[Azergoth]

Hi everyone ;-)

I have a relatively stupid question...
I know that the aperture of a lens has an impact on autofocus (phase detection, not contrast with LV etc.) reliability and accuracy. The wider the aperture, the better. But, apparently, other factors seem to have an impact too. Here is what I observed:

• If I try to autofocus (using trap-focus, K10D) with my 135mm, the autofocus works until f/8.

• I have red that AF confirm even "works" with a f/10 lens (same camera, K10D) and with a f5.4 + 1.4 converter

• BUT, with a Samyang 500mm f/6.3, AF does NOT work. [EDIT: trap focus of AF-confirm]

So, I conclude that there is another factor affects autofocus... but which?

Thanks for passing by ;-)

---

Last edited by Azergoth; 05-19-2015 at 04:36 AM.

[Oldbayrunner]

If I try to autofocus (using trap-focus, K10D) with my 135mm, the autofocus works until f/8.

Trap focusing isn't designed for autofocusing, it's designed for auto shutter release when the focus is obtained manually.

BUT, with a Samyang 500mm f/6.3, AF does NOT work.

Perhaps this is because this lens is not an AF lens, It's manual only.

[Azergoth]

Trap focusing isn't designed for autofocusing, it's designed for manual use.

--> it's the same, it uses the phase detection... just not performing changes on the lens ;-)

Perhaps this is because this lens is not an AF lens, It's manual only.

Yeah, I know I meant focus trap or AF-confirm ;-)

(i'll edit my post)

[Oldbayrunner]

--> it's the same, it uses the phase detection... just not performing changes on the lens ;-)
(i'll edit my post)

So, are you really going to try to convince us one is auto focusing when using trap focus? Anytime I have used it I have had to manually rotate the focus ring to obtain shutter release. I think your intention is different than what is initially written.

PS. It has worked at every Aperture I have used as long as the lighting is adequate. This isn't something new it's been around since the Yashica 230 Af SLR first introduced it.

---

Last edited by Oldbayrunner; 05-19-2015 at 04:58 AM.

[Azergoth]

So, are you really going to try to convince us one is auto focusing when using trap focus? Anytime I have used it I have had to manually rotate the focus ring to obtain shutter release. I think your intention is different than what is initially written.

PS. It has worked at every Aperture I have used as long as the lighting is adequate. This isn't something new it's been around since the Yashica 230 Af SLR first introduced it.

I know, and I'm not trying to say AF and trap focus are the same. But it uses the same phase detection sensor.

And no, it does not work at any aperture, and int's not only a matter of amount of light. (there must be enough light, this is true!)

[Na Horuk]

Yep, there are many factors to AF.
Subject: - contrasty, well-lit, colourful helps. Some shiny or odd materials might make things more difficult. And subject movement can be problematic as well. Or if there are two good subjects both on the same AF point, but at different distances - the camera might lock onto one you don't want
Light conditions: - good, even light. An f8 lens will work if you have enough light, but even an f1.4 lens might fail if you have none. Also, the light would ideally be full spectrum, not just a narrow wavelength LED
Camera: - I think modern digital cameras have special AF algorithms for the lenses that they recognize, allowing for speedier focus. In this case, it would be limited to modern DA and DFA lenses (and maybe FA 31, 35, 43, 50, 77). This would mean that other lenses don't get this boost. Then there are the camera's AF detection points will be important. How many, where are they, how big are they, how sensitive, are they cross type, etc. And also whether the lens and camera are calibrated together or not. If a lens is miscalibrated, then you might have difficulty getting good focus
Lens: - the lens is between the subject/light and camera. The lens might have a slow aperture, or it might have really low contrast and other problems. Another problem is CiF with ultra wides, since the DoF is so great. I find my Samyang 14mm is not very good for CiF at all. Mirror lenses might be poor for AF, as well (low contrast, high aperture, low resolution)

And this doesn't even take into account user error, like setting the CiF wrong, turning the focus ring too fast, mistaking handshake blur for wrong focus, and so on. Also, some lenses will not work with CiF because their mount doesn't short out the AF contacts. Some Takumar and Helios 44-2 have a narrow mount, which doesn't cover the contacts. So you can only get CiF if you add a flange or buy a flanged adapter. Some mounts are plastic or covered with non-conductive paint. CiF also cannot work with lenses that are always in AF mode (like DA 35mm or DA L 18-55mm), because CiF needs lens to be MF and camera AF.

That's all I can say. If you need further help, you would have to do some tests (2sec timer, in natural sunlight, using tripod and lens hood, with an obvious subject at appropriate distance) and upload a couple photos (resized or 100% crops, not full size 16MP photos; also they should contain exif data)

[Azergoth]

