Sony's (lack of?) lens quality control

[jonoslack]

N,

#1 It is not a rant, it's a discussion, if I was actually ranting it would be very obvious, trust me.

#2 I am glad to hear you are familiar with statistics, thus you should know why this is not a sample set of 5. You can also consider following: What is there to assure us sample set it not much bigger but we just didn't hear about it? What makes you confident that we didn't hear about it because owners do know how to test for it, did test it, and it did pass the test? Could it be that people are having them without being aware there is an issue? Wouldn't be the first time.

HI Zoran
Look at it this way -

About 10 years ago I took 5 successive copies of the Nikon 17-55 lens back to a professional camera supplier in Cambridge - they quietly replaced them, agreeing with my samples that it was definitely not OK (I wasn't familiar with the term 'decentered' then, but I sure knew what 'soft' meant )

When I finally got a good one, I asked their Nkon specialist if they often got lenses returned from customers - he said that I was the only person that year who had returned one.

Of course, I can't prove it, but it seems likely that there were lots of sub-standard lenses going out (or at least less than optimal). It's just that people weren't looking, added to which, back then, most pros were using film, and it was less easy to realise.

Whatever - I have quite a lot of statistics as well, and as N says - these are not statistical samples.

In this case lots of people are buying these new lenses, and they have been warned about sample variation and are looking for it.

Back to Roger Cicala's excellent article at lens rentals: All of these lenses vary quite widely in quality, and they are all going to be decentered to a greater or lesser degree - only you can decide what you find acceptable, but I'd be really surprised if Sony's QC was worse than others (especially since long personal experience (not statistically significant sadly) has shown me how reliable their cameras seem to be).

all the best

 

[jonoslack]

Have to say that I am having difficulty accurately assessing the 24-70.
Question.
1.Are others still finding it soft on edge/corners even when stopped down to F8 at the wide end of the range (24-32.
2. Does having OSS off make a difference to IQ when at higher shutter speeds. In other words is it better to have it switched off.
Thanks

HI David
I'm finding mine pretty hard to evaluate as well - but mine is softer on the edge/corners even when stopped down (perhaps more so) - but not very soft, just not as sharp as the centre. Mine is also very slightly less sharp on the right edge at 70mm, but seems to be even at other focal lengths. On the other hand it seems to come up with the goods. I do feel that curvature of field is often the problem, but that doesn't seem to be very consistent either!

It's a difficult balance between being realistically careful and pointlessly picky. Generally speaking mine sounds to be similar to Tim's.

 

[Viramati]

1) Yes, but much less so than the Nikon 24-70 F2.8 - one thing to try: because of odd DOF and possible midl focus shift and changing field curvature effects, the wide end often has better edges at F4 than F8. Try it on a couple of scenes and see how yours is.
2) it seems not to make any difference I can see. It even seems fine with a tripod and OSS on. But I do try to remember to switch OSS off when using tripod or higher shutter speeds, just in case!

Thanks Tim. I had oddly found that sometimes edges are better at F4 but it seems to vary so I was wondering about OSS. I will go back and look at some of my old D700 and 24-70/2.8 images to try and get an idea of what I am looking for. It could also be that I have only used primes for the last few years and am just expecting to much.

P.S. Have started to look at my old nikon images and I have to say that the Nikon 24-70/2.8 is one hell of a lens (at least on the D700)

 

[Viramati]

HI David
I'm finding mine pretty hard to evaluate as well - but mine is softer on the edge/corners even when stopped down (perhaps more so) - but not very soft, just not as sharp as the centre. Mine is also very slightly less sharp on the right edge at 70mm, but seems to be even at other focal lengths. On the other hand it seems to come up with the goods. I do feel that curvature of field is often the problem, but that doesn't seem to be very consistent either!

It's a difficult balance between being realistically careful and pointlessly picky. Generally speaking mine sounds to be similar to Tim's.

