"standards" for testing of manual focusing lenses?

jf56 · Joined: 02 Feb 2010 Posts: 10 Location: newport, oregon

Hello to all again. I am curious if there are any "standards" or testing procedures for determining proper lens functioning of older manual focusing lenses? In particular, I have read a statement that APO lenses in particular are pretty sensitive to manufacturers tolerances in order to focus "correctly"? And, that if lens groups or individual lenses within the lens barrel are out of synch or alignment in any way that they will give less than stellar performance.

I am not talking about front or back focusing issues. That can be tested and corrected for somewhat by using the Michael Tapes Lens Align Pro kit with he long enumerator, and then adjusting camera settings.

Also, is a tiny bit of "play' while focusing with the focusing ring 'normal', or should there be absolutely no play whatsoever?

Attila · Expire: 2025-11-18

Put your lens on matched camera body , wait for good lights , try to find nice subjects and shoot Smile

Pretty much this is standard, there is other methods too. Shoot in room bookshelf, teddy bear, neighbor roofs , but these are less popular. Premium test is make portraits from beautiful young ladies Laughing

Orio · Expire: 2012-12-04

Seriously, this is an interesting topic, that deserves a discussion.

Personally, I'm not really inclined to tests, for two reasons:

1- they require discipline, and when it comes to photography, I'm a rather indisciplined person, as my urge to create something new is always bigger then the will of following a tight schedule

2- they require free time, and I don't have much of it, so when I have a timeframe for photography, I usually choose to use it for something more creative than testing.

Having that said, I have my own idea of how a lens test should be, and it does not involve traditional things like bookshelf tests or chart tests, for the following reasons:

1- lenses are usually not made to photograph bookshelves. I mean it both aesthetically and technically. Technically speaking, most lenses are optimized for infinity use. So, unless someone has a living room that is 1 km long, I doubt that he can make a fair test by photographing a bookshelf Wink

2- photographing objects like resolution charts requires absolute precision and consistency (of distance, angle, leveling, illumination, focusing, flatness of surface, et c.). In most if not all cases of such tests that I have seen performed by non-professional people, one or both these elements were missing at least once during a test

3- things like copy variations, manual focusing accuracy, autofocus accuracy, focus shifts (especially focus shifts!) are rarely, if ever, took into due consideration by non professional testers.

So, either someone performs such tests with professional scientific accuracy and instruments, or, it's basically just wasted time.

So what kind of tests I do in the rare occasions when I feel so inclined?

Basically, I take the lens out and try it in different situations. I have developed a routine, or protocol if you prefer to call it that way, that I wish I was so determined to always follow, but alas I am not Laughing

but someone more constant than me might find it useful. This is my routine:

1) VIGNETTING TEST. I photograph a flat and evenly lit surface at all apertures. Often I don't have such surface at hand so I photograph the sky

2) SHARPNESS TEST. I photograph three subjects, one close (at or around minimum focal distance), one in the middle ground, and one at infinity. I do that with at least two apertures, one wide open and another one two stops down.

3) BOKEH TESTS. I make a series of photos of different objects, at different apertures (at least 3), and with different backgrounds, to compare the results. I try both backgrounds with highlights and evenly lit backgrounds. I also test foreground bokeh, which is something that most people forgets about, but can be important in some photographic situations. I do not only evaluate the specular highlights, I also try to evaluate how a lens "draws" the OOF objects at the different apertures. And I try to evaluate the "gradation" of bokeh, that is, the transition from in-focus to out-of-focus. This is best done by photographing round and deep enough objects (a vase, a human head, a column, etc.)

4) CHROMATIC ABERRATIONS TEST. I photograph subjects with high contrast. Typical subjects are metallic objects, naked trees over overcast skies, falling waterdrops lit by sunlight. I take them both focused and defocused, and at 3 apertures at least (or until the phenomenon disappears), from wide open and proceeding stopping down 1 at a time

5) FLARE TEST. I remove the lens hood if present, then photograph scenes with a strong back light, usually the sunlight if possible. I try three situations: sun in the frame freely, sun in the frame but partially obscured by objects (usually branhces), sun just out of the frame.

6) GEOMETRIC DISTORTION TEST. I photograph objects with a sufficient number of vertical and horizontal lines, for reference. They should be present both in the centre of the frame, and at all four edges. Care must be used in handling the camera so that it is as much as possible levelled (in order not to add camera inclination distortions)

7) 3D TEST. I find a scene that has at least three elements distributed one in the foreground, one in the middle ground, and one in the background. I then try different photos focusing on slightly different points, and using at least two apertures, one wide open, the other stopped down of 1 or 2 stops, depending on the base speed of the lens (slow lenses are stopped down of just one stop)

8 ) DIFFRACTION TEST. I almost never do this one because I am convinced that I would test the camera more than the lens. Anyway it should simply be like taking a picture with the lens completely stopped down. Normally that would mean a landscape, or a deep enough scene, and normally you would focus on a fairly close (but not closest) object.

