New Olympus 17mm 1.8

Started Mar 7, 2014 | Discussions thread
Anders W Forum Pro • Posts: 21,468
Re: Okay, I think I got it-

madmaxmedia wrote:

Anders W wrote:

In many earlier threads on the subject, I have pointed out that there are indeed conditions where the AF of the 20 will be slower than that of other lenses. One of them is when it hunts (although it is not more prone to hunting than other lenses, just slower when it does it). Another is when you focus between extremes, like from minimum focus distance (0.2 m) to infinity (but 0.75 to infinity as in my test is not a problem). A third situation for which the lens is not suited is AF-C (on Pany bodies it won't even be allowed to enter that mode) and a fourth is video (if you want to AF during a clip).

However, I am hard pressed to think of a situation where it is practically important to quickly AF from 0.2 m to infinity and as to hunting, the practical solution is simply not to allow the lens to hunt or hunt to completition. This is largely a matter of knowing your equipment, its limitations, and how to overcome them.

I think you point out right here pretty well why the 20mm gives less optimal AF experience than other lenses for many people, and why there have been various disagreements between you and some others.

It's easy to say, well just pick good AF targets and your lens won't hunt, or hey you should know that focusing from one extreme distance to another will slow down this lens so you should prepare accordingly.

But consider the type of candids one might be trying to capture with the 20mm (or other lenses.) You're in the backyard at sunset with the kids, the light is great and they are running around in the backyard and it's a beautiful scene. Great photo op- especially if you can be somewhat inconspicuous and not ask them to pose, etc. because you want to catch them in spontaneous moments and expressions.

In situations like this, focus hunting and racking focus from near/far and far/near are going to happen, due to unpredictability of subject movement, low light, only reasonable contrast in terms of catching detail on faces, etc. It's easy to end up racking focus when you try to focus on child, but he's moving and so you miss and the camera focuses on the distant background, etc. The end result is that there's a good chance you are going to miss more shots with the 20mm than other lenses, even if you're trying to be careful and use good contrast AF technique.

I see what you mean. But racking focus between so widely different focus distances that it would put the 20 at a clear disadvantage wouldn't normally happen in the scenario you describe. The test of mine that I linked to varies the focus distance between infinity and 0.75 m. Still no problem. You'll see the difference between the 20 and other lenses only if you go to extremely short distances (close to the minimum focusing distance of 0.2 m) and then all the way out to infinity again.

But the really close part of the range (say 0.2 to 0.5 m), which is short as measured in meters but very long in terms of "focus throw", would hardly be used in the scenario you describe. Several tests I have seen fall into the trap of testing speed across the entire range from infinity to 0.2 m (or vice versa) although fast AF between those extremes is of hardly any practical relevance.

When it comes to hunting, I haven't really encountered much of a problem in scenes of the kind you describe either. The 20 is a fast lens aperture-wise, which in AF terms means that it takes really low light before the body has difficulties locking focus on any reasonably contrasty target, so low light that the option to do much in the way of "action shooting" is gone anyway. Note that the test I did was at a light level corresponding to an exposure of about 1/1.7 and 1/15 at ISO 1600.

The main practical consideration with regard to hunting is to find a reasonably contrasty target at the right distance and not to permit the lens to hunt to completion if it nevertheless starts but quickly let go of the shutter button, readjust the target, and try again.

Furthermore, the appropriate strategy in many cases is to prefocus. Being used to MF, where there is no other option if you want to capture the right moment, I nearly always do that, and find that it is nearly always possible, especially with a fairly short FL like 20 and with the DoF that MFT affords. If you wait to AF until you actually see the right moment in the EVF or on the LCD, it is frequently too late because the moment will be gone by the time the shutter opens. This is true not only with the 20 but with other lenses too, so I don't treat the 20 differently in this regard.

Finally, the main reason why the 20 is slower than other lenses when it refocuses across the entire focus range (from close to minimum focus to infinity) or when it hunts is that it is only then that the lens AF mechanism (which has lower max speed with the 20 than with other lenses) becomes the main bottleneck. Under normal circumstances, the main bottleneck is body processing, which is pretty much the same no matter which lens we are talking about (at least as long as max aperture is roughly the same).

So I don't disagree at all with your testing methodology and results. When there is no hunting and you're not racking focus over distance, the 20mm is actually pretty fast! But as you point out, there are factors that lead to a slower overall AF experience with the 20mm lens, than others. And these factors may come into play more often than not for some people, or in certain situations.

I understand all of the above. The main reason, however, that tests are insufficient is that there are too many potentially different situations for tests to cover them all well enough. Consequently, I too listen to the experiences of others. Sometimes, however, I find reason to question that reports about such experiences are valid descriptions of the objective facts. As I hope you realize, the step from subjective experiences to objective facts is sometimes a very long and complicated one.

The 20/1.7 and its AF behavior is a case in point here. I have seen quite a few forum members recognize, after discussing the matter in detail, that perceptual complications play a part in the impression that the AF of the 20 is slow. One of these is that you hear the lens AF (in sufficiently quite surroundings), which is likely to strengthen the impression of duration. Another is that the slow hunting behavior is what is most likely to make a lasting impression.

I agree with all these points too.

Actually, knowing about the good and bad aspects of the 20mm AF performance, will probably help try to maximize AF performance. And that may lead to a subjectively better experience, and opinion of the lens.

I would think so yes. My point is not at all that the AF behavior of the 20 is ideal. Of course, newer lenses with internal focus are better. Rather my point is that the shortcomings of the AF mechanism of the 20 are in many cases of minor importance, especially if you are aware of the shortcomings and try to adjust accordingly.

 Anders W's gear list:Anders W's gear list
Panasonic Lumix DMC-G1 Olympus OM-D E-M5 Olympus E-M1 Panasonic Lumix G Vario 14-45mm F3.5-5.6 ASPH OIS Panasonic Lumix G Vario 7-14mm F4 ASPH +28 more
Post (hide subjects) Posted by
(unknown member)
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark MMy threads
Color scheme? Blue / Yellow