Yes i also wonder about field curvature as this is obviously a very complex design and seems to have some significant distortional defects that need correcting in PP. Back in the good old days of film they wouldn't have been able to get away with a lens like this !!!

 

[jonoslack]

Yes i also wonder about field curvature as this is obviously a very complex design and seems to have some significant distortional defects that need correcting in PP. Back in the good old days of film they wouldn't have been able to get away with a lens like this !!!

. . . but then we wouldn't have the lens, it isn't just that they couldn't have got away with it, they couldn't have made it at all.

Let's face it, the results are pretty sparkly

 

[f/otographer]

And all of this is one reason I choose to shoot with 30 and 40 year old lenses. Nothing is without its pitfalls, but I have found that many of these legacy lenses were manufactured to higher standards originally and have stood the test of time well.

 

[Jeff Kott]

By the way, here's an article written by Roger on a fairly simple test for determining if your lens is decentered. I though this might help some of us.
LensRentals.com - Testing for a Decentered Lens: an Old Technique Gets a Makeover

 

[jonoslack]

By the way, here's an article written by Roger on a fairly simple test for determining if your lens is decentered. I though this might help some of us.
LensRentals.com - Testing for a Decentered Lens: an Old Technique Gets a Makeover

HI Jeff
Thanks for that - It's actually the first test I used with the 24-70 (flying colours). not absolutely certain it's good for all distances, but it's certainly a start.

What would we do without Roger - a jewel of common sense and information in an internet world of misconception and disinformation.

 

[jonoslack]

And all of this is one reason I choose to shoot with 30 and 40 year old lenses. Nothing is without its pitfalls, but I have found that many of these legacy lenses were manufactured to higher standards originally and have stood the test of time well.

Yes, a good lens is a good lens . . . . . but vintage lenses have their fair share of pitfalls too (including decentering). I wonder whether QA was really better in the good old days?

 

[Viramati]

. . . but then we wouldn't have the lens, it isn't just that they couldn't have got away with it, they couldn't have made it at all.

Let's face it, the results are pretty sparkly

I take your point.

 

[philip_pj]

What would you choose if Sony offered a special release series of accredited lenses for say $100 extra, in addition to the standard output lens? These could come with a certificate and test chart.

So the FE 55mm might have a small run added to the general production line which could be offered for sale for say $1100 rather than the regular price of $1000? I chose that lens deliberately because I have lost count of the number of internet experts who would 'never pay $1000 for just a normal lens'. Sony is already up against buyer pressure to constrain costs, and is being kicked for making such a fine lens at a still quite fair price.

I'd pay the surcharge in a heartbeat but that's me, living in splendid isolation here in Australia where we already pay much more for most photographics in a small market.

Special series might be extended to lenses that have no cheaper counterparts, which is really what Zeiss and Leica do anyway. Zeiss already test each copy of the 15mm ZEF Distagon for MTF, they told Lloyd that. I bet the Otus is very well assembled and QA is very good.

What is happening is that buyers range from folks like Tim and Zoran right through to folks who just start shooting with whatever they buy and get upset only at really obvious defects. The makers try to strike a balance with a wary eye on their bottom line.

 

[Annna T]

What would you choose if Sony offered a special release series of accredited lenses for say $100 extra, in addition to the standard output lens? These could come with a certificate and test chart.

So the FE 55mm might have a small run added to the general production line which could be offered for sale for say $1100 rather than the regular price of $1000? I chose that lens deliberately because I have lost count of the number of internet experts who would 'never pay $1000 for just a normal lens'. Sony is already up against buyer pressure to constrain costs, and is being kicked for making such a fine lens at a still quite fair price.

I'd pay the surcharge in a heartbeat but that's me, living in splendid isolation here in Australia where we already pay much more for most photographics in a small market.

Special series might be extended to lenses that have no cheaper counterparts, which is really what Zeiss and Leica do anyway. Zeiss already test each copy of the 15mm ZEF Distagon for MTF, they told Lloyd that. I bet the Otus is very well assembled and QA is very good.

