PostPosted: Sun Aug 26, 2001 3:30 am
Hullo All!
Problem is, Ian knows too damn much! <g> Let's look at a hypothetical example of profiles in use, and see exactly what happens when we profile chromes. Some of this has already been hinted at in previous posts, but redundancy never bothered me before. <bg>
Imagine that I have photographed my pretty friend Cathy in her French maid's outfit, sitting nest to a Kodak 18% grey card. (She exists, so does the outfit. We are just friends. Sigh.) My exposure is perfect, and so is Cathy's. I am using my 5500K studio lights (these are imaginary) and have used three different chrome films. Emulsion A is slightly warm toned, tending to a pleasing bit of yellow in the skin tones, with lowish contrast. The effect is similar to some Agfachromes. Emulsion B is a bit contrasty, with a tendancy to bluish tones in the mids and highlights, a bit like some Ektachromes. Emulsion C is my amazing new Perfectochrome. It is perfect. Viewed on a D50 lightbox, colors, densities, etc. are a perfect reflection of the original scene. I run my three rolls through the E6 line, dry 'em and head off to the scanner.
Using my incredible knowledge, I have created proprietary software which has allowed my to profile each of these three films under my studio lighting to absolute perfection. Let's examine what this means. I will have shot an IT8 or other profiling target in the studio. The target contains known RGB values. The software "knows" what each of these values should be. It looks at each of the target patches as they are reproduced by the different emulsions, and maps them to their correct values. It then creates a series of curve functions that will cause all intermediate tones to also map correctly. (Remember, my software is perfect, like Perfectochrome and Cathy.) Thus the 18% grey tone in the original scene will map to a precise neutral, the white highlight in Cathy's blouse will map to (say) 245R 245G 245B, and the black in her velvet collar will map to (say) 5R 5G 5B. Get it? Got it? Good!
Now let's look at what all of this MEANS! I scan all three emulsions through my trusty LS 2000 (I do have one of those.) and lo and behold, the three files look (and are) IDENTICAL! Emulsion A had made the grey card a bit yellow and warmed up Cathy's skin, but the software, by means of the profile, "knew" that this would happen and "fixed" it! It also expanded the contrast range, so that white and black would map as shown above. The converse occured with Emulsion B. Now both A and B look just like C. (I simplify of course, and assume "perfect" profiles which cannot exist in the real world. Nevertheless, the actual effect of this workflow would produce essentially these results.)
Is this what we want? NO! If I use Emulsion A to photograph Cathy, it is because I WANT the particular color palette it possesses. I might also choose B to photograph a landscape, where blue sky and water needed a bit of saturation, and where increased contrast might be desired. Now here we get into philosophy. Why not use Perfectochrome all the time, and alter the tonal balance in Photoshop or Silverfast? Because it doesn't exist! And if it did, we might prefer to use Emulsion A for our glamor shots, which would put us into the color range we wanted without any software manipulation, which unavoidably discards data. Having the color balance correct on the chrome also allows us to have a "proof" for us or our service bureau to match when producing the final output. I suppose one day we will have 150 megapixel digital cameras that fit in our shirt pockets and have an infinite number of user-addressable color presets and perfect white balance adjustments. Until then, the range of color available from film is one of its distinct advantages.
Hope I don't get flamed too badly on this one...
Les
Problem is, Ian knows too damn much! <g> Let's look at a hypothetical example of profiles in use, and see exactly what happens when we profile chromes. Some of this has already been hinted at in previous posts, but redundancy never bothered me before. <bg>
Imagine that I have photographed my pretty friend Cathy in her French maid's outfit, sitting nest to a Kodak 18% grey card. (She exists, so does the outfit. We are just friends. Sigh.) My exposure is perfect, and so is Cathy's. I am using my 5500K studio lights (these are imaginary) and have used three different chrome films. Emulsion A is slightly warm toned, tending to a pleasing bit of yellow in the skin tones, with lowish contrast. The effect is similar to some Agfachromes. Emulsion B is a bit contrasty, with a tendancy to bluish tones in the mids and highlights, a bit like some Ektachromes. Emulsion C is my amazing new Perfectochrome. It is perfect. Viewed on a D50 lightbox, colors, densities, etc. are a perfect reflection of the original scene. I run my three rolls through the E6 line, dry 'em and head off to the scanner.
Using my incredible knowledge, I have created proprietary software which has allowed my to profile each of these three films under my studio lighting to absolute perfection. Let's examine what this means. I will have shot an IT8 or other profiling target in the studio. The target contains known RGB values. The software "knows" what each of these values should be. It looks at each of the target patches as they are reproduced by the different emulsions, and maps them to their correct values. It then creates a series of curve functions that will cause all intermediate tones to also map correctly. (Remember, my software is perfect, like Perfectochrome and Cathy.) Thus the 18% grey tone in the original scene will map to a precise neutral, the white highlight in Cathy's blouse will map to (say) 245R 245G 245B, and the black in her velvet collar will map to (say) 5R 5G 5B. Get it? Got it? Good!
Now let's look at what all of this MEANS! I scan all three emulsions through my trusty LS 2000 (I do have one of those.) and lo and behold, the three files look (and are) IDENTICAL! Emulsion A had made the grey card a bit yellow and warmed up Cathy's skin, but the software, by means of the profile, "knew" that this would happen and "fixed" it! It also expanded the contrast range, so that white and black would map as shown above. The converse occured with Emulsion B. Now both A and B look just like C. (I simplify of course, and assume "perfect" profiles which cannot exist in the real world. Nevertheless, the actual effect of this workflow would produce essentially these results.)
Is this what we want? NO! If I use Emulsion A to photograph Cathy, it is because I WANT the particular color palette it possesses. I might also choose B to photograph a landscape, where blue sky and water needed a bit of saturation, and where increased contrast might be desired. Now here we get into philosophy. Why not use Perfectochrome all the time, and alter the tonal balance in Photoshop or Silverfast? Because it doesn't exist! And if it did, we might prefer to use Emulsion A for our glamor shots, which would put us into the color range we wanted without any software manipulation, which unavoidably discards data. Having the color balance correct on the chrome also allows us to have a "proof" for us or our service bureau to match when producing the final output. I suppose one day we will have 150 megapixel digital cameras that fit in our shirt pockets and have an infinite number of user-addressable color presets and perfect white balance adjustments. Until then, the range of color available from film is one of its distinct advantages.
Hope I don't get flamed too badly on this one...
Les