This is my personal story with Sony lens warranty, and not purely gear advise. I posted about this on DPReview last year and got some helpful input, but I'm curious to hear this community's take as well.
I bought a A7IV together with the 24-70 GM II a few years ago and was happy with the setup. After a while — probably as my shooting style shifted towards longer focal lengths — I noticed that my shots were not consistently sharp comparing the left and right edge of the frame. I mean that the left edge was considerably sharper than the right edge. This was most visible towards the long end of the zoom range, but also at larger apertures (see test shots linked below). I know that lenses are usually softer the more you move away from the center, but here the softness was inconsistent between different sides of the frame.
I remembered some stupid mistake I did right after I got the camera: I tried to mount the 1.4x teleconverter backwards, attaching the lens side to the camera body. This meant the tube that normally extends into the rear of the lens was instead pressing against the sensor. I twisted gently before realizing my mistake. I don't think I used force, but I know this could have bumped something out of alignment.
I did a few test shots. This looked like decentering. I contacted Sony here in Switzerland and I could send it in for repair. The repair center is third party operated and does anything from cameras over playstations to washing machines. They took the lens and my camera. It was important to me that they check the camera as well and I explicitly asked them to check sensor alignment. They found nothing wrong with the camera, but replaced some washers on the lens. I tested the lens and the sharpness was basically the same as before. So I sent it back in and they didn't find anything wrong this time. Working "within parameters". On the bright side both inspections and the repair was free of charge.
In the meantime I bought a Tamron 28-75 G2. I did the same tests and the lens showed perfect sharpness on the right edge, but slight softness on the left edge. Exactly opposite the Sony lens! At this point I concluded the camera was in fact fine and my Tamron was just also slightly decentered. Since the Tamron was a lot cheaper, I just accepted this.
Since my Sony GM lens still wasn't sharp, I send it to a private lens repair shop over in Germany. Interestingly they told me the lens has some fall damage and some of the "glass bearings" (probably not the right words) were shattered. I didn't have any accident with the lens, also it had just been at Sony's repair shop. I don't know if this was a shipping damage, but I did pack the lens well. I got the lens back and ... still the same softness. The repair shop handled this well and offered to look at the lens again for free, but couldn't find anything wrong with it now.
So this leaves me with a few options: 1) the lens is indeed fine, the camera is fine, this is just tolerance. 2) those GM lenses are just very hard to align correctly 3) the camera sensor or bayonet is misaligned and the sony repair shop missed it.
I did not find any other repair shop that looked capable. At least not in Switzerland or Germany.
This all happened last year and I have been using the Tamron lens exclusively since then. On the technical side I have a quite expensive Sony lens that just seems to underperform and a Tamron lens that is quite adequate. On the personal side this whole journey somehow took the joy out of using my Sony gear and I am thinking of selling it all and starting fresh.
I am curious to hear your thoughts here. What could be wrong? Did you ever encounter unfixable issues?
A misaligned camera mount is more likely to affect wider focal lengths, so it's likely the lens. Asymmetric sharpness s usually caused by a tilted lens element rather than decentering, although many use the terms interchangeably.
It's a dirty secret of the industry that many expensive lenses come out of the factory with subpar symmetry. Some claim resolving this would result in the lenses being even more expensive, due to the yield and labor costs involved. I'm of the opinion the practice continues simply because users generally allow and accept it. Most repair facilities will claim ignorance and gaslight you that the lens is fine, especially manufactuers' facilities. They have a vested financial interest in doing so.
There are well-centered copies out there for every lens design. The question is how much effort and trouble each individual is willing to expend to find one.
Lenses vary but it is also easily possible to chase a defect that just isnt't there or not of practical importance. In theory lenses are corrected for a flat field, in practice one takes photos of different subjects with a 3D arragement of surfaces. Then there is the distance at which a lens is tested, it might perform better at infinity.
What do you think about posting some test photos? 100% crops of the center and the corners, distance not too short, e.g. 5meters. One set focussed in the center, on set focussed where the crop is from. Open aperture and 2 stops closed.
This is not just a Sony problem. In the past, I have had some significant de-centering problems with Sony, Nikon and Canon. I always had the same response from OEM repair: "within factory spec".
Now, I thoroughly test my lenses right after purchase, so I can exchange directly with the dealer, immediately. The first GM 24-70mm f2.8 II I purchased was noticeably de-centered, but the second was much better (still not perfect). As a note, this particular lens has a reputation for de-centering problems, based on a number of forum posts, I have read. I tried a Tamron 28-70mm f2.8 and it wasn't great, either.
Again, your greatest insurance is to thoroughly test each lens right after purchase.
I have noticed that Sony's more recent releases have been much better than in the past. My GM 28-70mm f2 and GM 85mm f1.4 II lenses have been darn near perfect.
The fact is, most folks do not even notice de-centering problems, which is why wide factory tolerance bands (Sony, Canon & Nikon) are essentially accepted by the buying community. I assume that, to tighten manufacturing tolerances, would increase the cost of these lenses.
