A Comparison of Four 90mm Lenses

[jaapv]

As someone else mentioned, it is odd how the Summicron images are so light --

The explanation is simple- Different lenses have different light transmission factors, due to different coatings and glass. The aperture is a mathematical value and does not take this into account, the stops on an aperture ring are calculated by the maker, not measured. If you measure exposure with an external meter you can run into variations in light transmission of over half a stop - as happened here.
A very interesting and informative test btw. Thank you 🙂

 

[RObert Budding]

The explanation is simple- Different lenses have different light transmission factors, due to different coatings and glass. The aperture is a mathematical value and does not take this into account, the stops on an aperture ring are calculated by the maker, not measured. If you measure exposure with an external meter you can run into variations in light transmission of over half a stop - as happened here.
A very interesting and informative test btw. Thank you 🙂

Sorry, but the shadow detail differs in the photos. That is 100% due to differences in exposure.

It really is hard to draw conclusions unless the exposures are equal. And I'd be willing to bet that the exposures with all lenses shot at the smae f-stop would be much closer than I'm seeing with the current methology. If the light isn't changing, then just meter once and shoot.

 

[raid]

I avoid making comparison photos when the light seems to be variable. It is clear that some people have strong feelings about certain lenses, and I don't think that the posted photos clearly distinguish a particular lens as being the "winner". I need to learn how to post several images side by side, as Ray did here. It makes comparisons more practical to follow.

 

[jaapv]

Sorry, but the shadow detail differs in the photos. That is 100% due to differences in exposure.

It really is hard to draw conclusions unless the exposures are equal. And I'd be willing to bet that the exposures with all lenses shot at the smae f-stop would be much closer than I'm seeing with the current methology. If the light isn't changing, then just meter once and shoot.

Of course the exposure is different when the light transmission factor is different- the effective f-stops are different. That is exactly the point.

 

[KoNickon]

While different lenses may have different light transmission factors, I have a hard time believing that's what accounts for the rather pronounced differences here. The coatings and glass can't be that different among these four lenses. I think indeed it's different exposures (whether caused by differing light or not).

I'm glad Ray did this test all the same. But Raid has a point -- a blind test might be more instructive, since I think folks may have been influenced by knowing which lens's image they were seeing.

 

[jaapv]

Differences of over 1/2 a stop have been recorded between different brands- notably Leica lenses can be considerable " clearer" than other makes. But of course there can be a host of different reasons for variations in exposure.

 

[Sparrow]

While different lenses may have different light transmission factors, I have a hard time believing that's what accounts for the rather pronounced differences here. The coatings and glass can't be that different among these four lenses. I think indeed it's different exposures (whether caused by differing light or not).

I'm glad Ray did this test all the same. But Raid has a point -- a blind test might be more instructive, since I think folks may have been influenced by knowing which lens's image they were seeing.

How would a blind test work?? I have to know the manufacture before I can have an opinion!! 🙂

Joking aside Nick I’m sure your right that would be the way to go

 

[telenous]

A proper blind test requires that you pick the right lens/photo combo blindfolded 😀

 

[raid]

I have just posted some results taken with a Canon 28mm/3.5 lens (for sure) and also some results that "could have been taken" with a Minolta Rokkor 28mm/2.8 or a Fuji Natura 24mm/1.9. Maybe having people inspect the photos will tell me which lens I used! Now that's a really BLIND test. Not even the user knows!

 

[RObert Budding]

Of course the exposure is different when the light transmission factor is different- the effective f-stops are different. That is exactly the point.

I'm suggesting that all shots should be made at the exact same exposures. For example, shoot at f/4.0 and 1/250 for all lenses. I'll bet that the light that hits the film is within 1/3 stop of each other. Closer, anyway, than the current test of varying the exposure for each lens.

 

[raid]

There will be another test soon for tele lenses in the range 85mm~105mm.
Maybe I will be able to find out new things then.

Raid

 

[ghost]

something's wrong with the exposure. some are too bright, others too dark.

the rokkor has the best bokeh, imo.

 

[Rich Silfver]

It was a great test effort and I enjoyed seeing the results (I prefered the results from the Leica and CV (did not care much for the Rokkor and Hexanon).

Having said that - the change in exposure does matter so even though this may not have been a 'test' where one can draw too many conclusions it was still great fun.


(To me, as boring as it sounds, a lens test is in-doors with controlled and consistent light where no parts of the light source (or any heavily reflected surface) is in frame).

 

[raid]

It was a great test effort and I enjoyed seeing the results (I prefered the results from the Leica and CV (did not care much for the Rokkor and Hexanon).

Having said that - the change in exposure does matter so even though this may not have been a 'test' where one can draw too many conclusions it was still great fun.


(To me, as boring as it sounds, a lens test is in-doors with controlled and consistent light where no parts of the light source (or any heavily reflected surface) is in frame).

Rich,
I recall your comment on my flare test in which some frames had a larger portion of window shown than others. On a clear summer day, the light is consistent here.

Raid

 

[raid]

I for one want to thank you Ray. Sure there is a difference in the exposure, but the images tell a lot about the lenses. It is a lot of fun to examine and compare the different qualities. Ignore the people that are negative. I did a test recently comparing a summicron and a sonnar but I haven't posted it here because of the negativity of some of the people concerning your post.

Thank you for doing it. Just want to let you know that some of us really appreciate your effort.

I support Patrick's statement above. Anyone taking the time to do such tests should be supported. If anthying can be improved upon, it should be mentioned in a constructive supportive manner. Thanks, Ray.

Raid

 

[ampguy]

Hi Ray, very interesting tests, and your daughter has a beautiful smile. From what I can tell of web photos, the Leica has that Leica "glow" which to me is low contrast, the CV sort of has it, so the CV is kind of close like an FSU lens, in fact I wonder how a J9 would have fared?

The Rokker is good, it's just too slow. I think both the Rokkor and Hexanon, with their higher contrast (or whatever 😉) are further from the "Leica glow" look but better overall lenses.

I wonder where the Elmarit 2.8 would fit in here?

 

[Trius]

A very interesting test, and I add my thanks. Yes, such a test could more "controlled", but Ray isn't a professional reviewer/lens tester so that's that!

I have seen Hexanon shots that I really, really like, but there are so many variables when shooting b&w negative material ... film, EI, developer, processing technique, scanning skills, etc., that my conclusion is ....

Given a lens that is in a top tier (or at least isn't a dog in one or more areas), a hybrid workflow in b&w can be developed to get very good to stunning results. So one person's shots with the Hexanon are optimized because of his vision, preferences, methods, testing, etc. Give him a Summicron and have him follow the exact same workflow as the Hex, and you may get very different and "inferior" results. But let him loose on re-jigging the materials/methods/workflow, and you may improve things drastically.

BTW, on this test I think the conditions favour the Summicron and CV in some areas, but the Hexanon and Rokkor in others; it's personal preference that would cause me to say I liked one over the others. But I ain't saying anything about preferences!

 

[back alley]

so much of this is truly personal.

some like that 'glow' but to me it looks like the contrast is off and that it lacks detail in the more shadowy areas.
the hex and rokkor, to me, have more detail and less glow but have a 'drama' that the other 2 lack.

joe

 

[boilerdoc2]

Agree. Hard to tell with these small images but you can see more detail in the shadows with the 'cron, methinks. Is that the famous "micro contrast" at work?
steve

 

[back alley]

i was thinking that the rokkor and hex had better detail in the face which i would prefer as this is a portrait and i would think that most viewers would look there first & formost.