p.4 #2 · FM Lens test for Centering (Mount tilt / swing)
I tested my new 16-35/4 (freshly minted from the factory, manufactured in Feb) and its fantastic. I tested at 16 24 and 35 and all is good.
This is quite different from my earlier copy which was never biting sharp anywhere in the frame. It was just sharp enough. I sold it thinking this is what you get with an ultra wide zoom. I was wrong.
p.4 #3 · FM Lens test for Centering (Mount tilt / swing)
From where did you buy it - may I ask?
nandadevieast wrote:
I tested my new 16-35/4 (freshly minted from the factory, manufactured in Feb) and its fantastic. I tested at 16 24 and 35 and all is good.
This is quite different from my earlier copy which was never biting sharp anywhere in the frame. It was just sharp enough. I sold it thinking this is what you get with an ultra wide zoom. I was wrong.
p.4 #4 · FM Lens test for Centering (Mount tilt / swing)
Dear Fred, dear everybody, find above the results for my new Sony 24-105G OSS.
Taken at ISO100, F4, Silent Shutter, 1/125sec and 24mm.
Lens is performing well, i just realized that many photos are softer in the left side than in the right side, so I did your test.. Do you think this is still acceptable or is this a case where i should send it in? I'm somehow torn, tend to get hung up in the details
p.4 #5 · FM Lens test for Centering (Mount tilt / swing)
I made some better light testing this morning and had the following findings:
- OSS on or off doesn't really matter for my results
- The problem is not only at F4, but also at F7.1 still
- i would say the left 25% of the image are affected
- the siemens star dot centering test as suggested by Roger Cicala does not show any asymmetries
Below you see a photo taken at 24mm F5.6, please see the trees on the left side compared to those on the right side of the image.. the picture really detiorates in the left.. and that is not even wide open. (Image has been compressed to get below the filehost's 10mb limit)
p.4 #6 · FM Lens test for Centering (Mount tilt / swing)
I've having difficult understanding what this means: "... but you will be actually taking a picture of the BOTTOM RIGHT and TOP RIGHT areas (Red Circle) since the camera is now upside down."
I guess it seems to me that the "picture" will be whatever is in the frame, which is largely the same throughout and includes the red circle and the subject (camera orientation nonwithstanding) ? I get that the relative positions of these will shift with respect to the camera's frame, the subject moving across each corner in succession. Is that all it means?
Or, to phrase it another way: is the technique to just put the subject in each corner successively by rotating the camera 4 ways?
EDIT: Perhaps someone could be kind enough to post examples of each of the full 4 images taken, in the raw orientation that they were taken (i.e., no cropping, zooming, or rotating) to help out my tired old mind? The images can be downsized greatly, of course, as they are not so much to assess the lens quality as the test itself.
p.4 #7 · FM Lens test for Centering (Mount tilt / swing)
jeetsukumaran wrote:
I've having difficult understanding what this means: "... but you will be actually taking a picture of the BOTTOM RIGHT and TOP RIGHT areas (Red Circle) since the camera is now upside down."
I guess it seems to me that the "picture" will be whatever is in the frame, which is largely the same throughout and includes the red circle and the subject (camera orientation notwithstanding) ? I get that the relative positions of these will shift with respect to the camera's frame, the subject moving across each corner in succession. Is that all it means?
Or, to phrase it another way: is the technique to just put the subject in each corner successively by rotating the camera 4 ways?...Show more →
jeetsukumaran
My understanding is that you're putting the target into the four corners of the lens, plus one in the center.
We are testing the lens sharpness in the corners as compared to the center.
If you would like references to other methods of testing lenses; other than sending them to a qualified lab; there are some good references out there. However, this is one of the easiest ways to test, that I've found.
p.4 #8 · FM Lens test for Centering (Mount tilt / swing)
Paul_Hannover wrote:
I made some better light testing this morning and had the following findings:
- OSS on or off doesn't really matter for my results
- The problem is not only at F4, but also at F7.1 still
- i would say the left 25% of the image are affected
- the siemens star dot centering test as suggested by Roger Cicala does not show any asymmetries
Below you see a photo taken at 24mm F5.6, please see the trees on the left side compared to those on the right side of the image.. the picture really detiorates in the left.. and that is not even wide open. (Image has been compressed to get below the filehost's 10mb limit)
Paul - I had a similar problem with a used 24-105 that I bought off the forum. Luckily for me, the lens was still under warranty and the seller provided an invoice. I called Sony, told them the problem, sent it in to Precision Camera (on their dime), and they repaired it and it came back sharp with no significant decentering/tilt. If your lens is under warranty hopefully they can fix it for you.
p.4 #9 · FM Lens test for Centering (Mount tilt / swing)
Paul_Hannover wrote:
I made some better light testing this morning and had the following findings:
- OSS on or off doesn't really matter for my results
- The problem is not only at F4, but also at F7.1 still
- i would say the left 25% of the image are affected
- the siemens star dot centering test as suggested by Roger Cicala does not show any asymmetries
Below you see a photo taken at 24mm F5.6, please see the trees on the left side compared to those on the right side of the image.. the picture really detiorates in the left.. and that is not even wide open. (Image has been compressed to get below the filehost's 10mb limit)
If I have done the test right, it seems to me that the lens is decentered on the left.
(1) Have I done the test wrong?
(2) Or am I reading the test wrong?
(3) From what I gather, I should expect some degree of decentering with a fast WA zoom like this. Would this level of decentering be acceptable to you? Or would you get the lens calibrated fixed?
If I have done the test right, it seems to me that the lens is decentered on the left.
(1) Have I done the test wrong?
(2) Or am I reading the test wrong?