Yep, there are many factors to AF.
Subject: - contrasty, well-lit, colourful helps. Some shiny or odd materials might make things more difficult. And subject movement can be problematic as well. Or if there are two good subjects both on the same AF point, but at different distances - the camera might lock onto one you don't want
Light conditions: - good, even light. An f8 lens will work if you have enough light, but even an f1.4 lens might fail if you have none. Also, the light would ideally be full spectrum, not just a narrow wavelength LED
Camera: - I think modern digital cameras have special AF algorithms for the lenses that they recognize, allowing for speedier focus. In this case, it would be limited to modern DA and DFA lenses (and maybe FA 31, 35, 43, 50, 77). This would mean that other lenses don't get this boost. Then there are the camera's AF detection points will be important. How many, where are they, how big are they, how sensitive, are they cross type, etc. And also whether the lens and camera are calibrated together or not. If a lens is miscalibrated, then you might have difficulty getting good focus
Lens: - the lens is between the subject/light and camera. The lens might have a slow aperture, or it might have really low contrast and other problems. Another problem is CiF with ultra wides, since the DoF is so great. I find my Samyang 14mm is not very good for CiF at all. Mirror lenses might be poor for AF, as well (low contrast, high aperture, low resolution)

And this doesn't even take into account user error, like setting the CiF wrong, turning the focus ring too fast, mistaking handshake blur for wrong focus, and so on. Also, some lenses will not work with CiF because their mount doesn't short out the AF contacts. Some Takumar and Helios 44-2 have a narrow mount, which doesn't cover the contacts. So you can only get CiF if you add a flange or buy a flanged adapter. Some mounts are plastic or covered with non-conductive paint. CiF also cannot work with lenses that are always in AF mode (like DA 35mm or DA L 18-55mm), because CiF needs lens to be MF and camera AF.

That's all I can say. If you need further help, you would have to do some tests (2sec timer, in natural sunlight, using tripod and lens hood, with an obvious subject at appropriate distance) and upload a couple photos (resized or 100% crops, not full size 16MP photos; also they should contain exif data)

Thanks for your great reply!

Concerning the Samyang 500mm, I have checked the contacts and they are shorted (the T2 ring was anodized but a scraped the coating. No I also have proper metering). Even under bright sunlight (ISO 100, f/6.3 and 1/300s, for example) and on contrasted subject, CiF does not work.

Concerning the camera, only the center focusing point works for CiF on the K10D.

The DOF of the 500mm is very narrow, even at f/6.3 that's why CiF would be great! On wider lenses, I often use the hyperfocal and don't even focus the lens 😆.

[Oldbayrunner]

CIF does not require contacts on a lens.

[Azergoth]

CIF does not require contacts on a lens.

They need to be shorted on te camera. So anodized or painted lens mounts need to be modified in order tu work...

[aleonx3]

Thanks for your great reply!

Concerning the Samyang 500mm, I have checked the contacts and they are shorted (the T2 ring was anodized but a scraped the coating. No I also have proper metering). Even under bright sunlight (ISO 100, f/6.3 and 1/300s, for example) and on contrasted subject, CiF does not work.

Concerning the camera, only the center focusing point works for CiF on the K10D.

The DOF of the 500mm is very narrow, even at f/6.3 that's why CiF would be great! On wider lenses, I often use the hyperfocal and don't even focus the lens 😆.

Do you mean you can never get a focus confirmation (green hexagon) on the viewfinder using this lens?

---------- Post added 05-19-2015 at 09:53 AM ----------

They need to be shorted on te camera. So anodized or painted lens mounts need to be modified in order tu work...

+1, what he said, the camera needs to detect a lens mounted on the camera.

[Azergoth]

Do you mean you can never get a focus confirmation (green hexagon) on the viewfinder using this lens?

---------- Post added 05-19-2015 at 09:53 AM ----------

That is correct. (I just checked)

[Oldbayrunner]

You don't need the superimposed hexagon for confirmation, only that the lens acquire focus for the shutter to activate.

[Azergoth]

You don't need the superimposed hexagon for confirmation, only that the lens acquire focus for the shutter to activate.

You mean for CiF?
I'f I'm not mistaken, the green hexagon means that the camera detects a correctly focused image (and thus lens). So, yes, I (or the camera) need that hexagon in order to activate the shutter (in CiF mode).

[stevebrot]

The minimum aperture restriction with the K10D is f/8 for the center points and somewhat wider for the others. This is a hard boundary and is based on the aperture size (actually, exit pupil diameter) and not the amount of light. There is a succinct explanation HERE. FWIW, f/5.4 with the 1.4x convertor is not f/10.

The fact that you have trouble at wider apertures may be related to subject contrast and/or the optical quality of the lens. Fuzzy over a wide focus range ==> unable to attain focus.


Steve

[Stone G.]

Concerning the Samyang 500mm, I have checked the contacts and they are shorted (the T2 ring was anodized but a scraped the coating. No I also have proper metering). Even under bright sunlight (ISO 100, f/6.3 and 1/300s, for example) and on contrasted subject, CiF does not work.

I would suspect the lens (: catadioptric mirror design). In contrast to refractor type lenses, light rays forming the image in a catadioptric design come from the perimeter of the lens, arrive at the sensor at a rather shallow angle and thus, precise focusing is VERY critical (if at all possible).

I have the same issue with my Tamron SP 350mm f/5.6 mirror lens: No Catch-in-Focus and no focus confirmation with any of my four DSLRs. Interstingly enough, when I put a 1.4X TC (f/8) or a 2X TC (f/11) CIF and/or focus confirmation work very well in most circumstances. I assume that this is due to the fact, that the beams arrive at the sesor at a steeper angle (but I don't know for sure, if that's the right explanation).

It may also simply be, that your (and my) lens needs some collimation. Again, mirror lenses are much more sensitive to proper collimation than are refractor lenses.