Previous news story    Next news story

Roger Cicala investigates accuracy of lens adapters

By dpreview staff on Sep 30, 2013 at 11:00 GMT

Mount adapters are incredibly useful for their ability to make lenses from one manufacturer usable with camera bodies from another. But do they have any impact on image quality? LensRentals' Roger Cicala, not one to take manufacturer's claims at face value, has done some investigation. Knowing that slight mis-alignments between a lens and even its native mount can cause softness in images, the added complexity with a lens adapter in the mix seemed likely to cause more problems. His findings are indeed interesting. 

In his blog post, Cicala takes a thorough and technical look at the effects of lens adapters on the measured edge-to-edge sharpness of a lens. His focus is mainly on whether commercially-available adapters would be good enough for testing lenses on an optical bench, which is a pretty demanding application. As usual, though, he brings a unique insight to the issue, from being able to test a large number of high quality adapters from the likes of Metabones and Novoflex.

In real-world use shooting 3D subjects (rather than planar test charts), we've found very little evidence of lens adapters having a serious impact on image quality. Still, his report is worth reading if you're interested in adapting your camera to accept different lenses.

Source: LensRentals


Total comments: 48
By alanscape (6 months ago)

I have posted a question on LensRentals site, asking if the Olympus MMF-3 would qualify as "acceptable for testing...............".
Owners of lenses like mine (the 7-14mm up to the 90-250mm) which I use on the OM-D E-M5 and E-M1 may find LensRental's reply interesting when and if they do answer.

E Dinkla
By E Dinkla (6 months ago)

The method I used is buying inexpensive but rigid adapters and tweak them for one lens so infinity is correct and image quality fall off to the sides is more or less equal. And keep that adapter fixed on that lens. This is not in conflict with Roger's findings. On M4/3 so far and with low weight C-mount lenses so even more within a safety zone. Despite that, two identical lenses gave different results. Most that go this route will use older and often secondhand lenses with different histories. There will always be a lottery aspect in this approach but it can be quite rewarding. Not just in a nice image quality but also to slow down your picture taking and manual focusing, aperture priority shots often deliver my best images.

Ernst Dinkla

By JDThomas (6 months ago)

I clicked on the link, saw MTF charts and immediately fell asleep. Can someone summarize it for me?

By PhotoKhan (6 months ago)

I don't get this...

If the only purpose of an adapter (the "pure" ones, not something like Metabones) is to reinstate the original Flange Focal Distance a lens was built to work with, why are they optically-bench-testing them?

The only appropriate testings to be done are mechanical and physical-dimensional ones, ie, how dimensional-accurate (and consistently throughout production output) is a particular adapter.

(I guess that would mean getting some measuring equipment they DON'T have...)

Comment edited 37 seconds after posting
1 upvote
By Murmillo (6 months ago)

People think that adding lens-free space between a lens & a camera body (for eg. extension tubes for macro; or these adapters) has no effect on IQ; but they are way wrong. Lens design includes not just flange distance & alignment but also depth of focus (not just DOF) which if not precise affects IQ on multiple levels. Some lens adapters have lens elements involved, which further complicates matters. So yes, MTF measurements, including measurements for LCA & LoCA are in order.

1 upvote
By PhotoKhan (6 months ago)

I explicitly excluded adapters with lenses (...for obvious reasons. More glass SHOULD equate to some amount, however minor, of optical degradation).

As "for adding lens-free space in between a lens & a camera body" it is the EXACTLY the same space the lens was designed to work with in the first place, right?

Comment edited 2 times, last edit 5 minutes after posting
By JensR (6 months ago)

You are right, Khan!
I share your assumption: They are lacking the necessary equipment to check the physical dimensions precisely.

1 upvote
By Klipsen (6 months ago)

Oops! I believe I misunderstood the subject ;-)

Comment edited 6 minutes after posting
By DVT80111 (6 months ago)

I didn't read the article but just went to the summary

"Putting a great lens on your camera via an adapter might still be better than an average native-mount lens. On the other hand, that great lens certainly wouldn’t be as good as it would be on its native-mount camera."

