Lens Tests

[Roger Hicks]

In the sense that I go out and use a lens then of course I test it, but what I don't do is 'test' things like edge sharpness, distortion . . . once you get around f5.6-f8, it's pretty much impossible to tell different lenses of similar focal length apart.

I shoot a fair number of buildings, inside and out, so distortion matters to me if there is a straight wall near the side of the picture; so I make a deliberate attempt, early on, to include such a shot, which otherwise might not come up for several rolls. Of course most manufacturers publish this data, so if I already know, there's no need to check.

I'd agree about the f/5.6 thing EXCEPT for older, cheaper reverse-telephoto wide-angles and zooms, where distortion is often horrible at all apertures and resolution may still be improving beyond f/8 or even f/11.

Contrast also varies quite widely, along with its cousin, flare.

Cheers,

Roger

 

[naren]

In the past the only "test" I've done was comparing the 2 versions of the Nikkor 35mm PC (perspective control) lens I have, using digital capture. Both were reputed to be of excellent quality, the later one was said to be the sharpest Nikkor 35mm lens at f/2.8. I saw no difference but I'm pretty sure that's because the digital sensor captured only the middle part of the image. I'm going to repeat the test on B&W film along with 3 other Nikkor 35mm lenses I have now.

I've been buying a lot of old equipment lately for shooting a documentary type project (and my God those prices!) So I've picked up a few duplicates, etc. I think you are trying to figure out what makes sense for you? And probably you are curious as well about hoopla on the boards. I am testing partly to satisfy my curiosity but mostly in order to do my best as a professional photographer. I think it is important not to waste your time and to consider in your testing - 1. What you are doing makes sense scientifically. 2. It makes sense in practical terms. For me if I know that I'm going to get a sharper image with my 35mm f1.4 and I know I am printing scenes with lots of detail at 11x14in for display, then after having done the tests I know it's worth toting the lens. If I am shooting flash, and I know that my 35mm f2.5 will give me just as good or close enough sharpness by whatever f-stop- I can take my much lighter lens. Heck I might even sell the f/1.4 based on these tests (I don't like the focusing action anyway). I've read over a number of times lens tests by Bjørn Rørslett (I like the info he gives and his style) and people who do that kind of testing are invaluable to pros who haven't had the chance or won't in order to find out themselves. Some photogs who aren't very technical will say, whatever, I'm using Nikkors or it's all good modern glass. I think the more you know in this Ultra-competitive field the better off you are. So you can consider also what the gentlemen says below but you don't have to test exactly as he says to learn what you want to know. Nothing wrong with printing 11x14's, use the same enlarger, etc.

Most amateur 'tests' are appallingly badly designed anyway -- colour neg film, commercial processing, scanner -- whereas if you are going to do it properly you need consistently processed film (and always the same film), a microscope for examining sharpness (and grain, in film tests -- and a photomicrographic outfit so you can compare pics rather than trying to remember which looks grainier), a shutter speed tester for leaf shutter lenses... As for comparing actual apertures with marked, this is REALLY difficult.

The following statement I don't agree with at all-

I generally try to avoid using lenses wide open, unless I'm looking for a specific effect, and my experience has been that once you get around f5.6-f8, it's pretty much impossible to tell different lenses of similar focal length apart.

Although I may avoid shooting wide open in many situations if you do read the tests people have done, some lenses are very sharp wide open. Most tests will describe some softening or minor loss of contrast if it's a good lens- but some also say "snappy" even wide open. It pays to know these things. As far as being able to tell different lenses apart at 5.6... I bet you can see a noticeable difference in quite a few lenses of a given focal length. I think with 35mm film the difference becomes noticeable at about an 8x10 print or 8X. I've seen shows with 16x20s from 35mm, and that guy knew a thing or too I'm sure about sharp lenses. I am very curious to see the diffence in my 11x14s between the Nikkor 24mm f/2.8 I just picked up (another great eBay score!) and the Zeiss ZF 25mm f/2.8 if I have get the dough for that lens. That will put a considerable difference in published lens test results into "real world" terms for me. I was holding out for the Zeiss lens because that was pretty much the last prime lens missing from my Nikkor line up and although it was a landmark lens, the 24mm Nikkor has not been reported to be stellar. The sample I got is gorgeous though. It has a the old hill and dale focusing collar I prefer and the Ai "kit" (if anyone can explain the details of those I would much appreciate) and a buttery focusing action.🙂

 

[Ade-oh]

As far as being able to tell different lenses apart at 5.6... I bet you can see a noticeable difference in quite a few lenses of a given focal length. I think with 35mm film the difference becomes noticeable at about an 8x10 print or 8X. I've seen shows with 16x20s from 35mm, and that guy knew a thing or too I'm sure about sharp lenses.

