Lens aberration impacts TechArt LM-EA3’s auto focus accuracy, part 1


It has been a year and half since I use TechArt LM-EA3 adapter, which adds Auto Focus function to old manual lenses. It was amazing news for classic (=old manual) lens users. I put pre-order and waited couple of months for the adapter. Then I wrote several reviews from unboxing to real world function test and firmware update.

I thought there is nothing left for this adapter to review, but unexpectedly I discovered that the accuracy of AF (auto focus) is heavily depended on the spherical aberration of the lens. That’s a flesh new finding, hence another review is here after 1.5 years using the adapter extensively…

What it does mean in short, if you use large diameter lenses (for example, 55mm f1.2 or 135mm f1.8), the auto focus will not be consistent, either front-focus or back-focus slightly sometimes not even noticeable easily. When I found this symptom, I suspect in the beginning the adapter’s malfunction but as soon as I stop down one or two click, it start to deliver pin-point accurate focusing, so it couldn’t be the adapter issue.

Even so, TechArt LM-EA3 is an amazing piece of modern technology which converts legacy lens to auto-focus one for digital camera. But also it does not perform the same as the latest AF lenses. Therefore it is important to understand the limitation as well as benefit to use the adapter appropriately and within boundary condition(s).

To continue this story, I need to explain in short what the spherical aberration of the lens is, how the Sony’s phase detection AF works, and how to test the focus accuracy.

3 examples of large aperture standard SLR lenses I use,  (top-left) Nikkor 55mm f1.2  (top-right) Canon FD 55mm f1.2  (bottom-center) Canon nFD 50mm f1.2


What is lens aberration?

Lens aberration is the phenomenon that the light coming into the lens does not focus to single point (or more precisely focal plane), instead scattered before and after of the focal point. Because of this, even if the subject is a single point, it appears as a blurred circle on the film or sensor .


 (Left) Lens with spherical aberration     (right) Lens without spherical aberration

In case of ideal lens without spherical aberration (right diagram), the light rays are gathered at one point. But actual lens, because of various optical characteristics, does not gather rays at one point (left diagram) thus image blurred and degraded. Modern lenses have dramatically improved performance, so not much worry about lens aberrations unless they are very low-cost lenses. However, it was big concern on older manual lenses when computer design and manufacturing technology was not available. Especially, old SLR lens tends to have a large aperture without caring about aberration on wide open because the brightness of the finder is directly related to the aperture of the lens. In film days, users welcome this feature to make their SLR screen bright for accurate focusing and accept stop down at least 2-3 steps when actual shoot. 

The size of aperture and lens aberration are inversely proportional

It is casually called lens aberration, but it can be divided further, spherical aberration, chromatic aberration, astigmatism, field curvature, coma, distortion … etc., depending on the cause and the result.

In general, lens aberration reduced by tightening the aperture (making iris smaller) until it reaches to diffraction limit. In case of ’35mm full frame‘, lens performance keep improving until reach to f8 or f11, and then start degrading due to diffraction around f16. (diffraction is the inherent physical property of light, not optical aberration)

The boundary condition of this general statement is the “full frame” because the diffraction phenomenon differs from sensor size (or film size). In the case of 645 or 66 format, it starts from f22 – f32, and in the case of 4X5, starts from f64 respectively. This is rough generalization but pretty much accurate for old lenses, but modern lenses which utilize aspherical element tend to stay out of this trend and show the best performance from f2.8 or f4.0 or so.

How to measure lens aberration?

Some manufacturers disclose MTF chart in public which is quantitative measurement of their lens performance, and 3rd party like DXO also measure and release various lenses’ performance for themselves, so lens comparison couldn’t be easier if you want to do so as long as they are new. However, even DXO does not perform old manual lens testing at all.

Individuals can do scientific lens tests similar to lens companies. One of the companies that provide such services is ‘Imatest.com’, which analyzes the aberration by taking test charts in the specified environment and conditions. It’s a kind of lens DNA test. But it is not free service, I wouldn’t think to spend money for it, but it is indeed possible if you want.



Aberration measurement of an old lens does not represent THE lens performance

The problem is that even if we measure one or two old lenses in the scientific and quantitative method, we still do not know or can not estimate others performance.

When new lens come to market, they have to pass the quality control standards set by the manufacturer, so it is possible to say how the lens will perform by measuring one or two samples. Even if there is variation unit by unit, at least they all passed a certain level (the quality control limit), therefore single unit measurement can represent the group of the lens performance.

On the other hand, old lenses discontinued decades have gone through different fate each individuals. Some are kept clean and cared well, some have been repaired multiple times due to malfunction or careless handling, and some are disassembled to be cleaned simply because owner is nervous about dust or haze collected over time. In summary, they were manufactured in a certain level of performance at the time of manufacturing, but now we do not know how they will perform at all. If we pick up excellent example, it performs excellent but cannot make conclusion the other lenses perform like that. Or if we encountered a poor performing old lens, we cannot judge that all others of the lens are in such a condition.


Old big guns, (left) Spiratone Tele-photo 135mm f1.8,   (right) Sigma XQ scalematic 135mm f18


Fatal action for lens, disassemble & cleaning

In a short-term perspective, dropping or drowning the lens is fatal to the lens life, but it does not happen too often. Even a careless person who drop lens bluntly, after two or three broken lenses, he or she will give more attention most likely.

In a long-term, the most detrimental thing to lens is ‘disassemble & cleaning’. Many manual lens manufacturer has gone to history, even the companies are still in operation now, for example Canon or Nikon, do not repair their lenses produced several decades ago. It is inevitable to rely on repair shops operated by individuals, and not expected to do the same measurement and adjustment as manufacturer did when the lens shipped from the factory. Some repair shops do not have basic optical equipment, such as a collimator, so adjustment has done only by the repairman’s hands on experience.

Of course, sometimes a master in bare hand can do better job than average repairman with optical tools, however, it is not possible every repair shop has a master and craftsmanship. There is always possibility that dismantled lens not to be restored as it was originally produced. And the worst thing is an ambitious amateur torn the lens apart and wrestles with them to put together as barely as a shape of the lens…


(Continued in Part 2)


글쓴이 Leo_KHIMME

사진, 사진기, 렌즈, 여행, 새로운 곳, 새로운 기술 & 신산업혁명
대만 타이페이 거주
Oldies but Goodies 오피넛 멤버