(3) From what I gather, I should expect some degree of decentering with a fast WA zoom like this. Would this level of decentering be acceptable to you? Or would you get the lens calibrated fixed?
Would this level of decentering be acceptable to you?
Now that's the $1000 question.
These lenses are mass produced to a specified level of tolerance. So what is acceptable? Perfect on all four corners?
I once read that we need to just take pictures and look at them; if we like the image, then fine. If the image quality is poor, have it fixed.
Your lens softness looks better than my Sony 85GM. See the photos of the FM method using a tree top. Images are posted on this thread a few weeks ago. My 85prime looks worse than yours. I am considering sending it to Sony for adjustment.
p.4 #12 · FM Lens test for Centering (Mount tilt / swing)
s4ducati wrote:
Would this level of decentering be acceptable to you?
Now that's the $1000 question.
These lenses are mass produced to a specified level of tolerance. So what is acceptable? Perfect on all four corners?
I once read that we need to just take pictures and look at them; if we like the image, then fine. If the image quality is poor, have it fixed.
Your lens softness looks better than my Sony 85GM. See the photos of the FM method using a tree top. Images are posted on this thread a few weeks ago. My 85prime looks worse than yours. I am considering sending it to Sony for adjustment.
Indeed, that's a subjective! It's one of those things that those with experience do not sweat, because they (a) understand the natural variation of lenses, good and bad; (b) recognize how bad "bad" needs to get before it impacts images to their satisfaction; and (c) recognize how much "good" is "good enough".
Those of us, like myself, that are less experienced can only hope to borrow their experience/standards.
This lens is still under warranty, so I can send it to be fixed/calibrated. Assuming that I am able to persuade the engineers that this needs to be calibrated and the case does not get dismissed as "within manufacturing tolerances"
p.4 #13 · FM Lens test for Centering (Mount tilt / swing)
I wanted to ask if anybody recognized a decentering over time. Means you test a lens and it's perfectly centered but over the time the results shift into negative territory!? I heard this phenomenon from owners of the 35mm f2.8 Sony Zeiss. No idea if this could be a contruction issue of this specific lens. Provided you handle the lens like a raw egg, how high are chances that lenses loose correct centering values over time?
p.4 #14 · FM Lens test for Centering (Mount tilt / swing)
jeetsukumaran wrote:
Indeed, that's a subjective! It's one of those things that those with experience do not sweat, because they (a) understand the natural variation of lenses, good and bad; (b) recognize how bad "bad" needs to get before it impacts images to their satisfaction; and (c) recognize how much "good" is "good enough".
Those of us, like myself, that are less experienced can only hope to borrow their experience/standards.
This lens is still under warranty, so I can send it to be fixed/calibrated. Assuming that I am able to persuade the engineers that this needs to be calibrated and the case does not get dismissed as "within manufacturing tolerances"
So I sent the lens in to Sony under warranty for adjustment. Just got it back. The note says: "Resolved power check/optical axis. A complete repair and return all functions to factory specs. Current FW ver 3. Complete cleaning". Will re-do the test and see how it does!
p.4 #16 · FM Lens test for Centering (Mount tilt / swing)
I'm curious as I want to do this test, but how do you know, since it's handheld, that the difference you see is not from the user moving? I guess on a far away subject it will make less of a difference.
NVM I just re read again and it's too be tested at infinity! Time to test my lenses 🤙
p.4 #17 · FM Lens test for Centering (Mount tilt / swing)
71sbeetle wrote:
I'm curious as I want to do this test, but how do you know, since it's handheld, that the difference you see is not from the user moving? I guess on a far away subject it will make less of a difference.
NVM I just re read again and it's too be tested at infinity! Time to test my lenses 🤙
When shooting at very high shutter speeds and your subject static and at infinity, it's like shooting from a tripod.
I usually do these tests under contrast light and my shutter speeds are close to 1/8000s.
If you try it once and repeat the experiment you will see that you'll always get the same results.
Use manual exposure, turn-off IBIS, focus at infinity once (at center) and do not re-focus until you take the 4 pictures. Try using high shutter speeds. (>1/4000s)
p.4 #18 · FM Lens test for Centering (Mount tilt / swing)
Fred Miranda wrote:
When shooting at very high shutter speeds and your subject static and at infinity, it's like shooting from a tripod.
I usually do these tests under contrast light and my shutter speeds are close to 1/8000s.
If you try it once and repeat the experiment you will see that you'll always get the same results.
Use manual exposure, turn-off IBIS, focus at infinity once (at center) and do not re-focus until you take the 4 pictures. Try using high shutter speeds. (>1/4000s)
Fred,
Can this method give a false positive?
I used this procedure to test a 85mm f/1.4 GM and got soft corners on the right (top and bottom) while tack sharp on the left; as well as center.
I then setup an ISO chart (using the process described by RogerC) on my wall and was careful to align the camera square to the target, using a framing square and strings. The results were all four corners were sharp.
I assume if I mis-aligned the camera to the ISO chart it would give false positive. But that did not happen here.
p.4 #19 · FM Lens test for Centering (Mount tilt / swing)
s4ducati wrote:
Fred,
Can this method give a false positive?
I used this procedure to test a 85mm f/1.4 GM and got soft corners on the right (top and bottom) while tack sharp on the left; as well as center.
I then setup an ISO chart (using the process described by RogerC) on my wall and was careful to align the camera square to the target, using a framing square and strings. The results were all four corners were sharp.
I assume if I mis-aligned the camera to the ISO chart it would give false positive. But that did not happen here.
Alignment is no longer a variable when doing this test at infinity.
I've never got a false positive using this test. If all four corners look the same, the lens is centered. However, if the lens is decentered or tilted, the test will always detect it regardless of how many times you try.
