I'm done with lens reviews

Truman Prevatt wrote:

KariP wrote:

Truman Prevatt wrote:

Jeff Biscuits wrote:

Truman Prevatt wrote:

Microcontrast

Actually it is quite real.

Fill me in in one thing, though…

I’ve done a limited amount of image processing in my career which has involved analysing local contrast/gradient. So “microcontast” as that concept of contrast within a small area of an image (whether analog or digital) makes complete sense.

Where it doesn’t make sense to me is when it comes to lenses. It’s a term that’s explicitly differentiated from just “contrast”, and I can’t get my head around how a lens can provide microcontrast to any different degree than it would provide “macro contrast”. In other words, how could a lens have the capacity to alter tonality just because of the proximity of their tones? It seems obvious that a lens can give more or less contrast, but surely there is no way two lenses with the same macro contrast can differ in terms of microcontrast.

So I’ve always assumed that in the context of lens reviews, microcontrast—though a perfectly real and measurable thing—is used in a somewhat mythical way, because in terms of a lens transmitting light, it’s just contrast.

Or is there something I’ve completely overlooked?

Every component of an optical systems reduces local contrast. Film reduces the local contrast of a scene. An outdoor scene can have as much as 20 stops of almost instantaneous contrast change. No film nor digital sensor can replicate that. In film days one had to chose the type of developer. D76 for example contained silver solvents. It was good for reducing grain but it made the edges smoother with slower roll of so reduced the micro contrast. Acutance ( a term for edge sharpness) developers like Ordinal produced negatives with much better micro contrast (sharp edges) but also made the grain more apparent since the hard edges were maintained.

Color filter arrays reduce micro contrast as the edges will be derived by interpolation from neighbor hooding detectors with different color. Cross talk in the sensor will cause a bleeding current across into neighboring detector hence reducing the contrast across the edges and smoothing the edges somewhat similar to the effect of silver solvents in developers.

The final component which may have the biggest impact is the lens. Lenses don't create micro contrast. No component of the imaging system create micro-contrast. That is a function of the scene. Any distortion will show up across edges as an effect of reducing micro contrast. Some aberrations are worse that others. Chromatic aberration resulting from the wavelength dependence on the refractive index will show up across sharp edges. Fixing one aberration can often result in making others worse.

https://www.lensrentals.com/blog/2010/10/the-seven-deadly-aberrations/

In the image they all show up on sharp boundaries between dark and light - the edges. When I ended up having to worry about R&D for both our signal intelligence/electronic warfare systems and optical systems I have to get an immediate education. The consolidation made sense as the big end image processing that the move to digital cameras for ISR was now simply the two dimensional version we did for signals intelligence and EW. For example an edge in an image is detected using the same basic approach as a radar pulse in detected in an EW system. But there is a lot more subtle to the imaging systems. Lens designs - we specified our lens designs to people like Kodak, Corning, etc. Different lens designs and glass formulations were needed for a camera flying at a nominal 30,000 feet on an airborne platform, from one flying on a U2 at 70,000 feet from one flying on the SR71 "blackbird" at 90,000 feet from one flying in space. But the goal was always the same - minimize the micro contrast degradation of the lens. In the 1960's NASA often contracted with Zeiss to design and build their lenses. Some attribute the "Zeiss pop" to the the fact these lenses degrade the scene micro contrast less than other lenses. That was important to NASA for the exploration of the moon during the 1960's and '70s. Often we would trade other properties of the lens to maintain high micro contrast.

So I expect that when people refer to their opinion that lens A has better micro contrast than lens B, what they mean is Lens A reduces micro-contrast to a less extent than lens B.

There is no perfect lens. Fourier optics tells us that. The very fact that there is an aperture between the scene and film/sensor impacts the recorded image. The point spread function tells us exactly what a point source looks like. The fact that all light rays landing on the sensor are not collimated and the interaction with the lens impacts the image. The fact that the index of refraction of all glass is greater than 1, impacts the image. The fact that there are air/glass boundaries (to per lens element) resulting in both reflection and partial reflection impacts the image. The fact that some of the photons scatter of the air/glass boundaries in random directions results in impacting the image. The major impact of all these things results in reducing the local or micro contrast.