What is happening is that buyers range from folks like Tim and Zoran right through to folks who just start shooting with whatever they buy and get upset only at really obvious defects. The makers try to strike a balance with a wary eye on their bottom line.

I think that certifying a perfect lens enters in the territory of diminishing returns and that a 10% price increase isn't enough to make it profitable for the manufacturer. And what would that mean with respect to the rest of the production ? If you don't buy a certified lens you are then certified to get soso ones ? For 10% all the people will want the certified and the manufacturers will end with too many rejects and increasing cost.. There is a reason why good lenses are much more expensive : tight tolerances, much more rejects and increasing production cost.

 

[henningw]

There is no lens that I have used and tested that didn't have a visible fault.

I've been shooting with decent equipment for over 50 years, and 40 years professionally. Maybe 500 different lenses for all sorts of formats. I've used Leicas since 1961 (including about 75 of their lenses) and pretty much every manufacturer's products in almost every format. Failed QC has gone from missing elements (twice), inverted elements (twice) to hairs and other junk in the lenses to most commonly, decentering. The more expensive German offerings (Leica, Zeiss, Schneider and Rodenstock) have been a fair bit better in this regard than Japanese mainstream produces, but hardly immune to flubs. For about 10 years I used the Konica Autoreflex system for my SLR requirements, and it was the only system that never needed repairs, worked to -40° and where I only twice readily noticed a decentered lens, but I wasn't as fussy then and that's not statistically significant anyway.

Whatever: all camera systems have problems at times, and when you have a 36mp sensor and can easily pixel peep every image you had better be prepared to pay close to five figure prices for every moderately complex lens if you want no easily seen decentering in every lens. Look at what cine lenses cost.

I check equipment after I buy it, but I also keep in mind what I intend to do with it. A macro lens that I will use for copy work better not show any decentering at f/8 at any magnification that I'm likely to use. A portrait lens can show a bit of decentering at a couple of wider apertures. I have one Canon lens which was selected as the best out of 6 that still has very noticeable decentering, but which is useful enough that I've kept it and used it for 15 years.

Professionally, I've always tried to be fussier than any of my clients. Being a huge amount fussier than your clients though isn't really worth it. For my own shooting, I can set my own standards, and those are often a lot higher, but I still try to keep realistic.

Print something at 24x36 and see if the results bother you. Then decide how much more you're willing to pay for something better.

Henning

 

[Guy Mancuso]

Well said and I agree. As much of a lens whore as I am I'm only willing to part with so much money and to pay thousands for incremental improvements sometimes seems hardly worth the effort. In short there always seems to be a alternate that works very close to your needs. I want great glass no question but I'm only willing to go so far anymore to get it. I have been down that road spending 7 k on one lens for optimum results. Great to have it but again clients don't always appreciate it either. There is a line in the sand and it's up to you to decide what's the true value of crossing that line. Like everyone else I love great glass but I find a good percent of the time if I go with a Zeiss lens I won't be disappointed. As I think they make some of the best glass around in old and modern mounts and they still are reasonable to purchase. The days of spending over 6k for a Leica R 35-70 2.8 rare as can be lens are over. I can look back and say I owned some of the best glass ever made like this one but again at what cost to get there is where I lost the interest. Buying lenses is a crap shoot anymore and sometimes buying three to get one great one is more the case than not but I have a hard time shelling out the big bucks as well. Maybe its a growing process many of us go through but in the end I am no worse off with less expensive glass than I was with 6 k per copy. When in doubt I always seem to find Zeiss glass at a reasonable cost plus great manual feel to boot that usually solves my quality standards. But until you print this stuff big it's really a unknown factor.

