AFMA - shift over time?

[rfdesigner]

I dot-tuned all my lenses when I first got my 6D about 18 months ago.

Recently I've been getting a bit dispointed with shots feeling many were just a bit soft.

Today some deer appeared on the green opposite our house, I grabbed the camera and missed focus on almost every shot. So later I tried rechecking the dot-tune and found it needed to shoft from -5 to +3.. i.e. it had shifted about half the total acceptable focus range.

Now it's back to perfect focus almost every time.

How long do others go between recalibrating lenses?

 

[Viggo]

My experience is that even the most durable bodies, like the 1dx and 1dx2, is miscalibrated from even small bumps to the body (or lens).

So I have experienced what you have, but since getting the 1dx2 I have been extremely careful with it and the values remain the same.

I live where we have very cold winters and pretty warm summers, and the values can vary a bit with large swings in temperatures.

 

[Don Haines]

It will shift over time, and it will shift at different temperatures.... particularly if you use it a lot and it gets bounced around a lot.

I usually AFMA my gear in the spring and fall, or if I drop something......

 

[Mt Spokane Photography]

The lenses do shift over time, usually not enough to bother with, but they do change. Bodies don't tend to shift much.

 

[[email protected]]

I found the dot-tune method nice and quick, but only as good as your camera's perception of acceptable focus, which is to say, not quite as good as going hard-core with a FOCAL target. You might see a +/- 5 error with dot tune - at least I did. I think with FOCAL I'm mostly +/- 2.

It might be that other autofocus systems or other brands have a stingier sense of a proper plane of focus, and the dot-tune method might be more useful. On my 7D2 and 5D3 (last bodies I used with it), it was a bit squishy.

Nowadays, I'm doing it by eye and trial and error, simply because it's too much of a hassle to roll out the computer, target, etc. That works great for Canon lenses, which are limited to in-body AFMA. The Sigma glass requires AFMA at different distances, though, so once you're suffering through that, you might as well make it precise with the more accurate method.

 

[Viggo]

I bought FoCal when it first came out and was always happy, but then newer versions gave me consistent front focus results, I tried everything, including multiple light sources, multiple versions and computers. I even bought the hard target directly from FoCal, always front focus.

I then tried dot-tune and found it extremely accurate, 100 times faster and repeatable. So for me, dot-tune works, FoCal doesn't (anymore).

True for both 1dx, 5d2, 6d and 1dx2, the ones I have tried.

 

[rfdesigner]

Thanks everyone, good to know it's just part of living with an AFMA DSLR.

I'm one of those that swears by rather than at dot-tune, I tried it as soon as I got my 6D and it's worked perfectly every time I've done it. I've now redone all lenses and they all needed to move plus 8~10 points... so I think something in the body moved a fraction.

living here in the UK we don't see the sort of temperature extremes one can get elsewhere, and I'm lucky that our house is pretty stable temperature wise. The gear is never left in the car to cook. I don't recall bumping the body but clearly something's shifted things..

anyway, all good now.

 

[[email protected]]

I bought FoCal when it first came out and was always happy, but then newer versions gave me consistent front focus results, I tried everything, including multiple light sources, multiple versions and computers. I even bought the hard target directly from FoCal, always front focus.

I then tried dot-tune and found it extremely accurate, 100 times faster and repeatable. So for me, dot-tune works, FoCal doesn't (anymore).

True for both 1dx, 5d2, 6d and 1dx2, the ones I have tried.

I hadn't heard of that experience. The inconsistency I've personally experienced with FOCAL has been when moving between *types* of light. Sunlight got a different result than a set of bright CFL bulbs. I don't understand how that would happen (at all), but I was able to repeat those results. Since then, I've tried when possible to use ambient sunlight, as that's what I shoot in mostly.

 

[Don Haines]

strictly speaking, light at different frequencies will focus at different spots..... this is what "chromatic aberration" is.... You MIGHT get different values if you AFMA under reddish light or blueish light.... The better corrected for chromatic aberration that your lens is, the less likely this is to be a problem, but the lesson is clear, AFMA your lenses with the light you shoot under.... If you are a studio photographer, AFMA under the studio light, if you are an outdoors photographer, go for sunlight...

