dtcls100
Well-known
Recent testing indicates to me that the inherent optical differences in resolution between the newest lenses and older vintage lenses (1970's and 1980's) is surprising very low under near ideal testing circumstances. What has a much larger effect on image sharpness in the real world is how a manual lens is balanced and how well it snaps into focus, given that even tiny focusing errors and camera shake can tremendously degrade image sharpness.
At the outset, I am not a particularly big believer in most lens tests I have seen of lenses' optical sharpness, given sample variation, manufacturing tolerances, focusing errors, poor testing methodology, testing equipment problems, etc.
That said, I recently splurged on my first mirrorless camera, a Sony A7R3, which I have been using with adapted OM and Nikkor lenses. In setting it up and figuring out how it works, I noticed that it would be a very good tool for testing lenses, given the wide range of mount adapters available, as well as its precise leveling function, focus magnification, electronic shutter, etc.
I have conducted some quick tests, using a sturdy tripod, electronic timer, electronic shutter (essentially eliminating shutter shock), focus magnification (over 100X), and carefully leveling the camera using the leveling function. I also used top notch, new Novaflex OM and Nikkor adapters, which mount like butter and have zero play, to eliminate/minimize the slop, misalignment, poor manufacturing tolerances present in many cheaper adapters.
I tested a 1979 Vivitar Series 1 70-210 f3.5 zoom (version 1) wide open at its weakest focal length of 210mm, in taking pics of a sheet of music. I have never previously found images from this lens to be particularly sharp on film cameras. Based on my quick testing, it appears that any lack of sharpness is essentially due to focusing errors and camera shake. Using the magnification function, I saw that virtually imperceptible differences in the focusing had MAJOR effects on the image sharpness, as well as chromatic aberration, and that the image bounced around quite a bit while focusing, even when the camera was mounted on a tripod (the lens lacks a tripod mount). After carefully getting essentially perfect focus, I noticed that the purple fringing disappeared. Using a 10 second electronic timer, I took some images.
The images were shockingly sharp for such an old lens and were very close in sharpness to images of the same subject taken with a new Sony 24-105G lens (even with this new lens being shot at 105 vs 210mm). The main difference was that the newer lens had a little more contrast (not surprising given that the 1979 lens has a decent amount of internal dust).
Some quick shots of the same subject taken with a couple of my other older manual focus OM lenses (35-70, 35-80 Zuikos) led to similar results. The images were all outstandingly sharp.
This confirms my longheld suspicion that the inherent optical sharpness of many older lenses easily meets very high standards, even with the latest 42 megapixel sensors. The reason why images on these older lenses may be lacking in sharpness is pretty much attributable to user error -- focusing errors and camera shake. Even tiny focusing errors lead to substantially more color fringing and camera shake destroys sharpness.
Of course, some lenses are much easier to focus and handle than others, even if they have similar focal lengths and apertures. For example, the Series 1 zoom I tested is quite long and heavy. More importantly, it is front heavy and doesn't snap into focus as well as some other zoom lenses I have. I always had trouble holding it steady (even more so with the additional length of the adapter and even using the IBIS in the A7R3).
Carefully focusing the Series 1 zoom on a tripod (without the magnification function) showed how much one can miss focus. What looked like perfect focus at ordinary magnification was clearly well off when magnified 100x (with markedly lower resolution and much more purple fringing), although correcting the error typically involved only a tiny, almost imperceptible, turn of the focusing ring. This shows the limitation of lens testing using an optical finder, even ignoring mirror alignment in SLRs and dSLRs.
The central point of what I learned here is that the importance of differences in lens resolution between different lenses is vastly overrated by lens test reviews, given the inevitable, although tiny, focusing errors from using optical finders. Even many older lenses are surprising sharp under more ideal testing situations using a magnification function. What lens reviews really need to focus more on (pun intended) is on the balance of lenses, how easy they snap into focus, and their handling characteristics, as this appears to have much more of a real life effect on image sharpness than any real optical differences.
