35L II > If every lens could be this good

J A C S wrote:

hermit383 wrote:

Tony Parenti wrote:

Dredging this thread up and I have to agree with Hermit, the 35L II just doesn't not render as beautifully has the original 35L. I highly regret unloading the original. Yes, it may be sharper edge to edge but the original 35 had such a beautiful rendering quality to it that can't be explained by charts and specs.

I think MTF charts may be limiting in measuring lens performance to the eye because of the MTF chart's nature: It measures " black" or "white" lines " yes " or "No" - The computer "Can"- "Can't" read the lines any further.

Actually, the tests read a blurred image of a sharp edge.

I think we need a test that incorporates fine tonal gradations, and instead, tests the difference in contrast between two very similar tones/graduations in tones.

This is a highly nuanced, and challenging topic to explore with quantifiable scientific data. However, I do believe it is worth pursuing the quantification of these attributes. At worst, it is wasted effort. At best it adds further to the artists ability to create meaningful images.

*I really try my hardest as a scientist by career to approach this the most factual and scientific way possible. *

This is where the challenge is: Being factual, and showing quantifiable data that support the hypothesis.

The Hypothesis: There are lens attributes other than ones measures by current MTF charts that meaningfully affect overall image quality and perception of aesthetically pleasing images.

Sure.

It is tricky because, as of right now, the quantifiable scientific data we have seems to miss some of the important lens attributes we see with our eyes.

This is where the challenge is: Being factual, and showing quantifiable data that support the hypothesis.

I will list some potential reasons ( *in my opinion*) this may be missed by current MTF charts:

1.) Due to being measures by computers in a 0 or 1 fashion, computers aren't programmed to examine "how clear" the lines that can be read are. When attempting to read the most fine amount of lines per millimeter possible, it comes to a point where it simply says either "Yes" or "no", as in: "No further lines can be read by the computer"

This is a gross misinterpretation of how the MTF tests are performed. There are several ways but the most common one is the slanted edge. It has nothing to do with counting lines.

Thanks for commenting. Not to sound like a cop out, But i totally forgot to add in here a section/part about people who know more about MTF charts to add in where/how I am wrong/ misinterpreting and correct me.

2.) Due to the nature of black lines against a white paper, and being only 2 "colors" this takes away the nature of reading tonal differences between other color.

3.) Most importantly ( *in my opinion*) Global contrast differences, EG: Black line against a white paper.

It could be done but in my understanding it is not done with the commonly used algorithms.

it would be cool to see this done in some way.

Versus: Microcontrast differences. This could not be accurately tested with either "black" or "white" lines. This would require difference shades very close to each other, and the ability of the lens to differentiate between those smaller increments of tone.

Also, I've changed my tune on this lens in particular. I think I would maybe very slightly still prefer the original a bit more for my use, But I would be really torn though considering the new versions performance at wide apertures when outside the centermost circle.