Since there is no native mount body for legacy lenses, it is a moot point.

By 24hrexposure (6 months ago)

This is a good quote from the article:

"What Does It Mean in the Real World? Like a lot of laboratory testing, probably not a lot. Adapters couldn’t all stink or people wouldn’t use them. Like a lot of tests, you can detect a very real difference in the lab that doesn’t make much difference at all in the real world."

Nuno Souto
By Nuno Souto (6 months ago)

Here we go again with the insane "optical bench tests"...
When will this kind stop being taken for granted?
They are destroying photography!

Let me see: a lens adapter mount has a major effect on sharpness, but the entire flip-mirror assembly in a slr doesn't and never has done so?
Oh wait: that flipped up piece of gear is the one primarily responsible for focusing said lenses?

Something that gets flicked up and down at high speed, is "inherently" more exact in its sub-micron positioning than a flat ring of metal held firmly in place by 6 or more screws?

Yeah sure. And I am henceforth to be addressed as Santa Claus.

PS: to the "dear Santa" commentators: wake up and start taking photos of real subjects. Not optical benches!

Mike CH
By Mike CH (6 months ago)

Are you quite sure that you know what exactly the mirror actually does?

Nuno Souto
By Nuno Souto (6 months ago)

Are you sure you do? Here is a hint: phase detection is the primary AF mechanism for dslrs. NO, they do NOT use only the imaging sensor, you are confused.

Comment edited 2 minutes after posting
Eric Hensel
By Eric Hensel (7 months ago)

Many of us understand the usefulness of standards, and quantification versus guesswork; and are also able to use less-than perfect equipment for excellent work. This isn't an either/or thing. I enjoy reading, and knowing, and refuse to let that stop me from using what I have. This kind of knowledge also helps one to work creatively within the limitations of equipment, when certain results are critical.

Comment edited 2 minutes after posting
By hmzppz (7 months ago)

There is of course no denying that many adapted lenses do not perform as well as on their native bodies. However, my humble opinion is that Mr Cicala is making wrong conclusions here. Adapters are not the main reason for the loss of image quality. The main deterioration occurs when a lens with a steep incident angle (like wide-angle rangefinder lenses) is used with a digital sensor. Leica makes (expensive) arrays of microlenses exactly to address this issue. In contrast, telecentric lenses (such as SLR lenses) should and do perform rather well.

That said, I secretly hope that after reading this article people will start selling their adapted lenses, so I can finally afford to buy some of the awesome Contax glass :)

By NancyP (7 months ago)

Roger also has the highest number of test subjects of any lens reviewer around. I want to read about "average" members of a specific lens design, not a single copy of a lens.

I think that his point is obvious - no free lunch - but if your use for an adapted lens is at f/8 for a landscape, perhaps with focus stacking, you may be able to tolerate minor losses caused by mis-alignment.

By 3systermuser (7 months ago)

now every time I buy a lens , check out his opinion, he is the best and most honest lens tester in the world.
Roger gets all credits and his work should be respected.

Comment edited 57 seconds after posting
By jhinkey (7 months ago)

Knowledge is power for those able to handle it. Yes, most people would not even know these issues were going on, but it is something to be aware of so that if you do have some off-axis sharpness issues you can perhaps know where to start looking.

I only had wished that he had taken more detailed data - like which direction the IQ problems are in the frame, decentering versus focal plan mis-alignment, etc.

Good stuff to know.

By Anfy (7 months ago)

I tried a Sigma 8-16mm f/4.5-5.6 F-mount on a Sony NEX-5n with a "pro" class adapter and saw some softness on the right side of the image, that - in more controlled shooting - I was not able to replicate with my Fujifilm S5 Pro (shot with the camera in portrait position, and then turned it upside down).
- the lens is an extreme design, can be more prone to optical flaws if used beyond the original specs;
- The 6+6MP Fuji S5 Pro resolution is not on par with the 16MP Sony 5n, maybe the latter resolving power can highlight some micro-misalignments of the glass itself;
but anyway an interesting (although a bit disappointing) test.

