24/1.4G: Actual Infinity vs Infinity Hard Stop, Overshoots

em_dee_aitch

Senior Member
Messages
3,675
Reaction score
106
Location
Austin, TX and San Francisco, CA, US
I encourage anyone who actually owns and uses a 24/1.4G to learn how to recognize and exploit the following behavior that I'm about to describe here.

As premise for this thread, you need to know the difference between optical infinity an the infinity hard stop . Optical infinity is the spot on the focus ring at which a very distant subject, such as the moon, is at its best focus. The hard stop is the place at which the focus ring will turn no further. These two spots are often not the same. On some lenses (most likely high dollar cinema lenses) the optical infinity may be closely aligned with the hard stop. On other lenses, the focus ring will move well past actual infinity focus, such that the image becomes blurry again. As a user of a lens, you need to learn how to recognize, by simply looking at the focus indicator window on the lens, when the camera has focused in the zone that falls between sharply focused optical infinity and the hard stop that is where it is for other reasons .

There are good reasons for this extra mechanical play on some lenses, such as to accommodate heat expansion (as noted by Erwin Puts of some Leica designs), but only the designer of a given lens can tell you why it's there on a particular model. I can't speak with blanket knowledge of whether going past optical infinity and the infinity hard stop will result in a worse picture on all lenses, but in the case of the Nikon 24mm 1.4G it most certainly does. That being said, I'll get to the point.

The Nikon 24mm/1.4G is a lens on which the focus ring will keep turning well past a sharply focused optical infinity to a hard stop at which your image will be quite blurry. Therefore this is not a lens for which, during manual focus, you should ever just turn the ring until it stops, as doing so would result in bad picture (due to the design of the default focus screen in most Nikon cameras, it is extremely difficult to see this with the naked eye in the viewfinder). Rather, if you are forced to guess at infinity focus (and don't have time to go to Live View), you should turn the ring until it stops then turn it back by a learned amount.

I believe at this point that I have what qualifies as a properly functioning and optically excellent copy of the 24/1.4G, i.e. this is not a "broken" behavior, or does not appear to be based on my experience with 2 prior copies of this lens. That being said, its behaviors are not perfect. At close distances it cannot focus on as many different types of textures as the 14-24mm/2.8G, for example; and, in what I believe is a related behavior, it often misperceives focus on distant objects. This is easily observable by using the focus indicator window on the lens. For example, this picture shows the correct optical infinity for an object that is about a quarter mile away:

(Please excuse the cell phone quality pic of my camera.)



This picture shows a re-focus on the same object which resulted in an incorrect setting beyond actual focus and closer to the hard stop:



The actual infinity hard stop on my copy of the lens is a bit beyond the left edge of the infinity mark on the dial. Because this is a short-throw lens, the small movements of the ring make very large differences in focus as you approach optical infinity. So the focus pictured immediately above is quite a bit off, though it could get worse by going all the way to the stop.

An important aspect of "mastering" or "dealing with" this behavior is to learn when it's likely to do it, then use the focus window to periodically check the focus results as you shoot subjects that are likely to invoke this behavior. So far I have learned not to trust focus on any landscape that lacks ultra-high-contrast detail.

But, barring that Nikon improve this behavior with a firmware update, the real solution for slow-paced landscape shooting it to make increased use of Live View. I recommend setting the multi-controller center button to Zoom during Live View (this is function f1 on the D3s, for example). Because Live View seems to be immune to this particular error, working in Live View ensures an optimal result on difficult focus targets. Once that custom function is set, you can drop into Live View, zoom in quickly, set your focus manually, then either stay in Live View to take the pic or go back out to compose in the VF. Of course if you do the latter, be sure you either turn off AF when leaving Live View or configure your focus not to be on the shutter release.

My lens has an AF Fine Tune value of +4 on this particular camera body. To eliminate AF Fine Tune as possible source of this focus error, I turned off AF Fine Tune. The result is that the focus error is the same amount either way.

--
David Hill
http://www.bayareaweddingphotographer.com
San Francisco & San Jose, CA | Austin, TX
Wedding Photographer and Apparent Gearhead
 
AFAIK, every Nikon ED lens exhibits this behavior. The glass that comprises Nikon ED elements suffers far more thermal variance than normal glass. This is also true for other manufacturers, hence the reason for Canon's signature white lenses. the white paint reduces heat absorption, lessening the thermal expansion of the ED elements. Because of that variance, it is impossible to have a hard infinity stop like the one present on older, non-ED lenses.
 
AFAIK, every Nikon ED lens exhibits this behavior. The glass that comprises Nikon ED elements suffers far more thermal variance than normal glass. This is also true for other manufacturers, hence the reason for Canon's signature white lenses. the white paint reduces heat absorption, lessening the thermal expansion of the ED elements. Because of that variance, it is impossible to have a hard infinity stop like the one present on older, non-ED lenses.
Thanks for that point. I knew that about the Canon fluorite glass (white lenses), but I wasn't sure about Nikon ED. To be fair to Nikon, some sources say Nikon glass is less heat sensitive than Canon.

