Do our eyes see linear?

[carioca]

I'm currently working on achieving a perfect linear digital negative for my alternative process. I obtain a pleasing print result that is good in tonal separation of each step (but not linear) printed from my original absolute linear PS created 21 step wedge.
Somehow, this non-linearity leads me to the fact, that our eyes do not see linear grey values with the same response as a computer calculates and shows them on screen. Whereas our eyes seem to easily capture differences in the highlights (of final print), shadow areas seem to take a while for the eyes to adapt and to see the separation of steps (obviously providing the fact that they are present!)
Shadows seem to me more mysterious in an image thus, a flatter shoulder but steeper beginning in a grey curve (of the darker printed values) somehow contradicts the linearity so much sought after.
A perfect linear scale appears a bit life-less to me, some punch seems to be missing. I like good contrast images and begin to question the need of a perfectly linear curve. (or formulated differently, can a perfect linear print look strong and punchy?)

Anyone made similar observations.

Sidney

 

[E Thomson]

No, they don't. It's discussed fairly often in fora like dpreview. In fact, I believe it's accurate to say that digital sensors output linear responses, which are curved in-camera according to the manufacturer's decision on how to match our optical response. In the Nikon world at least, there is a program called Curve Surgery which can alter the .NEF* file's proprietary curve for creative reasons. This differs from Photoshop work where you are curving the curve.

*Nikon Exif File, Nikon's Raw file format.

Edit: In addition, Nikon dslrs will load a "custom curve" which replaces the default .NEF curve, usually for purposes of brightening mid-tones or similar. But you could theoretically replace it with a linear response. Google "fotogenetic" for a good site on the subject, or search for, say, "D70 custom curve."

 

[Ron-san]

I'm currently working on achieving a perfect linear digital negative for my alternative process. I obtain a pleasing print result that is good in tonal separation of each step (but not linear) printed from my original absolute linear PS created 21 step wedge.
Somehow, this non-linearity leads me to the fact, that our eyes do not see linear grey values with the same response as a computer calculates and shows them on screen. Whereas our eyes seem to easily capture differences in the highlights (of final print), shadow areas seem to take a while for the eyes to adapt and to see the separation of steps (obviously providing the fact that they are present!)
Shadows seem to me more mysterious in an image thus, a flatter shoulder but steeper beginning in a grey curve (of the darker printed values) somehow contradicts the linearity so much sought after.
A perfect linear scale appears a bit life-less to me, some punch seems to be missing. I like good contrast images and begin to question the need of a perfectly linear curve. (or formulated differently, can a perfect linear print look strong and punchy?)

Anyone made similar observations.

Sidney

Sidney-- I think you are missing the point. Once you get an image on the computer screen so it looks good to you, all we ask of the computer-printer-sensitized chemicals on paper is that they translate the tones we liked on the computer exactly to the same tone on the print. That is what we mean by "linearity" in the digital neg business. It has nothing to do with how our eyes view tones or whether or not our eyeballs are "linear" devices. We just want the machines to talk nicely to each other and not throw in any tonal changes of their own making. If your dig neg system is truly linear and you do not like the final print, then you did not make it right on the original computer screen.
Cheers, Ron-san

 

[Neil Poulsen]

What Ron-San said.

This works, because we're dealing with monochromatic color. In full color, linearity would have this condition for each of the printing colors, C, M, Y, and K.

As another example, in Ansel Adams zone system, zones progress in roughly even steps to our perception. Yet, each zone is twice the light as the previous. So, our perception of evenness progresses in a geometric versus a linear fashion.