Comment edited 14 minutes after posting
1 upvote
By Murmillo (6 months ago)

Could be your converter workflow or even testing technique/methodology that affected results vis-a vis resolving power. Eg: Vibration is an issue even when mirror lockup is used.
However it could be that your pro adapter (not present on the DSLR) is misaligned. Your next step is to determine if this is a design flaw or if you have a defective adapter.

1 upvote
By Shamael (7 months ago)

Imagine that some 50 years ago. Anyone who made photos in that time just looked at his prints. The good cameras where known, the good lenses too, and there was a calssification in pricing, so that in every price range, you could find an acceptable gear. Today, when I read all those specialized forums and sites, whooow, bokeh, one stop better at 6500 iso, not good over 1600 iso, makes noise at 12500, and so on. Why do all of you not waste your time with going around and taking pictures. Any of today's correct digitals do better shots than most of our better old boxes. I have made my first shot with a Kodak Instamatic in 1964. I stepped then fast over to a Pentax, and later to Nikon FE, and after that to Mamyia C and Hassy. Now, lets look a Hassy today, 40000$, limited to 800 Iso, 400 for acceptable quality. And what do we do? Benchmarking, DXO, comparing, criticizing, claiming, and so on. You wanna compare, take your pictures, there is lots to see and talk about in there, I presume.

By Murmillo (6 months ago)

Your utter mixing apples with sour grapes (not just citrus fruit); or as they say in some languages your mixing of vegetables & flatulence noise, does not make you post any more sensical in its repeating of populistic clichés that still are horribly wrong,
neither does the number of likes make your post any better ;-)
(IOW: you have no idea what you're talking about)

By AngryCorgi (7 months ago)

Meh...I have seen very, very few complaints on DPR of people getting bad results from adapters, and I swear the pickiest people on Earth use this website. In practice, its just not a real issue.

By Paullubbock (7 months ago)

Exactly, if someone is greatly bothered by the very possibility of losing a microscopic degree of performance on their lenses with an adapter, then just don't use them. The rest of us will enjoy them.

By netudiant (7 months ago)

Perhaps because Roger's analysis is more exacting than the DP Review audience?
Does that mean that we are past the indifference point for digital image quality?

By Jogger (7 months ago)

i bought a $20 Nikon to NEX adapter off ebay and have been really happy with it. Will my photos be better with a $500 adapter?

1 upvote
By kodachromeguy (6 months ago)

I think the point is that inexpensive adapters may have wide manufacturing tolerances. You may have a real good example, but another one of the same brand may be worse. Presumably the expensive ones have tighter tolerance.

By Murmillo (6 months ago)

You've got it upside-down, reversed & trodden upon:
One of the major values of this type of testing is to compare results from differently prices adapters, so prospective buyers might thus…………get the idea? ;-)

1 upvote
By Paullubbock (7 months ago)

There is no such thing as a perfect lens, or adapter. This is the equivalent of a pixel peeper review of lens performance. Taken at minutia, it is just AWFUL, but taken as a whole or with a grain of salt, it will never limit your creativity. An adapter is a fantastic tool to expand your creative potential and the only thing it can introduce is perhaps a slight but negligible miss alignment or difference in lens distance that may or may not show off the quality of performance of the edges of the lens. So be it. Other then that, there is no extra glass in between your lens and the sensor to change anything else.

By yabokkie (7 months ago)

focusing and zooming will do more harm, that we have to live with.