The reason I chose to bring up this point on this lens is simply that it's the only lens I have that does the overshoot behavior often enough to notice. As you suggest, being aware of this on all lenses would be good. Another thought: As you move between weather extremes or stand in the sun for an extended period, Live View provides an obvious way to monitor if and when the actual optical infinity changes positions relative to the indicator scale and hard stop.

--
David Hill
http://www.bayareaweddingphotographer.com
San Francisco & San Jose, CA | Austin, TX
Wedding Photographer and Apparent Gearhead
 
All my Nikkors do this to some extent. Actually, my 17-55 does it the most and is easily noticeable on stars.

As far as the 24/1.4 is concerned, I spent a few hours with a friends, and it was super accurate at close focal with high contrast. I set up a test scene using a knife with a small LED light pointing at the tip of the knife, with the blade pointing almost directly at me. I used the center focus point and switched to AF-C and half depressed the shutter on a D700. I moved so that I was at minimum focal distance to the tip of the bright knife tip and moved ever so slightly from focusing on the tip to focusing on the back wall about 20 feet away, and then moved back to the knife tip. The 24/1.4 nailed it quickly and every time. I then used room light alone, and it still got it just fine. I tried it again outside against the moon and it was just as good, but a tiny bit slower - longer to turn to infinity I suppose.
 
This is not really a new discovery.

All my Nikon lenses do this.!!!

--
Geoff
Gold Coast, Australia
http://www.alldigi.com
 
This is not really a new discovery.

All my Nikon lenses do this.!!!
I'm not saying this is new or a discovery, btw. I just wrote my post so that it can make sense for different knowledge levels.

When you say that all your lenses do this, do you mean that all of them overshoot infinity on a regular basis, or simply that they all have the extra play? That's what I'm interested in terms of others' experience.

I was certainly aware that others had the play, but I don't have any other lens which so often overshoots infinity. Thus the others have not merited attention on this front... Perhaps Nikon should enhance the fine tune settings to have a user-definable infinity limit as an additional setting. Of course you would have to remember to recheck it after a major temperature change, but it would prevent the issue and save time while shooting.

--
David Hill
http://www.bayareaweddingphotographer.com
San Francisco & San Jose, CA | Austin, TX
Wedding Photographer and Apparent Gearhead
 
Generally they all overshoot infinity. As most are aware, it is due to allowances for temperature changes. Generally they are in focus at infinity but not necessarily when beyond infinity, although the difference may be small with some lenses.

--
Geoff
Gold Coast, Australia
http://www.alldigi.com
 
Generally they all overshoot infinity.
Sorry to ask you to clarify, but we already agree they can physically overshoot infinity when manually put there, but are you saying that all of your Nikon lenses overshoot infinity in error when using auto focus?

--
David Hill
http://www.bayareaweddingphotographer.com
San Francisco & San Jose, CA | Austin, TX
Wedding Photographer and Apparent Gearhead
 
They auto focus at the infinity mark when focussed at distance. But if the autofocus is overridden by manually focussing to the hard stop they will not be in focus.
--
Geoff
Gold Coast, Australia
http://www.alldigi.com
 
They auto focus at the infinity mark when focussed at distance. But if the autofocus is overridden by manually focussing to the hard stop they will not be in focus.
I guess you're saying that your lenses don't go near the hard stop in error. Mine don't either (at least not often enough to notice), with exception of the 24/1.4, which does get near the hard stop under AF, without any manual assistance. On the later part of your statement we obviously agree, that manually focusing to the hard stop ends in a bad result.

--
David Hill
http://www.bayareaweddingphotographer.com
San Francisco & San Jose, CA | Austin, TX
Wedding Photographer and Apparent Gearhead
 
So what if the 24/1.4 focuses closer to the hard stop or past the infinity mark? It's just a symptom of the same thing. The infinity mark on the focus ring isn't going to be any more accurate than the hard stop. The same can really be said with all the other distance markings on an ED lens. Your best bet is to just ignore it. It's good for rough knowledge of focus distance, but it's not accurate for any precise work. Same can be said of any lens.
 
All autofocus lenses go past infinity. A phase-detection autofocus system has the ability to tell if the focus is too close, or too far. Thus, to autofocus on infinity, it needs to be able to go past it.

I don't think the heat expansion issue is very significant compared to this.

Most manual focus lenses stop at infinity.
 
All autofocus lenses go past infinity. A phase-detection autofocus system has the ability to tell if the focus is too close, or too far. Thus, to autofocus on infinity, it needs to be able to go past it.
I get that. The point is whether or not a particular lens does it too oftenWhile the extra mechanical play must be there to facilitate the mechanics, I think the camera ought to be able to use software to limit the overshoot via AF to be a very rare occurrence (because there is no photographic reason to go into that extra zone of mechanical play, other than once per time interval to register the actual infinity point, after which the extra travel could be soft limited out), and my observation with 24/1.4 is that it's not a rare occurrence. I'm wondering why this is not controlled better in software as I think some of the other lenses are. Of course exactly how the AF system "thinks" is proprietary info, and Nikon isn't saying. It's known that the camera bodies do have some degree of lens-specific database, because they use that in vignetting correction, etc. How much is this used for AF correction? Unknown. I wish it were more.
Most manual focus lenses stop at infinity.
That's not really true. Look at a Leica 180mm APO, for example. At temp of 70 F, it goes way beyond infinity. So do many others. Some of the Zeiss ZF lenses have a hard stop that is beyond actual infinity, like the 35/2.