At the outset, I am not a particularly big believer in most lens tests I have seen of lenses' optical sharpness, given sample variation, manufacturing tolerances, focusing errors, poor testing methodology, testing equipment problems, etc.
That said, I recently splurged on my first mirrorless camera, a Sony A7R3, which I have been using with adapted OM and Nikkor lenses. In setting it up and figuring out how it works, I noticed that it would be a very good tool for testing lenses, given the wide range of mount adapters available, as well as its precise leveling function, focus magnification, electronic shutter, etc.
I have conducted some quick tests, using a sturdy tripod, electronic timer, electronic shutter (essentially eliminating shutter shock), focus magnification (over 100X), and carefully leveling the camera using the leveling function. I also used top notch, new Novaflex OM and Nikkor adapters, which mount like butter and have zero play, to eliminate/minimize the slop, misalignment, poor manufacturing tolerances present in many cheaper adapters.
I tested a 1979 Vivitar Series 1 70-210 f3.5 zoom (version 1) wide open at its weakest focal length of 210mm, in taking pics of a sheet of music. I have never previously found images from this lens to be particularly sharp on film cameras. Based on my quick testing, it appears that any lack of sharpness is essentially due to focusing errors and camera shake. Using the magnification function, I saw that virtually imperceptible differences in the focusing had MAJOR effects on the image sharpness, as well as chromatic aberration, and that the image bounced around quite a bit while focusing, even when the camera was mounted on a tripod (the lens lacks a tripod mount). After carefully getting essentially perfect focus, I noticed that the purple fringing disappeared. Using a 10 second electronic timer, I took some images.
The images were shockingly sharp for such an old lens and were very close in sharpness to images of the same subject taken with a new Sony 24-105G lens (even with this new lens being shot at 105 vs 210mm). The main difference was that the newer lens had a little more contrast (not surprising given that the 1979 lens has a decent amount of internal dust).
Some quick shots of the same subject taken with a couple of my other older manual focus OM lenses (35-70, 35-80 Zuikos) led to similar results. The images were all outstandingly sharp.
This confirms my longheld suspicion that the inherent optical sharpness of many older lenses easily meets very high standards, even with the latest 42 megapixel sensors. The reason why images on these older lenses may be lacking in sharpness is pretty much attributable to user error -- focusing errors and camera shake. Even tiny focusing errors lead to substantially more color fringing and camera shake destroys sharpness.
Of course, some lenses are much easier to focus and handle than others, even if they have similar focal lengths and apertures. For example, the Series 1 zoom I tested is quite long and heavy. More importantly, it is front heavy and doesn't snap into focus as well as some other zoom lenses I have. I always had trouble holding it steady (even more so with the additional length of the adapter and even using the IBIS in the A7R3).
Carefully focusing the Series 1 zoom on a tripod (without the magnification function) showed how much one can miss focus. What looked like perfect focus at ordinary magnification was clearly well off when magnified 100x (with markedly lower resolution and much more purple fringing), although correcting the error typically involved only a tiny, almost imperceptible, turn of the focusing ring. This shows the limitation of lens testing using an optical finder, even ignoring mirror alignment in SLRs and dSLRs.
The central point of what I learned here is that the importance of differences in lens resolution between different lenses is vastly overrated by lens test reviews, given the inevitable, although tiny, focusing errors from using optical finders. Even many older lenses are surprising sharp under more ideal testing situations using a magnification function. What lens reviews really need to focus more on (pun intended) is on the balance of lenses, how easy they snap into focus, and their handling characteristics, as this appears to have much more of a real life effect on image sharpness than any real optical differences.
Ko.Fe.
Lenses 35/21 Gears 46/20
Where are plenty of old lenses which are as sharp as modern ones. It is not something unusual. The real problem I see with many old lenses is how dull they are on colors if in use on digital cameras. Sharp resolution and dull on colors, this is what many, if not most of old lenses on digital cameras are for me.