9) COLOUR TEST. This is another of quasi useless tests, because in this case too you end up testing the camera instead of the lens. Anyway it can be useful to try the lens in at least the three basic situations, which according to my own experience are: 1) testing the reds (excruciating moment for all digital cameras) 2) testing the ability of differentiate different shades of green (photograph some vegetation where you can see at least three different shades of green) 3) testing the human skin (how natural the lens can render human skin). In my experience, only this third test (human skin) shows significant variations based on the lens used. The other two situations are much more influenced by how the camera behaves.

10) SHADOW DETAIL TEST. This is a more important test. I photograph scenes where there are shadow parts that contain details (typical examples: a shrub in the shade, a wall in the shade), talking a normal exposure (that is, not exposing for the shadows). With a normal exposure that includes also the sunlight parts, the shadow parts will end up underexposed. How well the details will surface from those shadowed areas is the purpose of verification. We could also call this one a micro-contast test, because we see more details in the shadows the higher micro-contrast a lens has.

11) FLAT DULL OBJECTS TEST. Or we could call this one micro-contrast test part two. Basically you choose a flat subject with dull diffused and undirectional lighting (typical example: a dirty window or dirty flat coloured wall when the sky is overcast, or at dusk, or when the object is in complete shade). Then you verify how well the minuscule details (a blot, a patch of dirt, etc) are visible and how well do they stand out. It is imoprtant that the objects have both a flat basic surface (monochromatic and dull), and that there are elements on it (like dirt or humidity or waterdrops) that do not contrast much with the underlying surface.

12) INFINITY FOCUS TEST. This is another basically useless test. I mean useless for the others - but useful for yourself. You take a lens, mount the adapter, then on the camera, then focus at infinity and check if you reach it or not. If your camera has liveview, one procedure could be to take one first shot at the infinity stop position, and a second shot focusing with liveview, to verify if the infinity point might be earlier than the stop point (it happens every now and then)

13) IMPACT TEST. You take your unprotected lens in the hand, raise it over your shoulder, then let it drop on a concrete surface and see what happens. No, you will not do that, especially as test #13, as it's sure to bring you bad luck. Rolling Eyes

14) EDGE/CORNER PERFORMANCE TEST. I select a scene where I have at least the four corners (but ideally all the edges) of the frame all filled by an object that is at the same distance from the camera, thus granting the most equal possible degree of focusing (typical subject: a wall that has enough surface detail to allow for a good evaluation). Then I take a series of photos starting from wide open at arriving at least to f/8. This should be done at infinity (although I understand it's hard to find a wall that big).

15) FOCUS SHIFT TEST. This is a crucial test and I was almost forgetting it. It pertains mostly to fast lenses (=faster than f/2.8 ), but can minimally happen also with slow lenses sometimes. Basically, you choose a flat subject at a diagonal angle, so that it has a part nearer and a part more distant. Typical subject: a square road sign, or hotel sign, or ad billboard, or vertical tombstone, et c.). You put the lens wide open and focus on a point (say, one letter of the tombstone writing). Take a picture. Then, without refocusing and without moving the camera, you stop down the lens of 2 stops, and take another picture. If you notice that the focused point of the second picture is different from the first, then your lens does shift the focus with the aperture.

As a final stage of the test, I take a series of pictures completely freely, following only my inspiration and visions of the moment. Although they are not specifically aimed at testing anything, they will often reveal useful things just the same.

I am sure I am forgetting something, in case I will add later.

martinsmith99 · Expire: 2013-11-18

This is a good question. Nodbody seems to like my tests so I don't post them. `Samples' is another thing, but they don't show comparisons.

LucisPictor · Expire: 2013-12-03

Good question, jf56!
And an excellent post, Orio!!
(I think this should be pinned to the top.)

I tried some objective testing but miserably failed which brought me to the conclusion that if you don't have a lab with an optical bench and a collimator etc., you cannot perform really objective tests.

Just use the lens to shoot whatever you prefer to shoot.

If I want to "test" lenses now, I do it in a similar way Orio describes. Checking for "flaws" in everyday situations. Sometimes these "flaws" add character to an image.

ZoneV · Expire: 2011-12-02

Regarding your question about the play in the focussing: I would say it is not normal. Some lenses may have play due to a complex construction (like zooms). But mostly I think play is a kind of wear.

I perform some simple lens testing at work - and because of this I don�t do that at home.
Long time ago I thought about a private lens testing setup for real MTF measurements - but fortunately I decided not to waste my time for that.
For my lenses I only make some pictures, and do some pixel peeping afterwards. Normaly I do not care about vignetting and edge sharpness.
I think tests with different objects / illuminations positions are not worth the effort. And tests that realy can be compared are realy high effort.

