[Chris Saper]

Hi Steven,

What an interesting discovery. It may be that the digital photograph escapes the value compression problems of negative (especially) based film, and I would love to know whether this is the case.

However with print film, what the eye may see as a range of 9 values is typically compressed onto the negative by about a third. The darkest four values are the most severely compressed, representing the equivalent of about 1-1/2 visual values. The lightest lights suffer the same fate, but not quite so severely. When the print is made, there is a little bit of relief in the middle value compression, but very little at either end.

So when you work from a conventional film print, your initial source has "clumped" the darks as well as the lights, and gives you a false reading of the distribution of values in the picture. If you then rely on this distortion to paint the values in the painting, and then photograph the portrait with film, you will have compounded this value compression yet again. That is the major problem with "second generation" photos.

I don't know whether the value compression happens with digital images, and plan to do a little research to see what I can find.

I'm not sure whether this explanation is intelligible or not, but I don't have permission to post the chart that shows this so beautifully (Eastman Kodak Company, "Copying and Duplicating: Photographing and Digital Imaging Technique", 1984, page 2). I'll hunt online for it to se if I can reference it here.

I look forward to seeing the revised portrait!

Nice to have you back.

p. s. Harley Brown's mantra (one of many, actually) is this: "You can use any color as long as the value's right, but remember that the color can't be right if its value is wrong."