The concept of micro-contrast

[jbielikowski]

MTF charts are cool but too detached and mathematical for my taste, best way to evaluate a lens is by putting it on a lens projector so you can observe it all: contrast, resolution, aberrations, distortions... The next best way is to have a good quality printed test chart:

Micro-contrast can be easily observed on USAF test target parts of the chart (those three bars getting smaller and smaller). So if you have a tiny digit with tiny bars and you can see it on a close-up then it's resolution. When you look at the whole image and those bigger bars are clearly black-and-white then it's contrast. But if those tiny digits and tiny bars are crisp (easily readable, with deep blacks and clean whites) on a close-up then it's micro-contrast.

Low-contrast/high-resolution lenses will have plenty of fine detail when observed on high magnification but overall image could look less sharp than a high-contrast/low-resolution lens. Some lenses even have different focus points of sharpness and contrast.

It's like looking at the distance on a clear winter day, you can count crows on a tree miles away.

 

[peterm1]

^^^^^ Agreed, some lenses have some magic in them in how they resolve. These are the lenses I prefer and seek out. HB is the easy way, but costs. OTOH we can get lucky with inexpensive lenses. It is kind of like horse racing, follow the bloodlines. I am fond of Sonnars and have had luck with them. And I have had good luck with some recent Chinese retro lenses, reasonably priced. So for people with eyes sharp enough to see the micro-contrast there are bargains around. Even the occasional US$100 Jupiter can be great. To quote Fats Waller, "One never knows, do one?"

What comes next is the question of degree of micro-contrast. I'll sidestep those discussions and be grateful for those lenses I have stumbled across.

I too love Sonnar rending. Other than that, it's sometimes a crapshoot as you say.

EDIT: I should have added that I have also had good luck with various old Schneider lenses from the 1950s/60s. One such lens, a 105mm f3.5 Xenar (essentially a Tessar style - 4/3 elements/groups ) is particularly sweet and it occurs to me that if any lens can be said to have excellent micro contrast it might be this one.

I recall reading an article many years ago in a very old photo magazine in which the concept of "plasticity" of an image was discussed which if I recall correctly was an old term sometimes applied to images which were sharp (ish) - certainly sharp enough for excellent portraiture, where they excelled - but nevertheless had a certain softness or "plasticity" or "roundness" imparted by the quality of the tonal and color gradations. Terms they used back in their day for thjis phenomenon. BTW if you can access a copy, see Ivor Matanle's "Classic Cameras" book at page 21 where there is a lovely photo he describes in this manner - the photo being taken with a Xenar, albeit on a medium format TLR.

All of which sounds suspiciously like my Xenar lens and most Sonnar ones I have used too. And also sounds, perhaps, to be what others are describing here as "micro contrast" - especially as these lenses have a low element count optical formula (a feature which is sometimes said to be conducive to good micro contrast unlike some modern, highly corrected glass with millions of elements.)

An example made with the Schneider Xenar 105mm f3.5. (I also have a Schneider Xenar 135mm f3.5 which I think renders in a similar fashion, but I cannot comment on their tele Xenar lenses as I have not tried them.)

Cafe Study 29 by Life in Shadows, on Flickr

 

[Freakscene]

Low-contrast/high-resolution lenses will have plenty of fine detail when observed on high magnification but overall image could look less sharp than a high-contrast/low-resolution lens. Some lenses even have different focus points of sharpness and contrast.

Contrast and resolution are only a trade-off where spherical aberration is limiting. In that situation large scale and fine scale MTF tend to be impossible to achieve at the same time. Modern lens design and manufacture have virtually eliminated spherical aberration and lenses can be high contrast and high resolution - functionally they can have very high mtf for large and small structures.

 

[boojum]

This is all fine for you folks whose degrees end in "E" or "S" rather than "A". But for me MTF charts and test images are a language as undecipherable as a Maya Codex or Trans-Carpatho-Ruthenian. I have to fall back on, "Oh, look at the smooth shift of color" or "Oh, look at the smooth shift of light." I appreciate the knowledge and discussion but somehow get the feeling that I walked into the wrong class to audit. ;o)

I can only fall back on how the image looks. I cannot tell you why or how it happened to look that way. But I enjoy it no less, regardless. The old watch analogy comes to mind and I mean no disrespect. You guys are streets ahead of me, but as the adage goes, "I asked what time it was and was told how to build a watch." I am just trying to figure out what it means when the big hand is on the seven. ;o)

 

[peterm1]

This is all fine for you folks whose degrees end in "E" or "S" rather than "A". But for me MTF charts and test images are a language as undecipherable as a Maya Codex or Trans-Carpatho-Ruthenian. I have to fall back on, "Oh, look at the smooth shift of color" or "Oh, look at the smooth shift of light." I appreciate the knowledge and discussion but somehow get the feeling that I walked into the wrong class to audit. ;o)

I can only fall back on how the image looks. I cannot tell you why or how it happened to look that way. But I enjoy it no less regardless. The old watch analogy comes to mind and I mean no disrespect. You guys are streets ahead of me, but as the adage goes, "I asked what time it was and was told how to build a watch." I am just trying to figure out what it means when the big hand is on the seven. ;o)

I know how you feel. 🙂

Worse than that I am really not terribly interested in the test charts and MTF charts etc. (although in the latter case I have a rudimentary understanding of how to read one - sort of). But I am apparently of a more artistic temperament when it comes to photography and do not really give a toss about those other things in practice. The "look" of an image is all that matters to me.

