[Marc Sabatella]

Is it worth the extra money for the Optibrite?Can you SEE the difference? Does it help the camera focus in low light?

Help *the camera* focus? The focus screen has no effect on AF whatsoever.

[Marc Sabatella]

I wonder if the effective f/2.8 maximum aperture of the focus screen is due to the light loss through the partially silvered mirror?

No, it's due to the construction of the screen itself - the microlenses in particular.

This is the best article I know on the phenomenon:

Manual Focus with AF DSLRs

[res3567]

Help *the camera* focus? The focus screen has no effect on AF whatsoever.

Hey Marc!

Thanks for your response.


I just wanted to see if there were any cures to the poor low light AF performance of my K10D.

I quess what you are saying is the AF sensor is totally idependent of the focussing screen and the problem lies there.

[NaClH2O]

Squier:

You might want to try some focusing exercises instead of getting a new focus screen. I found these exercises posted by Godfrey over at DPR. Yes they are boring, boring boring. But they work. As an ex professional musician I found practicing scales boring boring boring, however I never got really good on an instrument until I mastered the scales on it. Same with photography, a little bit (ok, more than a little bit but not much more) of practice makes a great deal of difference in your shots. Try the exercises posted in the link and at worse you've wasted a little time. But I think you will find them very helpful.
Re: Is a KatzEye required for those using manual focus?: Pentax SLR Talk Forum: Digital Photography Review

NaCl(how do you get good? ...practice)H2O

[Squier]

That link is a good read Marc - fascinating.

And NaC - definately worth a shot, cheers.

[OregonJim]

No, it's due to the construction of the screen itself - the microlenses in particular.

This is the best article I know on the phenomenon:

Manual Focus with AF DSLRs

Good article - so that means my Chinese screens (cut from K1000's) solve this problem as well - cool!

[stevebrot]

Is it worth the extra money for the Optibrite?Can you SEE the difference? Does it help the camera focus in low light?

Yes, you can see the difference, but only with lenses that have slower maximum apertures. I use the 18-55/3.5-5.6 kit lens and have a couple of manual focus lenses in the f/3.5-f/4 range that were just a little dim for my taste. As for aiding in dim light focus...I can't really say that it makes a difference.

Steve

[Marc Sabatella]

I just wanted to see if there were any cures to the poor low light AF performance of my K10D.

If you're using slow-focusing lenses, then replacing them with faster-focusing ones certainly helps. In general, wide maximum aperture means more light for the AF sensors; contrastier lenses gives the AF system clearer targets to work with; smaller/lighter lenses have less bulk to move than larger/heavier lenses; and lenses with a short focus ring throw can go from minimum focus distance to infinity much faster, which reduces hunting time. And better-made lenses probably have less resistance than more cheaply-made ones.

The DA40 ends up getting good marks on all counts; you'll probably never beat it. But if you're trying to use the kit lens at 55mm & f/5.6 in low light, good luck with that...

I quess what you are saying is the AF sensor is totally idependent of the focussing screen

Correct.

[res3567]

If you're using slow-focusing lenses, then replacing them with faster-focusing ones certainly helps. In general, wide maximum aperture means more light for the AF sensors; contrastier lenses gives the AF system clearer targets to work with; smaller/lighter lenses have less bulk to move than larger/heavier lenses; and lenses with a short focus ring throw can go from minimum focus distance to infinity much faster, which reduces hunting time. And better-made lenses probably have less resistance than more cheaply-made ones.

The DA40 ends up getting good marks on all counts; you'll probably never beat it. But if you're trying to use the kit lens at 55mm & f/5.6 in low light, good luck with that...



Correct.

Thanks Marc!

I have a 16-45 on the way.

I have heard good reviews about this lens.

I hope to get a good copy and if it matches some of your criteria for good AF performance in loe light, I will be quite happy.

[Marc Sabatella]

Well, the 16-45 isn't particularly small, nor is it particularly fast or especially well built, but it is sharp & contrasty wide open from all I hear. I have no idea how long the focus throw is.

Overall, by these criteria, I'd expect it to be around average for a zoom; probably nowhere as fast as the average prime. I'm not sure what lenses you're comparing against, but I doubt you'll see a huge improvement in terms of focus speed.

[res3567]

Well, the 16-45 isn't particularly small, nor is it particularly fast or especially well built, but it is sharp & contrasty wide open from all I hear. I have no idea how long the focus throw is.

Overall, by these criteria, I'd expect it to be around average for a zoom; probably nowhere as fast as the average prime. I'm not sure what lenses you're comparing against, but I doubt you'll see a huge improvement in terms of focus speed.