--
David Hill
http://www.bayareaweddingphotographer.com
San Francisco & San Jose, CA | Austin, TX
Wedding Photographer and Apparent Gearhead
 
So what if the 24/1.4 focuses closer to the hard stop or past the infinity mark? It's just a symptom of the same thing. The infinity mark on the focus ring isn't going to be any more accurate than the hard stop. The same can really be said with all the other distance markings on an ED lens. Your best bet is to just ignore it. It's good for rough knowledge of focus distance, but it's not accurate for any precise work.
Again, I'm not advocating this method for "precise work." I'm recommending it for troubleshooting and workaround purposes. The "so what" of learning to spot this is by occasionally checking it you can spot major focus errors that are not visible on a standard focusing screen in viewfinder. And even if you did have an enhanced focusing screen, the precision of the screen comes into play, as a screen is subject to placement/calibration errors in the same sense that the phase detect AF array under the mirror box is.

--
David Hill
http://www.bayareaweddingphotographer.com
San Francisco & San Jose, CA | Austin, TX
Wedding Photographer and Apparent Gearhead
 
Thanks for sharing this information!!

I don't use Manual focus very often with this kind of lens, so I never knew this was something common to all ED or modern lenses. In fact, I was baffled why my results were off when using manual at infinity when I did try it some very few times.

Will keep a look out for this issue, and thanks for teaching me something.

--

Sincerely,

GlobalGuyUSA
 
I like your post. Good information to consider, and well written.

Thanks
 
They focus beyond infinity to make manual focus easier.

Most photographers claiming to be a "pro" realized long ago infinity is at the infinity position on the focus ring.

Whilst manual focus through the viewfinder requires reasonable eyesight, those who have reasonable eyesight should be able to judge precise focus at infinity with the 24mm f1.4
I have no problem doing it with the 24mm zooms at f2.8 on the D3/300 series

One thing that cannot be done is to judge depth of field at f1.4 through the viewfinder as the eyepiece limits DOF to no wider than f2.8 on a D300/s and no wider than f2.6 on a D3/x/s.

--
Leonard Shepherd

Practicing and thinking can do more for good photography than buying or consuming.
 
They focus beyond infinity to make manual focus easier.
That's totally wrong, Leonard. The reasons are for heat expansion/contraction and mechanical considerations during assembly, etc. If anything, manual focusing would be far easier if infinity were always at the exact hard stop, but in reality it is not.
Most photographers claiming to be a "pro" realized long ago infinity is at the infinity position on the focus ring.
Leonard, as usual you're the one here proving your ignorance. If you check five different AF Nikkors, you're likely to find five slightly different infinity positions relative to the marked position, though all will likely be near the center of the infinity mark. They should not be near the hard stop, relatively speaking; i.e. they should be much closer to the center of infinity mark than to hard stop. Also, the exact position of infinity will change with temp and probably with individual camera body, due to AF Fine Tune.
Whilst manual focus through the viewfinder requires reasonable eyesight, those who have reasonable eyesight should be able to judge precise focus at infinity with the 24mm f1.4
Again, that is patently false, because modern default focus screens don't reveal detail below the equivalent of f/2.0, or possibly 2.5. To see the detail of f/1.4, or closer to it, you would require a different focusing screen. I've got well corrected eyesite, and I have a hard time spotting in VF when this happens.
I have no problem doing it with the 24mm zooms at f2.8 on the D3/300 series

One thing that cannot be done is to judge depth of field at f1.4 through the viewfinder as the eyepiece limits DOF to no wider than f2.8 on a D300/s and no wider than f2.6 on a D3/x/s.
That totally contradicts what you just said above, and it's essentially what I just threw back at you.

--
David Hill
http://www.bayareaweddingphotographer.com
San Francisco & San Jose, CA | Austin, TX
Wedding Photographer and Apparent Gearhead
 
Two updates:

First, I confirmed that the 50/1.4 and 14-24 cannot be made to overshoot to the hard stop as does the 24/1.4, at least not my copies of those lenses on my D3s bodies.

Secondly, I spoke with Rick Houghton at Nikon, and he agreed that it would be unusual for an AF Nikkor to overshoot into the extra play range which he said was indeed there to accommodate extreme weather (though in saying this he did not exclude other possible functions). He said he didn't have an answer as to why the 24/1.4 is doing this, but he would look into it.

--
David Hill
http://www.bayareaweddingphotographer.com
San Francisco & San Jose, CA | Austin, TX
Wedding Photographer and Apparent Gearhead
 
What I don't get is, in the first post, how did re-focus result in incorrect focus when the first focusing on that quarter-of-a-mile-far object resulted in correct one?
 

Keyboard shortcuts

Back
Top