By ProfHankD (7 months ago)

As I just explained in the NEX forum, there are 3 types of glassless adapter issues. Slightly missing infinity (but still in DoF) will cause slight IQ loss at infinity (floating element designs might suffer worse), but is easy to detect. Decentering of the lens as a whole, as opposed to lens elements relative to each other, is generally negligible -- (camera mounts often are not quite centered on the sensor). A tiny tilt will be magnified by the Scheimpflug principle, and can cause lousy performance on a precisely aligned optical bed, but will be compensated if you base alignment of a flat target on focus rather measurements (effectively aligning with a tilted plane) -- or hidden if your scene isn't flat.

I think Cicala is measuring tilt that is usually not photographically relevant.

By photosen (7 months ago)

Cool except for this bit, which I'll change to another industry: Windows PCs couldn’t all stink or people wouldn’t use them. Oh really?

1 upvote
Eric Hensel
By Eric Hensel (7 months ago)

Ignore the troll...

By BRPWS (7 months ago)

So was the test flawed?

Roger Cicala said:


We’re trying to figure that out a bit. It’s very consistent with several different mounts.

"One thing that has recently been suggested is that the left side of the machine is against a wall, the right side exposed to a very bright room – and the front of the lens is open to that ambient light during testing. I’m going to feel really stupid if turning out the lights makes a difference after running hundreds of tests"


Comment edited 50 seconds after posting
1 upvote
By Plastek (7 months ago)

Good test. Guys at made similar tests (only never published the results) and came to a similar conclusions - adapters are completely random and getting an acceptable one - perfectly centred with accurate flange distance - is impossible. That's why they test lenses on native bodies instead of comparing all of them in a single body and single sensor (what would allow cross-system comparisons).

So much for all these people thinking that shooting mirrorless with adapters is a valid way for photography.

1 upvote
Andy Westlake
By Andy Westlake (7 months ago)

There's a world of difference between lens testing, and real-world shooting with adapted lenses on mirrorless cameras. For one you need absolutely perfect alignment, for the other, you don't.

It's probably worth bearing in mind that teleconverters and extension tubes will necessarily show the same effect, but that doesn't stop anyone using them. In fact if you look really closely, you'll probably find that all of your lenses appear to point in slightly different directions - and it simply doesn't matter.

Comment edited 32 seconds after posting
By Plastek (7 months ago)

What's the purpose of spending over 1000$ on a lens if not having it as good as it possibly gets?
If you think that "there's a world of difference between lens testing" than you can happily shoot with any lens what so ever. I can't - I want a quality for the money spent. If it doesn't deliver - I sell it and buy something that's up to the standards.

"It's probably worth bearing in mind that teleconverters and extension tubes will necessarily show the same effect" - That's different. TCs benefit you by providing longer focal length without a need of taking a second lens into your backpack. That's a huge weight-saver for the price of IQ. Completely different situation.

"In fact if you look really closely, you'll probably find that all of your lenses appear to point in slightly different directions" - Again: If I buy a decentered lens I give it back. I understand that some people don't care and will shoot with whichever lens. I do care.

By joejack951 (7 months ago)

"If I buy a decentered lens I give it back."

How are you checking for decentering? Unless you are using equipment similar to Roger's you are very likely getting inconsistent and/or unreliable results. But so long as your real world shots agree with your test, you accept the lens, right?

Remember no lens (or body) is perfect so if you aren't seeing decentering you just aren't looking hard enough. Roger is looking pretty hard though so it's not surprising that he's finding issues.

By Plastek (7 months ago)

joejack951 - There's a difference between decentering that's acceptable within production standards (==insignificant) and a decentering because lens is broken or it uses adapter.

I'm not looking for perfection. I'm looking for acceptable quality where one both sides of picture are equally sharp on my camera. It's not really difficult to get if you don't use adapters - most of the lenses are fine. Older glass often got an issue with it - especially plastic lenses. So far I seen 3 lenses that got noticeable decentering without any adapters - 1 brand new Tamron I bought and 2 older lenses (that was to be expected).

Most of the people care about image quality - that's why they test glass on brick walls (lol) and do other stuff like that. That's why "bad copy" is a standard expression in photographers dictionary. IDK why suddenly with the growth of mirrorless people don't care about quality any more and can settle down with adapters. IMHO that's quite silly.

