[splash_fr]

Hi all,
Does anybody know the technical reason, why I can only use the central AF point when using MF lenses (A series)?
I.e. when switching a DAxxx to MF I can select any AF point, when using an A-series lens I can only use the central one... Is this only a software thing or a real hardware problem? I thought the AF sensors are in the camera body and should also work with old MF lenses?

rgds,
Gerd.

PS: I use a K-3II, but this should be the same on most recent bodies...

[mcgregni]

I don't know the technical reason, but it is specified that only the centre af point will work for focus indication with non-af lenses. There is no point to selecting any af point with manual lenses. The only technique you could use is focus with the centre confirmed then recompose. ...... At least you don't have to hold focus lock

It's the same for catch-in-focus working... Only the centre.

[pathdoc]

Philosophically I've got no objection to only being able to use centre-point - after all, this is what the cameras which used these MF lenses were limited to.

My guess is that that the optical path at the periphery and the centre is not necessarily the same, and this might potentially throw the AF algorithms off. The AF lens can tell the camera what its formula is and how to compensate; the MF lens can't, and the camera has been programmed to do the safe thing under those circumstances and default to the centre path as being the reliable one under all conditions.

It would be interesting to compare which lenses offer multi-point AF with the list of lenses that offer distortion correction in-camera.

[UncleVanya]

If I understand it correctly there is data about the lens needed to use the alternate points - I think only lenses than can identify themselves can be associated with this data. But I am not certain.

[stevebrot]

Does anybody know the technical reason, why I can only use the central AF point when using MF lenses (A series)?

I don't know of any technical reason. After all, focus confirmation uses only the most basic functionality of the AF system.

PS: I use a K-3II, but this should be the same on most recent bodies...

It has been that way on all Pentax dSLRs at least as far back as the K10D (2007) and I believe all of the *ist D series as well.


Steve

[stevebrot]

It would be interesting to compare which lenses offer multi-point AF with the list of lenses that offer distortion correction in-camera.

F, FA, DA, and D FA all are supported for multi-point manual focus.

FA 31, FA 43, FA 77, DA (all) and D FA (all) are supported for in-camera lens correction.*


Steve

* Depends on body and firmware version.

[splash_fr]

Ah, in-camera lens correction! That must be the reason!
Thanks,
Gerd.

[bygp]

Hi all,
Does anybody know the technical reason, why I can only use the central AF point when using MF lenses (A series)?
I.e. when switching a DAxxx to MF I can select any AF point, when using an A-series lens I can only use the central one... Is this only a software thing or a real hardware problem? I thought the AF sensors are in the camera body and should also work with old MF lenses?

rgds,
Gerd.

PS: I use a K-3II, but this should be the same on most recent bodies...

I guess it is intentional, I also can't do that on my K200D, so I use either center or auto modes, auto is nice in MF because it gives you transitional informations, mainly for close subjects, if you focus very far, auto or center don't make a big difference, except that auto seems to answer faster, and if it doesn't show the center point, it is often a point close to the center. My eye is still ok so I prefer to tune my focus with it than with AF indicators, I usualy get better focus this way.

[UncleVanya]

F, FA, DA, and D FA all are supported for multi-point manual focus.

FA 31, FA 43, FA 77, DA (all) and D FA (all) are supported for in-camera lens correction.*

Ah, in-camera lens correction! That must be the reason!
Thanks,
Gerd.

No, I think you missed an important point. Only the FA limiteds of the F and FA lenses are supported for in-camera corrections. However all F and FA lenses are supported for multiple-point manual focus (at least according to Steve). I can't verify this at the moment but I can verify that in AF mode multiple point focus works with my F lenses.

[stevebrot]

(at least according to Steve)

This is in the camera user manuals, though I confirmed with lenses from my own collection. Why the AF point is not selectable with non-AF lenses puzzles me, though it may have something to do with needing to know focal length for off-axis points.


Steve

[bygp]

This is in the camera user manuals, though I confirmed with lenses from my own collection. Why the AF point is not selectable with non-AF lenses puzzles me, though it may have something to do with needing to know focal length for off-axis points.