This is a few hundred meter away. The focus is always set to the center, then switched to manual focus without refocusing when moving the subject to the side.
I could to another round of shots where I actually refocus every time, but then again, I do expect a corner to corner focused image at those distances.
InFocus2014 wrote:
This is not just a Sony problem. In the past, I have had some significant de-centering problems with Sony, Nikon and Canon. I always had the same response from OEM repair: "within factory spec".
What did you do with those lenses? write them off as a loss?
InFocus2014 wrote:
Now, I thoroughly test my lenses right after purchase, so I can exchange directly with the dealer, immediately. The first GM 24-70mm f2.8 II I purchased was noticeably de-centered, but the second was much better (still not perfect). As a note, this particular lens has a reputation for de-centering problems, based on a number of forum posts, I have read. I tried a Tamron 28-70mm f2.8 and it wasn't great, either.
Again, your greatest insurance is to thoroughly test each lens right after purchase.
That sounds like a solid strategy. Unfortunately, no retailer offers free returns in Switzerland
Maybe it’s my eyes — but none of those pictures look sharp anywhere in the frame. Sure the ones focusing in the corners there appears to be a sharpness difference but maybe it’s the atmospheric haze in the scene… or just a poor scene to test a lens against. I’d probably choose a scene with more contrast / less haze to test against
Jazzgear296 wrote:
Maybe it’s my eyes — but none of those pictures look sharp anywhere in the frame. Sure the ones focusing in the corners there appears to be a sharpness difference but maybe it’s the atmospheric haze in the scene… or just a poor scene to test a lens against. I’d probably choose a scene with more contrast / less haze to test against
I agree. As a test it would be much better to find a object in middle distance (say 5-10 meters, preferably in good light). I would also like to see the whole frame to better understand how much sharpness falls down when going to the corners.
I have used two copies of 24-70GM II. I was a little concerned about the first one being decentered in wide end. Taking a good 24mm test test shot is not that easy... After a while I started to trust in that lens, and it was fine in practice as is the second one. I am not saying that there are no dud copies around, but Sony seems to have done pretty good work with this one.
I second the idea of testing a lens upon receipt. The optical quality of the Tamron 28-75G2 seems to be often underestimated. I have a copy of the Tamron and a Sigma 24-70 DN II, the tamron has better sharpness and focus accuracy. I don't think there is anything wrong with the Sigma I have.
The testing can be a bit tricky. I like to use a large brick wall or the wall of a stadium for more distance. The critical part is getting the lens perpendicular to the wall. Testing at infinity, especially with stars at night also works.
Nifty Fifty wrote:
Some people wonder if we're obsessed with fast lenses. I wonder if we're obsessed with decentering tests.
I think OP has detailed a clear lack of obsession. It's all there to read, it took significant use to identify a problem which was then consistent. No decentering test was attempted before that time.
It's frustrating, but typical.
You sent the camera with the lens for the first repair attempt.
So the repairer had a minimum of 2 bodies to test.
Softer on one edge at longer focal lengths.
So couldn't they at least report they tested on 2 bodies and whether the same edge was softer on both?
Or that they did not notice one edge softer on either?
Jazzgear296 wrote:
Maybe it’s my eyes — but none of those pictures look sharp anywhere in the frame. Sure the ones focusing in the corners there appears to be a sharpness difference but maybe it’s the atmospheric haze in the scene… or just a poor scene to test a lens against. I’d probably choose a scene with more contrast / less haze to test against
I tried something at infinity focus since this was the setting when I noticed it first. I'll try to find a mid distance subject and wait for good light here and report back.
That's ill advice for testing sharpness of a lens.
You want to test close to infinity as Thomas correctly did, since this is the most challenging distance for any lens and it will reveal any issues such as tilt, decentering, contrast, and general lack of sharpness or contrast, etc. At very close distance, and this is unfortunately how Swiss Sony repair partner Sertronics and Co. are testing customer lenses with a dumb Siemens Star chart, real issues are usually masked, except for more extreme cases such as very strong tilt defects. Heck, any decent lens and copy will look good at 3-5 meters, severe tilt problems aside.
If I spend over 1000 USD for a G or GM lens today, I expect excellent landscape performance as one its main traits, except maybe for Macro lenses, no matter if zoom or prime. I have tested surely over 100 lenses and often various copies of the same lens with identical testing setup, and the difference in sharpness and centering performance between the same lens model - be it new or old design, or low-end or high-end (G/GM) - can be shocking at times. So this is real issue, and has always been for decades with any lens manufacturer. Even just this forum alone hosts many examples, tests and reports about this issue. Those claiming that lens sample variation is a myth or overblown issue simply don't know what their talking, i.e. chatty armchair "experts" or just simpletons
thomas001le wrote:
What did you do with those lenses? write them off as a loss?
That sounds like a solid strategy. Unfortunately, no retailer offers free returns in Switzerland
To answer your question, I eventually sold them. Fortunately, the buyers were very pleased with the lenses (if they were not, I would have accepted a return and refunded their money plus shipping costs). This helps prove that most folks have no problem with de-centered lenses.