Realistically, this won't matter for most people as we shoot under a wide range of conditions and lighting, but if you shoot consistently under one set of conditions, it is something to think about....

 

[[email protected]]

Don, thanks!! It's been a mystery to me, but this makes complete sense. The effect was significant even with lenses that are pretty well CA corrected. But the light temperature difference is enormous in the room that I find most convenient (painted peach!). There's a lesson learned.... Don't accede to my wife's pick of peach. They should sell paint by the kelvin

strictly speaking, light at different frequencies will focus at different spots..... this is what "chromatic aberration" is.... You MIGHT get different values if you AFMA under reddish light or blueish light.... The better corrected for chromatic aberration that your lens is, the less likely this is to be a problem, but the lesson is clear, AFMA your lenses with the light you shoot under.... If you are a studio photographer, AFMA under the studio light, if you are an outdoors photographer, go for sunlight...

Realistically, this won't matter for most people as we shoot under a wide range of conditions and lighting, but if you shoot consistently under one set of conditions, it is something to think about....

 

[neuroanatomist]

IIRC, FoCal recommends against using a CFL light source.

 

[SecureGSM]

correct. CFL and LED lighting is a big no-no for running AFMA calibration as light source filcker affects AF consistency tremendously. One will end up with very inconsistent calibration data. Same valid for runing AFMA calibration in low light conditions with some lenses more affected than others.

IIRC, FoCal recommends against using a CFL light source.

 

[Random Orbits]

Thanks everyone, good to know it's just part of living with an AFMA DSLR.

I'm one of those that swears by rather than at dot-tune, I tried it as soon as I got my 6D and it's worked perfectly every time I've done it. I've now redone all lenses and they all needed to move plus 8~10 points... so I think something in the body moved a fraction.

living here in the UK we don't see the sort of temperature extremes one can get elsewhere, and I'm lucky that our house is pretty stable temperature wise. The gear is never left in the car to cook. I don't recall bumping the body but clearly something's shifted things..

anyway, all good now.

I wouldn't be surprised if the camera was bumped inadvertently. I had a 5DII that dropped a few feet onto the dirt and that moved everything by about 5 clicks. That is one good thing about sending in a camera for clean and check... the two times I've sent it in for work, it's come back set accurately to 0.

 

[Viggo]

I got repeatable results with the exact same AFMA value in both sunlight and using my halogen work lamp. This I also
Tested when I first got FoCal. This because I know lights flicker and wanted to test both the sun and others. It gave consistent front focus with the sun and other sources, on any lens/camera combo.

 

[leadin2]

Thanks everyone, good to know it's just part of living with an AFMA DSLR.

I'm one of those that swears by rather than at dot-tune, I tried it as soon as I got my 6D and it's worked perfectly every time I've done it. I've now redone all lenses and they all needed to move plus 8~10 points... so I think something in the body moved a fraction.

living here in the UK we don't see the sort of temperature extremes one can get elsewhere, and I'm lucky that our house is pretty stable temperature wise. The gear is never left in the car to cook. I don't recall bumping the body but clearly something's shifted things..

anyway, all good now.

I wouldn't be surprised if the camera was bumped inadvertently. I had a 5DII that dropped a few feet onto the dirt and that moved everything by about 5 clicks. That is one good thing about sending in a camera for clean and check... the two times I've sent it in for work, it's come back set accurately to 0.

That's what I did. The Canon technician told me he did a reset to factory default, no calibration done on my 5D2. Everything was back to normal after that. How I wish I can do the reset myself instead of sending it in. Of course, that can only be done if it's a body AF alignment issue. So far I have no lens AF issue.

 

[SecureGSM]

that would be a correct outcome, in my experience, for both halogen and incandescent light. However CFL and LED that were mentioned above is a different animal.

I got repeatable results with the exact same AFMA value in both sunlight and using my halogen work lamp. This I also
Tested when I first got FoCal. This because I know lights flicker and wanted to test both the sun and others. It gave consistent front focus with the sun and other sources, on any lens/camera combo.