Probably I do that for some lenses, lets say 10 different 50 mm lenses. But this should be done on one time. I do not have the space for a test setup that is unaltert for some years.

I think the lenses I like most, would be pretty bad at the f-stop I use them most:
Minolta Rokkor 58/1.2 and Meyer Trioplan 100/2.8.
The Zeiss 50/1.4 would probably be sharper than the Rokkor, but I do not use it. The Zeiss 135/2.8 or Sigma 90/2.8 will probably have more contrast - but I do not use them for good pictures.

Bruce · Expire: 2014-11-22

Orio · Expire: 2012-12-04

poilu · Expire: 2019-08-29

cooltouch · Joined: 15 Jan 2009 Posts: 9097 Location: Houston, Texas

Regarding thorough testing, I'm wondering why reinvent the wheel? Why not instead do a bit of research and see if Modern Photography magazine ever performed a test on the lens? Or an equally well-respected European publication? Then the search is on for a copy of that issue. Libraries have -- or at least used to have -- back issues on microfiche, selections from which could be printed out.

I've always preferred Modern Photography's lens tests, not just because they published resolution and contrast results, but because they also had a state-of-the-art optical lab in which they were able to perform all sorts of other tests to evaluate lens aberrations. The results they achieved have always been good enough for me. I may not have always liked those results, but I've always trusted them.

cooltouch · Joined: 15 Jan 2009 Posts: 9097 Location: Houston, Texas

fergus · Joined: 21 Jan 2009 Posts: 61 Location: Surrey, UK

Regarding test #13, I would not want to personally conduct this test intenionally, although I think the manufacturers perhaps should and report how tough their lenses are Laughing

. I suspect the belief that the old heavy metal and glass lenses are more resilient to mishaps than modern day plastic lenses may be somewhat wishful thinking. I would much prefer to drop my ef-s 18-55mm 3.5-5.6 Canon kit lens than my heavier soft bodied aluminium lenses any day Laughing

Don't get me wrong, (I've got my materials engineer's hat on) Wink

I prefer the old metal to plastic, it feels better. But, don't kid yourself, in what would fair better in a drop test between a modern plastic lens and and old heavy metal lens.

Orio · Expire: 2012-12-04

cooltouch · Joined: 15 Jan 2009 Posts: 9097 Location: Houston, Texas

I have a Canon EF 75-300mm -- pretty plasticky, but it's an older one and not as plasticky as the latest version. I had it sitting on my desk, walked by the desk and brushed against the lens, and knocked it to the carpeted floor. A distance of about 30" (~75cm). Carpeted. No external damage is apparent, but now the focusing ring gets tight toward infinity, and the internal motor doesn't want to push against the resistance. I've accidentally knocked a couple of older mf lenses from this same desk, but they were not harmed.

So I can understand the argument that lenses made from more modern materials might be more resilient against impact damage. And perhaps they are. Mine didn't show any external damage, after all. It was the internal damage, probably due to the design/manufacture of materials and structures that can be made as cheaply and assembled as quickly as possible. That's where I pin the blame.

fergus · Joined: 21 Jan 2009 Posts: 61 Location: Surrey, UK

Yes, unfortunately it is purely down to a few unwanted personal experiences and just how the lens lands. To be honest, I would hope or expect your Carl Zeiss lens to be more resilient than a lesser lens and would certainly prefer it, though I would try very hard not to drop it Laughing

Yes, I learnt not to leave lenses on table tops when my cat is around, as of yet, I've fortunately only lost a few coffee cups, ornaments and the like. You have got to admire their curiosity or laugh at the occasional clumsiness.

The only lens I have dropped from a height is my old Rikenon 2/50 cheapo plastico fantastico and that is still good after nearly 30 years apart from the non-engraved lens markings rubbing off. Though I have had filter threads going out of round on metal lenses from quite gentle bumps.

At the end of the day it depends on what plastic or metal is used in construction and its design, I just get a little annoyed with the blind view that �metal is better than plastic� and that's from me preferring metal to plastic in a feely sort of way.

enzodm · Joined: 11 Sep 2010 Posts: 350 Location: Italy

Orio · Expire: 2012-12-04

jf56 · Joined: 02 Feb 2010 Posts: 10 Location: newport, oregon

Back again, I want to thank Orio and cooltouch (?) for their responses. It is nice to be taken seriously, although I too do have a sense of humor when the time calls for it. For Instance, if someone you love lies bleeding on the ground and you ask for help from a stranger,...you would not appreciate being handed a cup of coffee at that moment....

Anyways, I did get some good feedback. Thank you.