Larry Cloetta
Veteran
Lloyd Chambers could not be reached for comment, but is still accepting donations to his Paypal account to duplicate your efforts.
Lux Optima
Established
Recently I presented a picture taken with a slow Summaron 3.5/35mm, at night, without tripod, on a simple Kodak Gold 200 at a contest in another forum.
The photo was critzised as "too sharp" - the lens is 62 years old...
The photo was critzised as "too sharp" - the lens is 62 years old...
Yeah, this stuff doesn't really matter much to me... most lenses are good enough IQ wise for the photography I do.
Calzone
Gear Whore #1
A lot of the old glass I own is really sharp, but to me the modern glass is more perfect to the edges and in the corners wide open. Really only an extension of of sharpness across the frame.
Cal
Cal
Ko.Fe.
Lenses 35/21 Gears 46/20
Cal, it is too technical. Most who are after old glass on digital are after bokeh - something sharp in the middle, the rest is soft 
.
.
peterm1
Veteran
The secret here, the secret I have to keep reminding myself is this simple fact. - To remain solvent, camera and lens manufacturers have to continue to convince us to buy new kit. Day in, day out. Month in month out. Year in year out. Once those sales stop the money stops, research and development stops and the production lines stop. The merry go round is stilled.
So they keep making pretty marginal improvements to their existing lines and turn out a few new ones. But when you think of it most of the improvements have been to things like lens coating to reduce flare (which also allows more complex lens designs), materials and production technology (which allows more comparatively cheap and light lenses), aspherics (ditto), lens flexibility (which allows some fantastic new zooms - also associated with the above factors) and of course computer designing which allows new lenses to be computed faster so new products can go out the door quicker. More recently we have also seen a flood of super fast lenses which also draw on all of the above and super wide lenses. None of these necessarily affect sharpness of lenses too much if at all for average lenses (by average I mean normal lenses such as 50s) But they have allowed new ranges of quite high quality lenses like ultra wides and ultra fasts. Over all, the little, marginal improvements have added up to big changes over time but I think not necessarily to sharpness.
I must confess, I am with others who value lens character over technical sharpness. Having said this the first thing I look for in a lens is acceptable sharpness, but not super sharpness above all else. Hence I have never felt compelled to buy and own many lenses that have a reputation for super sharpness across the field above all else unless they also have other characteristics I value.
So they keep making pretty marginal improvements to their existing lines and turn out a few new ones. But when you think of it most of the improvements have been to things like lens coating to reduce flare (which also allows more complex lens designs), materials and production technology (which allows more comparatively cheap and light lenses), aspherics (ditto), lens flexibility (which allows some fantastic new zooms - also associated with the above factors) and of course computer designing which allows new lenses to be computed faster so new products can go out the door quicker. More recently we have also seen a flood of super fast lenses which also draw on all of the above and super wide lenses. None of these necessarily affect sharpness of lenses too much if at all for average lenses (by average I mean normal lenses such as 50s) But they have allowed new ranges of quite high quality lenses like ultra wides and ultra fasts. Over all, the little, marginal improvements have added up to big changes over time but I think not necessarily to sharpness.
I must confess, I am with others who value lens character over technical sharpness. Having said this the first thing I look for in a lens is acceptable sharpness, but not super sharpness above all else. Hence I have never felt compelled to buy and own many lenses that have a reputation for super sharpness across the field above all else unless they also have other characteristics I value.
View Range
Well-known
I have the old two piece Nikkor 400mm f4.5, 600mm f5.6 and 800mm f8.0. I also have a Nikkor 400mm f2.8 AFS II. The old lenses are very poor in comparison to the newer lens. I don't think the OP's conclusions are valid for telephoto lenses.
Robert Lai
Well-known
If anyone wants ultimate lens sharpness, stop down 2 stops, use fine grain film, and use a tripod. That's a formula from the oldest photography texts that I have, from decades ago. This "secret" hasn't changed yet.