Fair enough: I'll take a closer look to see if I can 'spot the difference'. I suppose in reality I rarely print much larger than 10x8 in 35mm and, in any case, there's the issue of the enlarging lens which comes into play then which presumably may be adding some distortion all on its ownsome.

Contrast also varies quite widely, along with its cousin, flare.

True, though I think I would bundle these in as part of the 'character' of a lens and work around it in normal use.

 

[raid]

Yes, but there's a big difference between testing something to see if it does what you want, i.e. taking pictures with it (which you can hardly avoid doing); testing resolution, distortion, etc., with test charts; and obsessively comparing it with another lens you own.

Most amateur 'tests' are appallingly badly designed anyway -- colour neg film, commercial processing, scanner -- whereas if you are going to do it properly you need consistently processed film (and always the same film), a microscope for examining sharpness (and grain, in film tests -- and a photomicrographic outfit so you can compare pics rather than trying to remember which looks grainier), a shutter speed tester for leaf shutter lenses... As for comparing actual apertures with marked, this is REALLY difficult.

A thourough test like Raid's is one thing, but most people's 'tests' are so casual that they tell you next to nothing that you could not learn from taking pictures.

Then there is the informal testing that most of us, surely, do when we first get a lens: a wall or other straight line near the edge to check for distortion, twigs or telephone wires against the sky for flare, maybe even an informal vignetting test at full aperture (sky, white wall, whatever). With half a dozen shots or fewer we can see if there are any nasty surprises; we don't need much more.

Cheers,

Roger

Hello Roger,

I did my lens comparisons with a lot of caution, knowing the missing elements of a professionally designed and excuted lens test. I simply do not have the time and energy to do what you described above as being necessary steps in a valid lens test. I still believe that my comparisons of lenses gave some useful results for the potential user of the lenses that were included in my tests.

same film
same developing
same subject
same light
use of tripod
use of cable release
replicate photos


Roland's efforts [Thanks!] in taking my images and creating with enlargements a higher level of comparisons made the last battery of 35mm-40mm lens testing more useful. We covered:

sharpness
flare control
bokeh
contrast

In the end, it all was a lot of fun.

Thanks for the heads up. 🙂

 

[Dektol Dan]

The Authority is Whom?

The Authority is Whom?

Yes, but there's a big difference between testing something to see if it does what you want, i.e. taking pictures with it (which you can hardly avoid doing); testing resolution, distortion, etc., with test charts; and obsessively comparing it with another lens you own.

Most amateur 'tests' are appallingly badly designed anyway -- colour neg film, commercial processing, scanner -- whereas if you are going to do it properly you need consistently processed film (and always the same film), a microscope for examining sharpness (and grain, in film tests -- and a photomicrographic outfit so you can compare pics rather than trying to remember which looks grainier), a shutter speed tester for leaf shutter lenses... As for comparing actual apertures with marked, this is REALLY difficult.

A thourough test like Raid's is one thing, but most people's 'tests' are so casual that they tell you next to nothing that you could not learn from taking pictures.

Then there is the informal testing that most of us, surely, do when we first get a lens: a wall or other straight line near the edge to check for distortion, twigs or telephone wires against the sky for flare, maybe even an informal vignetting test at full aperture (sky, white wall, whatever). With half a dozen shots or fewer we can see if there are any nasty surprises; we don't need much more.

Cheers,

Roger

Shots at test patterns? Oh please! Thank God my tests are so amateurish! Good lenses have a 'look' a 'personality', it takes a professional to nurse a lens to its best and edit the results for others to to see without bias as presented in the photographs themselves. However, experience cannot be separated from opinion in these matters, not everyone can afford to play with the best, or the most rare and unique.

There is much more to a lens than lines per inch and believe me, 4000 DPI scans reveal far more than any loop. What is wrong with using negative scans for God's sake? I can't even follow that one! There isn't a color positive film made that can match the latest black and white negatives and besides when is god enough good enough? Color aberration and comas can be seen quite well from scans of negative film, thank you. When is grain apparent? Why at 2800 DPI. We are doing far better than that nowadays.

There is no one catch all lens test. I look at many images on Flickr, so far that is my best measure: pictures from the real world: real day to day use.

 

[Roger Hicks]

I did my lens comparisons with a lot of caution, knowing the missing elements of a professionally designed and excuted lens test...

Dear Raid,

Sorry if I created the wrong inpression. I meant specifically that your tests WERE well designed, unlike most -- which you have just confirmed.

Cheers,

Roger

 

[Roger Hicks]

Shots at test patterns? Oh please!

