Specifically for the discussion of the A-mount DSLR range
Forum rules
No more than three images or three external links allowed in any post or reply. Please trim quotations and do not include images in quotes unless essential.
I realise this topic has probably been done to death but still interested in your thoughts.
My A700 seems to exhibit consistent back focus using particularly fast lenses ie my SAL50/1.4 and shiny new Tamron 60/2. It's not really noticable on my slower lenses, even the constant f/2.8s (KM28-75, Sigma 70-200, Min 200). The "problem" lenses appear to be perfectly fine on my KM5D. Is this typical of a general focus issue, or is there more to it?
The local repair centre is halfway across the country (OK, not a big country ) and they would want lenses to test against after the adjustment. I'd definitely send the 50/1.4 but not so sure on the Tamron, given it took over two months for it to arrive!
Just do the three screws in the camera base. Use an L-shape hex key, use paper to record the position and turn each screw the same amount exactly. Try one quarter turn clockwise for each screw. Test.
Just make a precise guide so you can return to the original positions perfectly if you want to. You will find it works, and has no adverse effects, and you can almost certainly fix the problem for most of your lenses.
I'm lucky that only 1 of my lenses seems to have focus issues. I sort of imagined that if I adjusted the screws for this one that it would adversely affect the other lenses. Am I wrong? It's an old Tokina 28-70 f2.8 ATX.
Mike
All my Sony SLT gear gone. Still got my RX100 though.
I still seem to be getting inconsistent results with the Tamron, I know it's supposed to be a macro lens but surely it should still focus at distances in excess of 5 metres or so. The SAL50/1.4 seems much improved so far.
Do not calibrate using the 50mm f/1.4 unless youi can stop it down to f/4 - the aperture related focus shift is too great. Calibrate with the Tamron at 10X focal length = 600mm or two feet.
David Kilpatrick wrote:Do not calibrate using the 50mm f/1.4 unless you can stop it down to f/4 - the aperture related focus shift is too great...
I was trying to calibrate my A700 2-3 times with the 50/1.4 lens but it's really hard to get it right, if is good close then mises at some distance and reverse.
This is my only prime, i have and KM 17-35D, so which is more suitable for focus adjustments and what aperture and distance to try in the test?
The 17-35mm will only activate the f/2.8 sensor when set to 17mm. The problem with any wide-angle is that the range for focus confirmation is very wide. As a simple test, set up a contrasty target, a good focusing subject at a medium distance like 2-3m. Set the 17-35mm to 17mm, and switch to manual focus and centre focus point. Now watch the focus confirmation light, and carefully adjust the focus both ways until the light goes out/comes on in each direction. You will find that there is a certain amount of movement of the focus ring, during which the light remains lit. This is the depth of field of the sensor, which may not be as critical as you requirements.
Used on autofocus, the lens can end up set anywhere within this depth of field. Now try it at 20mm, which will switch off the f/2.8 AF sensor and revert to the f/7.1 sensor. In theory, the range of 'focus OK' confirmation will increase to about double the 17mm finding.
All AF cameras when made are calibrated to a distance of about 2 metres, using a 50mm lens. So in theory you might want to adjust A700 focus for exactly this, ideally using a 50mm f/1.7 because that's what they used at the factory. You then get most chance of the adjustment being accurate.
But if you regularly shoot landscapes or use a specific lens at close range a lot and get errors, you might want to calibrate for infinity with a wide angle, or 1.2m with a 70-300mm at 300mm. A camera calibrated that way may then not agree with some other situations. The 50mm at 2m is an average. For example, my 100-300mm APO D consistently back focused on all small animal shots taken at 300mm and closest focus on the A700, but it was fine for distant views.
Generally the most critical calibration will be with shorter focal lengths and closer to infinity. Errors in macro and tele focusing are much less dependent on the body calibration.
So - try the wide-angle test, you will then realise that actual AF is not a fixed point, but a range of 'OK' and that focus will vary depending on which direction the lens has focused from, and other conditions including the colour of the target. Try testing focus on a red target, and a blue target. You will find a difference with most lenses. This is why cornflake packet or drinks can tests are not reliable, often the subjects picked have strong primary colours.
Thanks, i see that the AF works in some acceptable range and sometimes makes a mistakes.
I have and Minolta 100-400 APO but this is my last problem right now, from my previous test saw that if my 17-35D focuses perfect at 17mm then mises at the long end so the things are very complicated.
So I'm trying to find some a settlement by compromise for my three lenses. I will try with 50/1.4 at 2m distance next time.
David Kilpatrick wrote:The 17-35mm will only activate the f/2.8 sensor when set to 17mm. The problem with any wide-angle is that the range for focus confirmation is very wide. As a simple test, set up a contrasty target, a good focusing subject at a medium distance like 2-3m. Set the 17-35mm to 17mm, and switch to manual focus and centre focus point. Now watch the focus confirmation light, and carefully adjust the focus both ways until the light goes out/comes on in each direction. You will find that there is a certain amount of movement of the focus ring, during which the light remains lit. This is the depth of field of the sensor, which may not be as critical as you requirements.
The worst DSLR I have had in this regard was the KM 7D. At any focal length of around 35mm or less it would almost always focus to infinity (or past infinity if the lens allowed) even for subjects that were just a few feet away. It would do this with Minolta lenses such as the 28-135mm, 35-105mm, and 35-70mm. It also would do it with a Sigma 18-125mm. The result were images that were clearly OOF and not within the DOF. The only wide lens that it would AF properly with was a Sigma 24mm f2.8. Did the KM 7D have an f2.8 AF sensor?
My A700 is better but it also is not all that reliable at short focal lengths, but the Sigma 24mm f2.8 is fine.
) and they would want lenses to test against after the adjustment. I'd definitely send the 50/1.4 but not so sure on the Tamron, given it took over two months for it to arrive!