I was recently testing a newly overhauled Contax IIa (Henry Scherer did the job). Testing out infinity focus and 1/1250 shutter speed at the same time. I focused on a cell phone tower about a mile away, used the 50mm Carl Zeiss (West German) f/1.5 Sonnar close to wide open. I was pretty amazed at how much detail came out in the cell phone tower, when pixel peeping a low resolution scan from Dwayne's.
Even lenses from the 1950s had plenty of punch to be competitive with today's lenses.
There is a series of medium format lens tests which showed that the Xenotar and Planars from Rolleiflex 2.8Fs are comparable in resolution to the latest and greatest lenses from the Mamiya 7.
I was recently testing a newly overhauled Contax IIa (Henry Scherer did the job). Testing out infinity focus and 1/1250 shutter speed at the same time. I focused on a cell phone tower about a mile away, used the 50mm Carl Zeiss (West German) f/1.5 Sonnar close to wide open. I was pretty amazed at how much detail came out in the cell phone tower, when pixel peeping a low resolution scan from Dwayne's.
Even lenses from the 1950s had plenty of punch to be competitive with today's lenses.
There is a series of medium format lens tests which showed that the Xenotar and Planars from Rolleiflex 2.8Fs are comparable in resolution to the latest and greatest lenses from the Mamiya 7.
JP Owens
Well-known
With old SLR cameras, and DSLR cameras for that matter before they had micro focusing per lens capability, focusing errors resulting from the assembly line nature of these cameras created enough body to body variations, and lens to lens variations, that it was pretty much just luck if you got super sharp images. We often blamed this variation lack of sharpness in the lenses.
Back in the day, I would send any new camera to CPS to have it matched to my lenses. Of course, that's now a non-issue with mirrorless cameras...and, with DSLR's that allow micro adjustments.
I too am amazed at how sharp these old lenses are.
Back in the day, I would send any new camera to CPS to have it matched to my lenses. Of course, that's now a non-issue with mirrorless cameras...and, with DSLR's that allow micro adjustments.
I too am amazed at how sharp these old lenses are.
dtcls100
Well-known
How did you test these lenses? The long telephotos with slow apertures that you cite are the very kind of lenses most prone to manual focusing errors and camera shake that will severely compromise their performance, compared to a more modern AF lens that are image stabilized. So your examples may actually support my point, rather than undercutting it.
Dante_Stella
Rex canum cattorumque
I can concur with the original poster - quite a few purportedly "worthless" zooms and primes for SLRs do very well on mirrorless cameras like the Leica M. In fact, even things like the ZM Sonnar do too. I would attribute this to several things:
Dante
- SLR lenses can only really be focused wide open. Their screens become unusable quite quickly, and many SLR lenses have focus shift that manifests at the wider but not widest apertures. Same with a rangefinder: it does not in any way compensate for aperture, which is why we are seeing so many apo lenses now. You're not actually upping the resolution as making the focus more predictable.
- Focusing on, say, a Leica EVF can be done at 5-10x magnification, which is massively more accurate than most SLRs, all of which reduce the size of the viewfinder image. This same magnification wipes out focus shift in older RF lenses too.
- You can use the EVF to calibrate your manual RF focusing technique.
- It is very easy to have a misaligned SLR mirror, which can throw things off (since the distance to film has to be identical to the distance to the focusing screen.
- It seems that offset microlenses are capable of cleaning up some of the CA mess that you would get with SLR lenses, even if the microlenses weren't designed to correct for those specific lenses. What's left can be addressed in Lightroom pretty easily.
Dante
I would say this is generally true, lenses from decades back are actually quite good, and very close in performance to equivalent modern lenses. This is also from experience using such lenses on mirrorless, although I find lens testing to be tedious work so I never pursued things as far as the OP.