Consider what I do for a living, and what magazine editors (used to) want to publish. Note also that I said MICROSCOPE, not loupe.

Cheers,

R.

 

[raid]

Dear Raid,

Sorry if I created the wrong inpression. I meant specifically that your tests WERE well designed, unlike most -- which you have just confirmed.

Cheers,

Roger

Hello Roger,
I understood your initial comment, and that's why I thanked you for the heads up. I can see the need for additional testing methods when comparing new lenses. It seems that almost all new lenses are fine performers for general applications.

 

[Roger Hicks]

. It seems that almost all new lenses are fine performers for general applications.

Dear Raid,

No disagreement there (except zooms).

That's why I think that 'objective/subjective' well-controlled comparisons like yours are so valuable.

The only caveat is that in magazines we say, "Trust what we say, not the pictures, because of the limits of photomechanical reproduction."

On the web, I'd trust what you say (based on direct comparisons of original pics) well ahead of what I see (on-screen images).

On this occasion, "you" means "Raid", not "one" (as in "one learns that") or "other testers".

Glad there was no misunderstanding. I'm almost tempted to write, "Respect!"

Cheers,

R.

 

[raid]

Dear Raid,

No disagreement there (except zooms).

That's why I think that 'objective/subjective' well-controlled comparisons like yours are so valuable.

The only caveat is that in magazines we say, "Trust what we say, not the pictures, because of the limits of photomechanical reproduction."

On the web, I'd trust what you say (based on direct comparisons of original pics) well ahead of what I see (on-screen images).

On this occasion, "you" means "Raid", not "one" (as in "one learns that") or "other testers".

Glad there was no misunderstanding. I'm almost tempted to write, "Respect!"

Cheers,

R.

Hello Roger,
Thanks again for your words.
I value your experienced input here.

Cheers,
Raid

 

[Tom A]

I dont "test" a new lens. I usually take it out with my regular film and a camera that I know is correct as to focus/speeds etc. I shot a batch of film with it (usually 5 rolls) over a period, Sometimes in one day and sometimes in a couple of days. If it is a lens with a focallengths that is close to one that I already have - a second body with a "known"lens will provide a "reference" point. My shooting is as per usual - a rather aimless wandering around with a couple of "set shots", the trees down at the beach, the stainless steel sculpture at the museum and the laboriously engraved poem about rain in Vancouver on arock on one of the beach walks.
This will tell me how the lens will perform in my type of shooting and that is all I am really interested in.
Years ago I used to shoot industrial stuff. It was paramount that straight lines were rendered straight and that colors were true and I did all the McBeth colorcharts, the US Ar Force resolution tests and focus checks that I ever wanted to do.
The amazing thing with todays optics are that they are really very, very good and that doesn't mean that only the expensive ones are good, some of the low or medium cost lenses today are quite spectacular performance wise.
Ergonomics, build quality, size and weight are more deciding factors today. I have the luxury that I dont have to deal with an editor or a client anymore and what I do I do for my own pleasure.
As I was reading this I tried to recall what bad lenses i have used over the last decade or so. Surprisingly few!! I had "dud" 35/1,4 Asph Summilux, an old 50/1,4 Nikkor with a misaligned element and there might have been more, but none springs to mind (a couple of russian 50's maybe).
As for lower cost "gems", the original 25/4 Snap Shot Skopar, the 28f3.5 Color Skopar. Midrange ones. The 25/35/50 (both 1.5 and f2 of the 50's) ZM line and then of course the expensive ones, the 50f1.4 Asph Summilux, the 75/2 Summicron, the 18f4 Distagon ZM and the 35mm f1.2 Nokton.
I have said if before and I will say it again. As rangefinder users, we have never had it this good! Unfortunately, neither can we blame our equipment either when the shot doesn't measure up!

 

[raid]

I own only one CV lens that I bought because I was curious about "that new lens line" a few years ago; it is the old style 25mm/4 Snap Shot Skopar. It really is a gem, as Tom states above.

some examples:

 

[Hacker]

Do you not take a newly purchased car for a test drive? Do you not smell a bottle of cologne before buying it, or do you buy it, spray it on and hope for the best?

No, and yes, I still have lenses that I have not shot with yet 🙁. I do, however, once in a while, shoot at the closest distance and at infinity to satisfy myself just before using the lens on a shoot.

I guess I have no pressure to have the shots perfect as I'm only an occasional photographer.

 

[Toby]

I always run a test roll through the camera, just to check everything is working OK, but I'm a portrait photographer so edge sharpnes isn't so important, most of what I look for in a lens is subjective - do I like it rather than technical specs. I love the old 105 SMC takumar I have for my pentax 67, but I doubt it's a very good lens in tests.