A900 AF micro adjust

[braeside]

I had a look at the AF micro adjust feature today and started with my Minolta 50mm/1.7 lens.

At closeup range at f/1.7 (around 3 feet) - I needed 0 adjust.
However at mid to long distances, I needed -12 for best results at f/1.7. The difference is clear even at smaller apertures at those longer distances. I always thought the f/1.7 was soft, but it was just OOF.

The Tamron 28-75mm appears to require +3, but I will need to carefully check at a few apertures and focal lengths (as well as distances).

It is apparent that you need to be careful when making these adjustments and not base them on just one distance.

I would appreciate observations of this by others.

edit: The Tamron 28-75 needed +3, not 0 as I originally thought. The 50mm/1.7 needs -12 on average to distance shots, but closeups 0.

[Greg Beetham]

David are you checking with actual photos?....(Just thought I'd mention that because I know the focus screen has been changed), and it might be worth double checking visually on a monitor.
Greg

[David Kilpatrick]

I don't yet fully understand how to use AF Micro Adjest best. I have been using the Sigma 12-24mm for the last week. It certainly focused badly without adjust, but I was unable to find an AF adjust setting which worked. The solution seemed to be setting the lens manually just under the official infinity mark (which it goes beyond) for most large interiors and street views, and I ended up using the lens with manual focus by scale - not through the finder. Setting the AF Micro adjust so the lenses ended up at the same setting as used on the manual scale still did not produce sharp pix, while the manual focus did. No explanation as the Sigma 12-24mm is not yet HSM and has no MF of its own, you must use the camera switching to get MF.

David

[braeside]

Now completed the AF adjust:

Lens Closeup Adj Distance Adj
Minolta 50mm/1.7 0 -12
Minolta 100mm/2.8 -11 -11
Minolta 80-200/2.8 HS-G 0 0
Sony 70-300G SSM -11 -11
Tamron 17-35/2.8-4 0 0
Tamron 28-75/2.8 +3 +3

Fortunately the only lens that seems to suffer from different results at close-up is the 50/1.7

All tests were done with real life objects, not test charts, in daylight, on a tripod with 2 sec timer and SSM off, using AF-S and centre point AF point only.

Reviewed on LCD display with the magnify function, allows the effects of poor adjustment to be clearly seen.

[braeside]

I really don't know how to interpret it, perhaps the sensor is slightly out of alignment with the AF sensors, but that does not account for the difference between individual lenses. I can see no pattern to it, some Minolta lenses needed no adjustments, some Tamrons also.
The end result is all that matters, so I am happy that the facility to make these adjustments is there.

[Bengeo]

David (K),

Do you think there might be a problem with micro adjust and Sigma lenses? Sony don't guarantee it for 3rd party lenses and Sigmas have often been the odd ones out in the past.

Andy

[David Kilpatrick]

The Sigma lens may 'pretend' to be another lens which would mean that as long as I do not have a lens with the same ID, the Sigma would work correctly. I do not seem to be needing adjustments with any of my lenses on my A900 but I know the 12-24mm does need something doing, and it needs more care than any other lens to get right, so it will get some attention. I think I can also see a slight hint of backfocus in a couple of the other lenses I am using but so far nothing which is consistent enough to need a permanent micro adjust.

Like David A, my 17-35mm appears correct, but my 70-300mm SSM either does not need any adjustment or has a slight back focus tendency on checking a wide range of files. It is very hard to tell what could be my error, what could be a calibration issue.

David

[caporip]

I don't yet fully understand how to use AF Micro Adjest best. I have been using the Sigma 12-24mm for the last week. It certainly focused badly without adjust, but I was unable to find an AF adjust setting which worked. The solution seemed to be setting the lens manually just under the official infinity mark (which it goes beyond) for most large interiors and street views, and I ended up using the lens with manual focus by scale - not through the finder. Setting the AF Micro adjust so the lenses ended up at the same setting as used on the manual scale still did not produce sharp pix, while the manual focus did. No explanation as the Sigma 12-24mm is not yet HSM and has no MF of its own, you must use the camera switching to get MF.

David

David

Was this using the lens indoors? I have the lens and it seems to be ok so far on all my external shots, without using the adjustment facility (I've not used it indoors yet) maybe you have a duff copy?

Richard

[David Kilpatrick]

If I use AF outdoors, it typically focuses well beyond infinity. The results look OK on casual inspection, but they are much sharper - incredibly sharp - if the scale if set back to just before infinity at 12mm (I rarely use this lens at anything except 12mm, that's why it is purchased!). However, for all the shots I have been taking last week, I have also been stopping down to f/11. I think it's essential to use smaller apertures with all ultra-wide zooms, just as it used to be with ultra-wide primes.

David

[braeside]

Today I tested the newly acquired Minolta 28-85mm lens and rechecked my Minolta 80-200/2.8 HS-G for AF more carefully as the previous measurement seemed at odds with all my other Minolta lenses.

Lens Closeup Adj Distance Adj
Minolta 50mm/1.7 0 -12
Minolta 100mm/2.8 -11 -11
*Minolta 80-200/2.8 HS-G -11 -11
Sony 70-300G SSM -11 -11
Tamron 17-35/2.8-4 -3 -3
Tamron 28-75/2.8 +3 +3
*Minolta 28-85/3.5-4.5 -11 -11

*= retested today

It is now making more sense, all Minolta & Sony lenses need about -11 on my A900 body that must be the sensor/AF sensor offset.
The Tamron lenses are different and appear to Front Focus relative to all the other lenses, why I have no idea, but at least they can be compensated in the A900.

[petervanl]

David,

In which light did you do your testing? Normal daylight, tungstun, fluorescent or ......?
Do you see any difference in focus accuracy when you compare focus accuracy in tungstun light and normal daylight?
My experience is that tungstun is the most critical for focus errors. (at least with the KonicaMinolta 7D and Sony A700).

Peter

[braeside]

Peter, I only calibrated in daylight, I rarely shoot in Tungsten or Compact Fluorescent (which is much more common here now). I know that in general there is an offset between the AF point in tungsten and Daylight with Minolta and Sony DSLRs, though I have not measured it yet with my A900.

[petervanl]

David,

If you do an AF accuracy test in tungstun light then please report the results, would set my mind at ease and can stop thinking about different brands.

I saw you have got a 28-85, how does it behave on the A900, especially in the corners?

Thx

Peter

[braeside]

Peter, I sold the 28-85mm. Didn't like it at all on the A900.
Old fashioned lens, rotating front element and no hood. Nicely built (like Beercan). MFD not great.
Rather odd that it is longest when zoomed out!