Modern lens design has progressed due to computer aided design, improved manufacturing capabilities, usage of lighter materials (better for autofocusing), among other things, but to get a significant incremental improvement over yesterday's lenses is very expensive. Many of today's designs are cost reductions of yesterday's.
Modern lens design has progressed due to computer aided design, improved manufacturing capabilities, usage of lighter materials (better for autofocusing), among other things, but to get a significant incremental improvement over yesterday's lenses is very expensive. Many of today's designs are cost reductions of yesterday's.
mdarnton
Well-known
When I started scanning film, the first thing I noticed was how many shots I'd made over the years where the focus wasn't quite perfect, but it didn't matter at all on normal-sized prints. We have the capacity now to pixel-peep and have convinced ourselves that since we have that tool, that it is a valuable tool, but looking at those old shots, I'm not convinced that those lenses and what I did with them was all that bad, just because I couldn't focus at 10X and then pixel peep on a computer.
teddy
Jose Morales
Great post. I have a Sony a7. It has more than enough resolution for anything. I use it up to ISO 800, like in film days. It is plenty sharp with my Leica R, OM Zuiko and Zeiss QBM lenses. I also have a Fuji XE-1 - what more could anyone want really. Plenty of resolution there also. I use my film Leicas, Olympus OM lenses with film and lots of other cameras. The point? Digital has more than enough resolution for anything. We've come to a point where digital has matured and the old lenses where excellent anyway. Sharpness is not photography - whats in the frame is. Specially film - where the process is a chemical one.
Koala Up a Tree
Sony a7, OM Zuiko 50/1.4 MC (late serial), VSCO Provia 100F filter, heavily cropped (low-res)
Koala Up a Tree
Sony a7, OM Zuiko 50/1.4 MC (late serial), VSCO Provia 100F filter, heavily cropped (low-res)
Bille
Well-known
Any digital image can be oversharpened in post. Does not say a lot about the lens involved.
Bille
Well-known
This is probably true for anything pre 1970. You should not see a lot of difference with newer, multicoated glass though.
In this age, low contrast and colors are basically a few clicks in Photoshop, unlike blown contrast and blown colors.
pgk
Well-known
I run a gallery and one of the best selling prints - the images area of which is approximately 26" x 17" (M9 - 18MPixels at 200ppi) was taken on a 135mm Leica lens designed over 50 years ago. The image is technically as good as it could be. Today's lenses are of advanced specification (ultra-wide zooms, fast aperture zooms and so on) which simply didn't exist in the past, but optically many older lenses are still excellent performers and will hold their own even today. Personally I'm much more interested in carriability and ergonomics in use than the nuances of test results which I'm always rather wary of without knowing just how good the tester actually is.
Pelle-48
Amateur photographer
More about lens sharpness but for LF and MF
I found this lens test for large format:
http://www.hevanet.com/cperez/testing.html
Here is medium format:
http://www.hevanet.com/cperez/MF_testing.html
Really a large work in testing but I am sceptical about the result. Not that is it wrong in itself but if you run a photo test in scale 1: 20 the focus must be perfect and film plane and test chart must be perfectly parallel. Look at the Perkeo fold camera! And look at the edge results for many LF- not very good. Compare also Mamiya 7 with all LF! 120 compared to around 60 in many cases....Even if 4"x5" has a much larger area than 6x7 it will not help for the result. How should an LF camera be tested? Most likely people will use them for landscape?
I found this lens test for large format:
http://www.hevanet.com/cperez/testing.html
Here is medium format:
http://www.hevanet.com/cperez/MF_testing.html
Really a large work in testing but I am sceptical about the result. Not that is it wrong in itself but if you run a photo test in scale 1: 20 the focus must be perfect and film plane and test chart must be perfectly parallel. Look at the Perkeo fold camera! And look at the edge results for many LF- not very good. Compare also Mamiya 7 with all LF! 120 compared to around 60 in many cases....Even if 4"x5" has a much larger area than 6x7 it will not help for the result. How should an LF camera be tested? Most likely people will use them for landscape?