The Sigma 18-50 f/2.8 EX, Pentax 16-50 f/2.8 and the Tamron 17-50 f/2.8

[Class A]

(I don't know what to make of this. I knew that AF lenses could have variations in FF/BF, but I thought with MF that WYSIWYG.

According to my understanding, your eye has the same trouble as an AF sensor. Both focusing systems have prisms in the optical path and somehow different lenses interact differently with these.

There is a logical explanation as to why an unaided AF sensor which does not take colour temperature into consideration (i.e., all pre-K-7 DSLR AF systems) will front-focus under Tungsten lighting: The prisms will direct Tungsten light to a different position compared to daylight, resulting in a phase shift and hence focus shift.

As to why focusing also seems to be dependent on lens characteristics is one of the "three great lens mysteries" to me. One potential explanation is lens-dependent CA (shouldn't affect manual focusing though, should it?) another, which I don't believe in, says that sensing works fine but lens control is different for different lenses, i.e., small tolerances in the drive/gear make for imprecise control (should not affect manual focusing significantly, to very very small amounts, if at all, because in manual focusing there is definitely a closed feedback loop).

[dave9t5]

According to my understanding, your eye has the same trouble as an AF sensor. Both focusing systems have prisms in the optical path and somehow different lenses interact differently with these.

There is a logical explanation as to why an unaided AF sensor which does not take colour temperature into consideration (i.e., all pre-K-7 DSLR AF systems) will front-focus under Tungsten lighting: The prisms will direct Tungsten light to a different position compared to daylight, resulting in a phase shift and hence focus shift.

As to why focusing also seems to be dependent on lens characteristics is one of the "three great lens mysteries" to me. One potential explanation is lens-dependent CA (shouldn't affect manual focusing though, should it?) another, which I don't believe in, says that sensing works fine but lens control is different for different lenses, i.e., small tolerances in the drive/gear make for imprecise control (should not affect manual focusing significantly, to very very small amounts, if at all, because in manual focusing there is definitely a closed feedback loop).

In retrospect, I think that the variation in MF from lens-to-lens (or zoom position on a given lens) can be attributed to these factors:

- Focus direction is really important. In my focusing screen calibration process, this was a largely uncontrolled variable. But it stands to reason that as one is viewing the focusing screen, they will consistently be slightly off from the focusing mark if they approach it from one direction only. The only way to ensure accurate results is purposely over-shoot back-and-forth in each trial, much the way that an AF makes it's final micro-corrections.

- Lens brightness is still important even with a split-image screen. That is, the split-screen is only as accurate as the DOF allowed by the lens. For example, at the slow end of a kit lens how can the split-screen focus more accurately than the DOF of f/5.6? I don't think it can. The split-screen still views the world through f/5.6. Therefore, the actual focus point could be off quite a bit from the apparent focus point on the split-screen .

- Conversely, with a ultra bright lens with ultra shallow DOF such as 85mm f/1.4, the DOF is so shallow that the actual focus point is so critical that even a split-screen does not allow accurate perception.

[OregonJim]

- Conversely, with a ultra bright lens with ultra shallow DOF such as 85mm f/1.4, the DOF is so shallow that the actual focus point is so critical that even a split-screen does not allow accurate perception.

The apparent speed of a lens when viewed through the viewfinder has a hard limit at the bright end. With the stock screen on a Pentax DSLR, I believe it is somewhere around f/2.8. This is due primarily to the design of the focus screen, but also to light losses through the AF beam splitter and pentaprism/pentamirror. So, you will never be able to view the shallow DOF of an f/1.4 lens through the viewfinder, which makes focusing in this range a challenge for both AF and MF.

[dave9t5]

The apparent speed of a lens when viewed through the viewfinder has a hard limit at the bright end. With the stock screen on a Pentax DSLR, I believe it is somewhere around f/2.8. This is due primarily to the design of the focus screen, but also to light losses through the AF beam splitter and pentaprism/pentamirror. So, you will never be able to view the shallow DOF of an f/1.4 lens through the viewfinder, which makes focusing in this range a challenge for both AF and MF.

Explicitly, we were discussing calibration of a (aftermarket) split-image screen. There is no stock screen of this type for Pentax dSLR AF body.

I'm not sure there is a correlation between viewfinder brightness and DOF resolution. Light losses would not affect DOF anymore than a ND or PL filter would?

The brightness of the stock screens may be limited in brightness to f/2.8 (?), but I think the stock matte screens are actually limited in DOF resolution to f/4 (?) due to the micro-prisms. Maybe I am getting those f/numbers mixed up.

I don't know what the DOF resolution on the prisms in the split-image screens might be. I assumed it should be limited by the lens used, but would be interested to know if it is not.