• Log in with Facebook Log in with Twitter
Photo

About contrast


  • Please log in to reply
12 replies to this topic

#1 stoppingdown

stoppingdown

    Advanced Member

  • Members
  • PipPipPip
  • 530 posts
  • LocationGenoa / Milan, Italy

Posted 25 July 2017 - 01:26 PM

Ok, I know this will be controversial, but... also interesting. I hope.  :)

 

I'm starting from a Klaus quote from the latest post about the 500mm:

 

If you have fine tonal variations in feathers or so that'll make a difference as well - resolution isn't everything.

But I've opened a new thread because it's a generic thing.

 

Today, is contrast of a lens still important, in fine-grained terms? I'm explaining what I mean. Let's imagine a fictional classification of lens contrast from 0 to 10. Clearly a lens with a score of 9 is better than a lens with a score of - say - 6. Such a different performance can't be compensated in post-production, because the amplification of contrast would produce coarse and not fine tonal variations. Ok.

 

But if I compare two lenses and they score 9.5 and 9, is it really something that can't be compensated in post-production? 

 

 

A clarification of my question. We have micro-contrast and contrast. Personally I understand micro-contrast is more important "in the lens", because trying to enhance it in post-production might create some "artificial" look. I'm thinking, for instance, of what happens when you overdo with "Clarity" in Lightroom.

 

But for the normal, overall contrast? Here I don't have a clear answer. If anybody has one, in either sense, does he have some evidence, I mean some test-case with comparisons?

 

PS Clearly I'm thinking of RAW post-processing, with plenty of bits...


stoppingdown.net

 

Sony a6300, Sony a6000, Sony NEX-6, Sony E 10-18mm F4 OSS, Sony Zeiss Vario-Tessar T* E 16-70mm F4 ZA OSS, Sony FE 70-200mm F4 G OSS, Sigma 150-600mm ƒ/5-6.3 DG OS HSM Contemporary, Samyang 12mm ƒ/2, Sigma 30mm F2.8 DN | A, Meyer Gorlitz Trioplan 100mm ƒ/2.8, Samyang 8mm ƒ/3.5 fish-eye II | Zenit Helios 44-2 58mm ƒ/2 
Plus some legacy Nikkor lenses.

#2 wim

wim

    Advanced Member

  • Members
  • PipPipPip
  • 1,145 posts
  • LocationMaastricht, Netherlands

Posted 25 July 2017 - 05:07 PM

You have to be careful when discussing contrast and microcontrast. Essentially, contrast is about dynamic range, and/or the contrast slope in a picture, and microcontrast is indeed about the ability to determine the difference, or rather, borders, between two different lines, adjacent to each other. And in the latter case, there are more things at issue, namely resolution, and MTF percentages.

 

In the past, we used Rayleigh criteria as the point to determine where we could still distinguish between two adjacent lines, i.e, contrast of 9%, whereas most modern lens reviews use MTF-50 or 50%. That is high contrast, and one can generally get a lot more from a photograph than the almost binary approach presented in reviews.

 

IMO, if you do have a lens with a decent resolution, you have a lot of headroom to play around with to get optimal microcontrast, i.e., plenty of details, without affecting overall contrast too much. Essentially, you'd have to determine which areas you find important, and process those accordingly, IOW, do not process the entire photograph as a whole, but do so for individual areas for optimum contrast and microcontrast.

 

Effectively, this more or less boils down to applying a Zone-type system to different parts of a photograph, with burning and dodging. Similarly, you could selectively up "clarity", or just make the colour more vibrant in specific spots. This is why I really like the (Google) Nik software tools (free). You could, f.e., treat a bird's feathers differently to the surrounds or environment, and that really works well.

 

All you need to do with your RAWs is to process them for an optimal DR, as you see fit, and then process them as mentioned above - in the end it is what you as a photographer see in a picture, and how you see the end result.

 

It is a bit more work than following a standard process, but to me it is worth it, especially for images I want to print (DR of only 6 in print).

 

For test cases I can only look once I have my desktop up and running again after its registry got badly corrupted by a software update, I am afraid. It'll take too much time otherwise.

 

HTH, kind regards, Wim


  • dave's clichés likes this

#3 stoppingdown

stoppingdown

    Advanced Member

  • Members
  • PipPipPip
  • 530 posts
  • LocationGenoa / Milan, Italy

Posted 25 July 2017 - 05:47 PM

Wim, I actually apply microcontrast with a brush, in Lightroom. There are a numbers of photos in the past where I did that in the wrong way - and also had a miscalibrated monitor - and I'm slowly re-process the old photos.

 

 

But the core of my question was not about the better technique for post-processing; rather whether they, if applied correctly, can compensate some differences in the native IQ of lenses.