Comment edited 2 times, last edit 9 minutes after posting
By keeponkeepingon (7 months ago)

"That's different. TCs benefit you by providing longer focal length without a need of taking a second lens into your backpack. That's a huge weight-saver for the price of IQ. Completely different situation."

How's that different? Your OK with loosing IQ to save weight/get more zoom, while other folks may be OK with a (slight) IQ loss to use a legacy lens.

It's the same thing. Trading IQ for to do something you normally could not do with your camera. Just because you value zoom and don't value using legacy lenses does not make it totally different.

One nice thing about pentax is there are some great legacy lenses and you don't need an adapter as the current bodies work with lenses made eons ago.

By Plastek (7 months ago)

keeponkeepingon - you assume that there's no other way to get certain focal lengths. Meanwhile usually there is - either by buying native lenses, or simply - being a signal that a problem you have is in picking inappropriate system for your photography.

That's a general thing - if you are really FORCED into using adapters for your photography - you most likely picked a wrong system for your needs and instead of toying - you should consider switching (it's not like joining a system you signed a chirograph).

Meanwhile on dPreview over and over again I see people encouraging adapters as if it'd be god-scent of photography, an ultimate tool for everything. Completely forgetting that there's a price for that and the lenses used through adaptors are NOT as good as they are natively (not to mention the warranty issues).

Again - TCs are better because they save you weight. You don't save any weight if you take an additional lens AND an adapter on top of that.

Comment edited 2 times, last edit 8 minutes after posting
By Iskender (7 months ago)

"That's a general thing - if you are really FORCED into using adapters for your photography - you most likely picked a wrong system for your needs and instead of toying - you should consider switching"

The system you are talking about doesn't exist. Every system is a compromise, as evidence by Canon EF users using adapted lenses.

Most people realise that camera systems are indeed compromises, and act on it. I save money and get excellent results by using adapters - this means I don't have to ruin myself just because of what someone's test bench showed. Many others use adapters and look at the picture first. If it's a good picture, they'll keep the adapter and save a lot of money.

Pixel-peeping gets frustrating and expensive fast.

1 upvote
By AbrasiveReducer (7 months ago)

Because of what Andy says above, I don't pay much attention ton what Lloyd Chambers says. In fact, I don't pay at all for his advice although unlike Ken Rockwell, it seems Lloyd Chambers understands the hairs he is splitting.

Plastek has a point. I love wide angle lenses and the first thing I do is see if all 4 corners are the same. Not good or bad; just the same. I had two Sony RX 100--a camera that people love--and each had a soft corner. The tests at Imaging Resource showed the same, as does the DPR test. But people love the camera and think it's worth every penny.

Comment edited 5 minutes after posting
By RCicala (7 months ago)

I just want to throw my $0.02 in: I agree with everything in this series of comments. Like Iskender said, pixel-peeping gets frustrating and expensive, and like Andy said, it usually doesn't matter.

The point behind a lot of what I do is basically trying to say there are no absolutes in photography. When people say 'adapters don't have any adverse effect at all', they're wrong. When people say 'adapters always screw up the image', they're wrong, too.

The coolest thing about photography, to me, is balancing all the compromises so we get the best images. I guess I see my bit as pointing out where some of the compromises are so we can make educated decisions.


By takumar (6 months ago)

I guess all the people who shot with a view camera for centuries are all "non valid" photographers by your standards, because on these, 100% exact centering is never achieved (and most of the time not even desired). Hopefully you will show your images soon, they will put the rest of us to shame for their immense qualities will demonstrate "valid" photography.

Comment edited 33 seconds after posting
By jp1774 (7 months ago)

Great. Now I'll have to buy bodies instead of adapters!

1 upvote
By Infared (7 months ago)

Just what I thought. Thanks Roger!

KL Matt
By KL Matt (7 months ago)

This guy is a true master of content marketing. And I like his content!

Total comments: 48