 100 iso may >> actually be closer to 80 or 125 - you need to run your own tests - >> just like >> the old days :-) >> >> tim a