 

[Viramati]

So I went to the camera store today and tried out another FE24-70 doing comparison shots at different focal lengths and f stops, Strangely both seemed similar at 24 and 70 but at 35mm the 2nd copy was noticeably sharper near the edges especially on the left compared to my first copy. So obviously have have changed the lens and will now see how this one performs over the next few days

 

[jonoslack]

So I went to the camera store today and tried out another FE24-70 doing comparison shots at different focal lengths and f stops, Strangely both seemed similar at 24 and 70 but at 35mm the 2nd copy was noticeably sharper near the edges especially on the left compared to my first copy. So obviously have have changed the lens and will now see how this one performs over the next few days

HI David
That sounds like a result - I'm with Tim about this lens - it's got some really nice characteristics, and some less likeable ones, but overall it seems to be producing nice shots for me, and that's the best one could expect.

all the best

 

[ZoranC]

HI Zoran
Look at it this way -

About 10 years ago I took 5 successive copies of the Nikon 17-55 lens back to a professional camera supplier in Cambridge - they quietly replaced them, agreeing with my samples that it was definitely not OK (I wasn't familiar with the term 'decentered' then, but I sure knew what 'soft' meant )

When I finally got a good one, I asked their Nkon specialist if they often got lenses returned from customers - he said that I was the only person that year who had returned one.

Of course, I can't prove it, but it seems likely that there were lots of sub-standard lenses going out (or at least less than optimal). It's just that people weren't looking, added to which, back then, most pros were using film, and it was less easy to realise.

Whatever - I have quite a lot of statistics as well, and as N says - these are not statistical samples.

In this case lots of people are buying these new lenses, and they have been warned about sample variation and are looking for it.

Back to Roger Cicala's excellent article at lens rentals: All of these lenses vary quite widely in quality, and they are all going to be decentered to a greater or lesser degree - only you can decide what you find acceptable, but I'd be really surprised if Sony's QC was worse than others (especially since long personal experience (not statistically significant sadly) has shown me how reliable their cameras seem to be).

all the best

Hi jonoslack,

#1 Let's agree to disagree.

#2 This discussion was not about was Sony's QC better or worse than others, it was about personal experiences and what each individual feels about it.

#3 IMHO if five random persons perceive issue with two out of three random copies chances are so will sixth, seventh, eighth ... person with two out of six copies.

#4 Reliability of cameras has nothing to do with QC of lenses, those are two separate production lines.

#5 Do you call number of reports of misaligned lenses on RX100 perceived issue with camera or perceived issue with lens, and can they be considered reliable?

#6 Were you really expecting Nikon rep will admit there is an issue when you asked him? Have you ever seen any rep ever do that?

Regards,

Zoran

 

[ZoranC]

What would you choose if Sony offered a special release series of accredited lenses for say $100 extra, in addition to the standard output lens? These could come with a certificate and test chart.

#1 I would want to see does paying extra really deliver me that value.

#2 I would gladly pay $200 if it does.

#3 I would still check lens I receive. Manufacturer might have made it "perfect" before it left their hands but they have no control how that package was handled during transportation (you should have seen how poor packaging of some shipment I got from Amazon was) or did I get copy that was handled and returned by somebody else.

 

[ZoranC]

Print something at 24x36 and see if the results bother you. Then decide how much more you're willing to pay for something better.

But until you print this stuff big it's really a unknown factor.

Can somebody please tell me what is the purpose of manufacturing/buying 36 MP bodies if one is not planning to use all of those pixels and is instead planning to reduce their gear to 12 MP or whatever is the lowest common denominator of the wekest link in their system chain?

Isn't it like saying "I purchased 200MPH car that manufacturer offers tires for that are capable of doing only 90MPH and that's OK because I won't be going over 55 anyway"?

 

[jonoslack]

Hi jonoslack,

#1 Let's agree to disagree.

Certainly

#6 Were you really expecting Nikon rep will admit there is an issue when you asked him? Have you ever seen any rep ever do that?

It wasn't a Nikon rep - it was the Nikon specialist in KP pro, and I'd been dealing with him for years . . . . . and he had no possible reason to lie!

Just saying