Zeiss and Leitz both got their start in designing microscopes that could increase the contrast between the subject and background so people could see the objects better. While there is a difference in the illumination sources the problem in a microscope and camera vs. a lithography systems for the production of semiconductor devices - the problem is the same. That is to maintain sufficient local contrast to maintain the necessary detail at the edges.

Yes,

i have read these facts many times and I think I understand what this is about - there is a measurable thing. BUT ! are basically and normally made good lenses really much different - somehow visually different . I have very (or too often ) often read how people can "see" in an image on a PC screen, that some lens (usually the expensive ones they own ) clearly has more "micro contrast" . I really doubt these observations. How can anyone see some fantastic micro contrast on a rather low resolution screen?

Interestingly enough the best way to see edge contract for a human is on an emissive display at a pixel to pixel mapping because the back lite makes the contrast stand out more to the eye. This is the same as the light box to help people evaluate their negatives. I sure would not through out the issue of the "I spend a lot of money on this puppy so by golly it better be good" bias - that's always a problem.

It is much more difficult to see the difference off a reflective display like a print. Clearly an optical bench with controlled illumination and the appropriate sensor to capture the output of the lens is the proper way. The only place I know that does this and basses tier lens results is lensrentals.com.

Today the major differences between the top of the line consumer lenses like Zeiss Batis, the Leica ASPH line, etc., some Fuji cinema lenses and the next tier down is coatings. The top lenses use very expensive nano-crystal coatings to mitigate reflections at air ground boundary and random angle scattering.

https://photo.stackexchange.com/questions/22151/what-does-a-nano-crystal-coat-do-on-this-lens

Can a person actually see the difference at 100% on a monitor? Some might the same way they can see the difference of the negatives by using a high quality light box and high quality magnifying lope or a high quality grain magnifier that is magnifying on a sharp edge in a negative. I know it is very easy to see for example that dilute Rodinal produces much better edge sharpness (micro contrast) than say a silver solvent developer like D76 with a grain magnifier. So on some lenses at 100% on a good display - some people might be able to tell while others would be all wet. However, inputting a 16 bit TIFF into a program like Matlab or Mathematica or many others, it could become pretty easy to determine. The question then becomes - "does it make in visual difference" in the image hanging on the wall?

The facts and science is OK , but I do not believe in people who claim that they clearly see micro contrast . And some decades ago I had too look in to a microscope quite often

Expensive microscopes were better.... and some binoculars are much better. And some really bad lens was really bad and I had to return it. All other lenses have been OK

Yes

I'm sure it is possible to see subtle differences in 100 or 200% view on my 27" iMac screen - I even believe I see something if I seriously try . But the lens comparing sites and reviews should mention money more often. A 3500€ cinema lens looks better in 200% view than a 750€ prime lens. But how much better ? I do not want to pay the price for a thing I cannot see in a picture.

I did some experiments with D76 and perhaps Rodinal (or something) some 40 years ago and I tried to adjust exposure and other thins to get best possible negatives. Not sure if it was worth the effort . Today we can just add some contrast , sharpening , noise reduction etc etc - and the results are great if the lens is ok and there is no motion blur

A friend inherited a rather expensive and large wine cellar from a relative who was really more than a basic hobbyist. My friend  knew something about wines , but really did not know what to do with all those too expensive bottles he inherited - and took a scientific course in Helsinki University ( not sure why there was such course...) Anyway, it was scientific and there are many methods to "read" and taste wines - and there are really many laboratory methods. The lecturers gave the final lectures and at the end of the course it was just a couple of really determining things left. Is it the price or age of the wine or what is the  real thing ? ... when you enjoy wine in a very good company of good friends , if the food is OK and if the place and the view from the windows is nice - the wine tastes much better. And if someone plays some nice music with real instruments and the discussions get more interesting turns , the wine tastes extremely good and many want to buy a load of it to take the great wine home.

After 2 weeks the wine  you bought is nothing special if you taste it at home alone .