  • dave's clichés likes this

stoppingdown.net

 

Sony a6300, Sony a6000, Sony NEX-6, Sony E 10-18mm F4 OSS, Sony Zeiss Vario-Tessar T* E 16-70mm F4 ZA OSS, Sony FE 70-200mm F4 G OSS, Sigma 150-600mm ƒ/5-6.3 DG OS HSM Contemporary, Samyang 12mm ƒ/2, Sigma 30mm F2.8 DN | A, Meyer Gorlitz Trioplan 100mm ƒ/2.8, Samyang 8mm ƒ/3.5 fish-eye II | Zenit Helios 44-2 58mm ƒ/2 
Plus some legacy Nikkor lenses.

#4 Klaus

Klaus

    Chief Editor

  • Moderators
  • 5,372 posts
  • LocationSydney, Australia

Posted 26 July 2017 - 12:09 AM

I was already afraid that the word micro-contrast might trigger a new discussion. ;-)

 

I think the topic has multiple dimensions (the sensor/AA filter also plays a massive role for instance). However, at the end of the day - if micro-contrast isn't there, it just isn't there. You cannot recover lost information. You can, of course, ease the issue to a certain degree which in turn is a question whether the base situation is "good enough". From there on we are having an opinionated discussion. :-)

 

The other aspect is ... when does it matter anyway? I'd state that it isn't overly important in many scenes because the world is a rather contrasty place to start with. The discussion started with birds. If you have a black or white bird, that's about as challenging as it gets. Having micro-contrast or not means the difference between a "plastic bird" or the real thing. Of course, if we are talking about the "real thing" further aspects will come up like e.g. moiree. 


Chief Editor
photozone.de

#5 stoppingdown

stoppingdown

    Advanced Member

  • Members
  • PipPipPip
  • 530 posts
  • LocationGenoa / Milan, Italy

Posted 26 July 2017 - 10:44 AM

I think the topic has multiple dimensions (the sensor/AA filter also plays a massive role for instance). However, at the end of the day - if micro-contrast isn't there, it just isn't there. You cannot recover lost information. 

 

 

 

Agreed. But:

 

Having micro-contrast or not means the difference between a "plastic bird" or the real thing. 

 

 

 

This looks like it's related to my "6 vs 9" scenario. But what about "9 vs 9.5"? It's true that the missing 0.5 isn't there and can't be recovered. But it's like the unsharp mask: it can't recover lost sharpness information, but can give some impression that there's a tad more sharpness. One can tell the difference by pixel peeping, but what about the overall image appearance?

 

Clearly, if I can reason with an infinite budget I don't put myself the problem: I'd always buy the best lens, whichever the price. In the real world that "9 vs 9.5" difference might imply several thousand euros. 


stoppingdown.net

 

Sony a6300, Sony a6000, Sony NEX-6, Sony E 10-18mm F4 OSS, Sony Zeiss Vario-Tessar T* E 16-70mm F4 ZA OSS, Sony FE 70-200mm F4 G OSS, Sigma 150-600mm ƒ/5-6.3 DG OS HSM Contemporary, Samyang 12mm ƒ/2, Sigma 30mm F2.8 DN | A, Meyer Gorlitz Trioplan 100mm ƒ/2.8, Samyang 8mm ƒ/3.5 fish-eye II | Zenit Helios 44-2 58mm ƒ/2 
Plus some legacy Nikkor lenses.

#6 wim

wim

    Advanced Member

  • Members
  • PipPipPip
  • 1,145 posts
  • LocationMaastricht, Netherlands

Posted 26 July 2017 - 11:07 AM

Unless you pixelpeep, you won't be able to tell the difference between a 9 and a 9.5, or a 9 and 10 for that matter. I even doubt you'll see the difference between an 8 and 10, unless you make huge enlargements.

 

Also, for best image appearance, you do need to optimize it for the medium you are using to show it. When you do that, for smaller viewing sizes, I think you can make a 6 look like an 8 or 9.

 

HTH, kind regards, Wim


  • stoppingdown likes this

#7 dave's clichés

dave's clichés

    Advanced Member

  • Members
  • PipPipPip
  • 1,621 posts

Posted 26 July 2017 - 12:36 PM

 

The other aspect is ... when does it matter anyway? I'd state that it isn't overly important in many scenes because the world is a rather contrasty place to start with. The discussion started with birds. If you have a black or white bird, that's about as challenging as it gets. Having micro-contrast or not means the difference between a "plastic bird" or the real thing. Of course, if we are talking about the "real thing" further aspects will come up like e.g. moiree. 

  That was a point that I nearly spoke about, what seems on a 100% crop like mediocre contrast, in the end is of little consequence when looking at the whole picture. An enemy for me is this super bright contrasty sunshine here, even when you haven't blown out the highlights getting the whole dynamic range in the final image is not always an easy task!

  The dynamic range of a black and white bird in bright sun is phenomenal, demanding RAW PP (adjustment brush) to curb the whites and boost the dark areas.  



#8 you2

you2

    Advanced Member

  • Members
  • PipPipPip
  • 993 posts

Posted 26 July 2017 - 02:41 PM

micro-contrast is important. As Klaus indicated the amount of micro-contrast is a combination or sensor and lens (new tests?). One of the fame of zeiss lenses is that many of them have very high micro-contrast. I don't know if this is due to optical design; lens coating or myth. My personal opinion is that it is more than myth but not unique to zeiss. Part of the issue is that in a non-control test overall contrast plays a role in how you judge a picture as well as  lighting.  I'm not sure if it shows up in current mtf but I presume that the lens portion could be measured. I think folks on the contax forum said it showed up in mtf at some point but then I was confused between micro-contrast and contrast.