Steve

I don't think this is the main reason. Because the way the camera knows when a point is in focus has nothing to do with the lens specs, it's only incoming image processing (software algorithms can easily tell if a small image crop is in focus or not). That's why some subjects are harder to AF than others, a polished glass or a mirror for example if they are clean. Also some lenses in certain conditions like the kit 18-55 lens getting too soft/blurry at 18mm are hard to AF with. That's also why it is harder to AF in low-light, as there is natural noise, less gradients, less raw image informations (less photons^^). I guess it is more an intentionnal design from Pentax engineers, selecting a specific point in MF is almost useless as you just need to aim the point you want to focus and then move back to where you want your shot centered, this is so fast to do, and that's how I used to do with old SLR MF cameras which had only the center circle on focusing screen. But if they wanted to add this functionnality, or for example add multi-focus points (having 2, 3, 4 points in focus at the same time), they could do it, this is just embedded software (they call it firmware) to upgrade.

But perhaps you're also right in the way that some special informations (but I'm not sure it needs to just be focal length) given by modern lenses may help the camera software to be more accurate when telling a point is or not focused. How exactly, I don't know, but maybe some AF lenses that are not designed for DSLR could also produce inaccurate AF, I have this issue with my Sigma AF-K zoom which seems to be a SLR AF lens, even if this lens allow me to select a specific AF point in MF mode. (I think it is inaccurate because the focal length information given is for use with FF films, so center AF point could still be accurate while other points could be miscalculated, that's why I use it only in MF or in center AF and why I'll probably sell it^^)

Anyway, if I understood what you want to do (validate an off-centered MF focused point), for this purpose I set AF point on Auto and this does the job perfectly even if you have to often release shutter button to re-calculate while MF. I've not tested that yet, but maybe setting AF in auto and AF.C (instead of AF.S) could help you get exactly what you want with continuous focus points scanning. I'll try that.

My bad, tested agan with a MF prime, only center point can be used, just mixed with how I used my AF lens in MF mode set to Auto... Tried all options (AF.C, change AE-AF correlations), seems you just can't do it because they decided you would not.

But tests I've just done with my 24mm MF prime shows that this focus validation method in MF with MF lenses is totally inacurate (with my K200), I can get red center spot even when center point is out of focus, so I don't rely on it and will maybe just uncheck this option to focus only with my eyes... It seems that accurate AF validation or lens control can only happen with modern lenses designed for DSLR, and even then, using entry lenses in worst conditions will lead to missed focus.

[stevebrot]

But perhaps you're also right in the way that some special informations (but I'm not sure it needs to just be focal length) given by modern lenses may help the camera software to be more accurate when telling a point is or not focused.

One of the main differences between the KA and KAF mount is the 7th contact for digital data transmission. Known data from the first generation lenses included: 1) lens ID code, 2) focal length (zooms too), 3) estimate of focus distance (position of focus rail), 4) true selected aperture, 5) lens dimensions (flash compatibility), and 6) exposure program data (!). (Source: Dimitrov's K-mount Page) KAF2 added MTF data.


Steve

[bygp]

One of the main differences between the KA and KAF mount is the 7th contact for digital data transmission. Known data from the first generation lenses included: 1) lens ID code, 2) focal length (zooms too), 3) estimate of focus distance (position of focus rail), 4) true selected aperture, 5) lens dimensions (flash compatibility), and 6) exposure program data (!). (Source: Dimitrov's K-mount Page) KAF2 added MTF data.


Steve

Yeah, that's a lot of inputs which explains that not all AF lens, even with KAF mount (my Sigma has a KAF but not KAF2 and is not designed for DSLR but FF SLR) would give you a performing and accurate AF. I badly need now a replacement focusing screen, most of my lenses are MF. Although having AF on a compact camera is almost mandatory, on a SLR or DSLR I almost hate AF, will only maybe use it if I start to take sports shots with a real digital designed zoom. Thanks for these infos anyway, we never end learning^^

[splash_fr]

In wanted to use the off-center points for portraits. With a 135/2 using center point and recompose is not accurate enough.

Thanks for the interesting discussion,
gerd.

[bygp]

In wanted to use the off-center points for portraits. With a 135/2 using center point and recompose is not accurate enough.

Thanks for the interesting discussion,
gerd.

What I would do for this purpose (if you don't mind about the number of shots taken) is bracketing while MF focusing, you would be very unlucky not to have 2 or 3 perfect shots from 10 If your subject is not stressed by shutter noise...