 

[boojum]

I know how you feel. 🙂

Worse than that I am really not terribly interested in the test charts and MTF charts etc. (although in the latter case I have a rudimentary understanding of how to read one - sort of). But I am apparently of a more artistic temperament when it comes to photography and do not really give a toss about those other things in practice. The "look" of an image is all that matters to me.

This is one of the great strengths of RFF: it has the spectrum of folks, from poets to physicists. And we all show up here and share what little or much we know and somehow get along. I think the relaxed guidance, collegiality of the members and the unwillingness to tolerate trollers helps a lot. It is a good board.

I have seen one like it, the old Gearslutz - new name now - had a Remote and Location recording forum that had the same feeling. The fellow who ran the forum had a general hands-off policy but was firm with serious transgressions. It was great. Another, as here, where I learned a ton.

 

[boojum]

I too love Sonnar rending. Other than that, it's sometimes a crapshoot as you say.

I have that lovely '57 KMZ J8. For some strange reason that one resolves very well in color and contrast and a 3D effect. I rolled a seven on that one and made my point.

Most of my lenses are good and happened through the counsel and assistance of the folks on this board. You know who you are.

 

[peterm1]

I have that lovely '57 KMZ J8. For some strange reason that one resolves very well in color and contrast and a 3D effect. I rolled a seven on that one and made my point.

Most of my lenses are good and happened through the counsel and assistance of the folks on this board. You know who you are.

I have one too. I am not sure of its date of manufacture but it is very good. 🙂

 

[boojum]

I have one too. I am not sure of its date of manufacture but it is very good. 🙂

The first two digits of the serial number are the year it was manufactured.

 

[boojum]

I have seen discussion on the differences and maybe validities of film and digital micro-contrast. This is interesting. There is that long running discussion of the differences and merits of each medium. And I do not have the tech chops to say that it is the gezoniblab whisltefist that causes the differences. And then there are film variances as in types and ISO's and in digital it is the sensors evolving and CCD vs CMOS vs CMOS BSI.

So a definitive answer about whose micro-contrast is the better is shooting at not one but two moving targets. I think that part of the preference may be habit. We are used to, accustomed to, habituated to one medium or the other. And that medium is our "gold standard." This, in digital is the M9 sensor being most Kodachrome-like. But is Kodachrome the best or the one we prefer? Likewise analog vs digital micro-contrast. And I do not think we have agreed on what micro-contrast is. Sheesh, I need another cup of tea.

 

[pgk]

Low-contrast/high-resolution lenses will have plenty of fine detail when observed on high magnification but overall image could look less sharp than a high-contrast/low-resolution lens. Some lenses even have different focus points of sharpness and contrast.

It works like this: when fine detail in the subject is of low contrast then only a lens with high contrast and high resolution will be able to project the image of this onto the sensor where it can be recorded as microcontast information (resolved fine detail) unless it is of too low contast ("<10% MTF"). High contrast lenses are able to resolve contrast better than low contrast ones because low contrast is caused by flare and residual uncorrected aberrations. So for resolving micro contrast (ie imaging fine detail in a way that it can be appreciated in the final image) you need to use a high contrast, high resolution lens. There is no 'magic' involved; lens designers use MTF as a predictive tool to analyse designs before finalising them. There are always trade offs making lenses, the main one of which is probably cost, based on precision of manufacture and quality control. The price of cine lenses reflects their high cost of manufacture, whilst cheap lenses tend to vary in performance due to poor quality control (although some are obviously well designed because some samples work well).

FWIW I took the same scene (near enough) a few days ago on an 1865 lens (two doublets) and a modern, more complex aspheric lens. At 'small' size (laptop screen say) they both look fine. Enlarge them and its obvious that the old lens loses resolution quickly. The modern lens undoubtedly has more detail than the sensor can record and records micro contast whilst the old lens cannot.

 

[peterm1]

I have one too. I am not sure of its date of manufacture but it is very good. 🙂

The first two digits of the serial number are the year it was manufactured.

I had totally forgotten about Russki serial numbers and dates. Just looked at mine. It's later than yours - 1975. Which I kind of expected because it is black.