rayford2 Loc: New Bethlehem, PA

There used to be advice not to buy a car that was made on a Friday or Monday. On Friday the employees are in a hurry, and on Monday they're recovering from weekend activities.

Julian Loc: Sarasota, FL

Just being picky: lens's is the possessive form of lens; lenses is the correct plural. Great topic anyway.

rpavich Loc: West Virginia

After buying and returning counless lenses because they were "soft" I've come to the following conclusion about lenses:


There is an acceptable tolerance in lens manufacture; let's call it a "1-10" scale.

There is also an acceptable tolerance in Camera manufacture; let's also call it "1-10" scale.


When you have a lens like the 50mm and it looks sharp on one camera and not so sharp on another here is what's happening.


Your camera tolerances are biased towards the high side...let's give it a value of 8.


Your sons camera is biased towards the low side...we'll call it 2.



The 50mm lens is biased at a value of 3.


Now anything within say "3" increments is considered acceptable.

Your son's camera and the lens almost match perfectly...the lens focuses great because the manufacturing tolerances are "built up" in the correct matching direction.


Unfortunately your camera, while good, is toleranced in the opposite direction and so you see this same lens as "bad"...out of focus...soft.

If you were to get another lens that happened to be biased more towards the "10" side you'd say that it was great and your son wouldn't be so pleased.


I learned this because I finally got a camera that has adjustable AF...I can tune each lens to my camera.


I found that ALL of my lenses have to be tuned to the negative side of the "0" tolerance scale due to my camera's manufacturing tolerance shift.


Some, more than others..but they all have had it.


If every camera could be tuned to every lens I don't think we'd have this kind of conversation very much.



Does that make sense?

jsenear Loc: Hopkins, MN.

It does, and that probably is the answer to my question. Thanks. Stand corrected on the lens's - lenses point.

TJ28012 Loc: Belmont, NC

I would guess that the variations in sharpness that you noticed were the result of focusing and not the lens construction.

I have found that if I take numerous photos of the same subject, from a stable tripod without touching the lens or camera, sharpness will vary very slightly from photo to photo. My tests were done using both AF and manual focus with similar results (more variations with AF).

Testing lenses with digital cameras costs nothing but time and is a good project for rainy days. The results will often surprise you.

jsenear Loc: Hopkins, MN.

We had fun doing it. I have always suspected differences in the same lenses. I don't think this was a focus problem. The differences were too marked.

rpavich Loc: West Virginia

That's been my experience.

I just got a 135mm that was VERY very nice....but not "stellar" as they said....I kept front focusing...too much for it to be me.

I adjusted my AF and it needed 15 units of adjustment..now it's razor/cornea slicing sharp...right on the money....

jsenear Loc: Hopkins, MN.

What cameras can you adjust auto-focus? Just wondering.

rpavich Loc: West Virginia

I have a Canon 5D II. I don't know which others.

jnenvir

I understand that some noticeable variation is common in the Canon 70-200 f2.8 lens. I've experienced autofocus function variation between lens, and have been told by Canon rep both lens were within tolerance. My guess is sharpness tolerances are also measurable by the human eye.

glojo Loc: South Devon, England

Hi TJ,
Totally agree with your valid point about focusing and that is a great way of detecting if there is an issue with the glass. Having said that I would NEVER dream of conducting any type of test whilst the camera is shooting in jpeg as it just has far too many variables built into the compression algorithms. (Do you do your tests using the RAW format?)

We are discussing here images that have been compressed from 15mb to less than 3mb.. Some EIGHTY percent compression.

Just like everyone else I TOTALLY accept nothing we purchase will be 100% perfect but we are talking about glass. (lens)

Glass that might not be fitted correctly which would then consistently show an error in exactly the same place.

There could be mould on the glass which again will show up an error in exactly the same place. The glass might be loose which would then probably show an erratic blurry type image.

The lens might have faulty connectors which would obviously play havoc. They might be fitted out of alignment and would never focus.

The glass might be scratched and again this would show a consistent fault but surely issues with glass are what I would call black and white.. or totally consistent. If it is defective, it will be defective, it will have a problem and one that can be consistently demonstrated, demonstrated and then replaced.

They might even be performing within this so called 'variation' :)

If however we take an image in the jpeg format and then let the camera reduce the quality of the 'negative' by 80% then EVERY image will lack something different from the previous exposure. The camera reduces the file size by 80% but each decision will be subjective although the bottom line will be an image that has 80% less information to play with which might mean a loss of sharpness, a loss of quality, zip, punch, contrast... You pays your money and takes your choices as each image will be different in what is taken out. Is this the fault of the lens or is it the 'fault' of using jpeg or maybe there is an issue with the camera's processing capabilities?

I just do not agree with blaming the glass without seeing the issues, BUT we all MUST accept there are differences in those images we have not seen. :thumbup:

If we decide to shoot in the jpeg format then we must surely accept the camera will be removing millions of little coloured dots which will take away numerous qualities which in the grand scheme of things will definitely show up when comparing like with like! Two jpeg files shot with the same camera with the same lens will most definitely NEVER, NEVER, EVER have the same little dots removed. Surely this should be a factor before blaming the glass (lens)

Please note this is HUMOUR
Perhaps we should all buy Nikon equipment :evil: :twisted: :mrgreen: (humour)

jsenear Loc: Hopkins, MN.

Agreed. The tests would have been better if we had shot in Raw format. Our tests consisted of brand new glass. So that answers some of your questions. I'm leaning towards adjusting auto-focus in the camera as the answer. At any rate, I never pretended these test were scientific. They were not. The test served it's purpose. We got rid of one 18-55 vr lens and got a much better one. Also brand new.

glojo Loc: South Devon, England

EXCELLENT result

:thumbup: :thumbup:

Much respect
John
from sunny Torquay

Roger Hicks Loc: Aquitaine

Lens age, build quality and history are major concerns in sample variation. Where a top-flight lens might survive a six foot drop onto granite with only a bent lens-hood (my 35/1.4 Summilux did), a cheap kit zoom might be knocked out of alignment by a three-foot drop in its original box, somewhere between the assembly line and the retailer. These are extreme but not unrealistic examples. Old (including 'repaired') lenses can also exhibit problems.

Come to that, some camera manufacturers who buy in lenses will order a dozen at a time; test them; and return the ones that are not of the highest standard. Linhof are famous for this, but Gandolfi and Corfield do/did it too.

Cheers,

R.

rpavich Loc: West Virginia

I disagree with the idea that the "back focus vs front focus" test would benefit by being done in raw format.

In fact; my software "Focal" uses JPG format to do the testing...not raw.