-

I will say that when I switched from olympus to contax I believed I noticed  a difference. Certainly I used the same film with my (then infrequently used olympus kit) as with my contax kit and certainly the zeiss lenses had higher contrast but I imagined that I frequently noticed significantly greater tonal range. My olympus kit was small (mostly vivitar 90f2.5 and 50f1.8 with an occasional vivitar 28f1.9 (which was a snazzy lens)); since my contax (aria and 6 lenses) was acquired 20 years after the olympus kit I had $$ to buy better lenses but even the 50f1.4 as well as the 35-70f3.4 showed what I presumed was significantly greater tonal range. The 90f2.5 was a fine lens but it had lower contrast and I believe noticeably less micro-contrast.



#9 borisbg

borisbg

    Advanced Member

  • Members
  • PipPipPip
  • 444 posts
  • LocationLos Angeles

Posted 27 July 2017 - 06:47 AM

Can this discussion be supported by sample pix?
  • stoppingdown likes this

#10 stoppingdown

stoppingdown

    Advanced Member

  • Members
  • PipPipPip
  • 530 posts
  • LocationGenoa / Milan, Italy

Posted 27 July 2017 - 09:24 AM

Can this discussion be supported by sample pix?

 

... which was my original desire, yes.


stoppingdown.net

 

Sony a6300, Sony a6000, Sony NEX-6, Sony E 10-18mm F4 OSS, Sony Zeiss Vario-Tessar T* E 16-70mm F4 ZA OSS, Sony FE 70-200mm F4 G OSS, Sigma 150-600mm ƒ/5-6.3 DG OS HSM Contemporary, Samyang 12mm ƒ/2, Sigma 30mm F2.8 DN | A, Meyer Gorlitz Trioplan 100mm ƒ/2.8, Samyang 8mm ƒ/3.5 fish-eye II | Zenit Helios 44-2 58mm ƒ/2 
Plus some legacy Nikkor lenses.

#11 JoJu

JoJu

    Advanced Member

  • Members
  • PipPipPip
  • 2,553 posts
  • LocationSwitzerland

Posted 27 July 2017 - 05:52 PM

Your contrast classification makes me think: Is he talking about the difference between max black and max white (meaning the range between straight black and white) or is he talking about the tonality in between the max marks? From a bad lens (contrast wise) with a rank of 3 I still can get contrasty pictures by raising the whites and lowering the blacks - but I loose a lot of tones in between.

 

Like Klaus said, lots information can't be recovered.

 

But with lenses of only 5% difference in contrast you need to compare the pictures directly and the comparison will be worthless, because the less contrasty lens could resolve 10% better and soon the better details would outrun the theoretically better contrast. Or a better colour fidelity. I just want to point out that contrast is only one characteristic of a lens, there are so much more and no picture (except test shots of test boards) can be repeated - so what's the point? A good lens should deliver sufficient contrast - if the tonality is maxed out, all the better and all the more versatile it will be.

 

But in dave'd samples if was not so much the contrast of the problem (the labels aside) but the contrast of the whole scene - and if the sensor doesn't cope with contrast, no lens can improve that.



#12 miro

miro

    Advanced Member

  • Members
  • PipPipPip
  • 435 posts
  • LocationI am here

Posted 28 July 2017 - 06:00 PM

I know that most members here have read this , but from what I read in this forum  I'm afarid that it is somehow forgotten

 

https://www.zeiss.co...cial_mtf_01.pdf

 

Is there any unclarity?

Do we need this discussion?



#13 stoppingdown

stoppingdown

    Advanced Member

  • Members
  • PipPipPip
  • 530 posts
  • LocationGenoa / Milan, Italy

Posted 29 July 2017 - 12:56 PM

Do we need this discussion?

 

Yes, because I don't know what you mean. I'm not able to read and understand all of that document, and I don't think all photographers are.

 

Secondarily, that is the theory. The question was related to practice - we all know that one thing is to test a lens on an optical bench and analysing diagrams, another thing is taking photos with it. The two things are not totally uncorrelated, anyway they are two different things.


stoppingdown.net

 

Sony a6300, Sony a6000, Sony NEX-6, Sony E 10-18mm F4 OSS, Sony Zeiss Vario-Tessar T* E 16-70mm F4 ZA OSS, Sony FE 70-200mm F4 G OSS, Sigma 150-600mm ƒ/5-6.3 DG OS HSM Contemporary, Samyang 12mm ƒ/2, Sigma 30mm F2.8 DN | A, Meyer Gorlitz Trioplan 100mm ƒ/2.8, Samyang 8mm ƒ/3.5 fish-eye II | Zenit Helios 44-2 58mm ƒ/2 
Plus some legacy Nikkor lenses.




1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users



© by photozone.de