MA - Does It Change?

I'm wondering if anyone else has experienced the same issue. I've spent a good deal of time getting my two primes nailed on my two bodies. The other day I'm shooting with my 5D and 50mm 1.4 and notice every shot is backfocused. MA is set 'back' 5. I put it to 0 and everything is perfect. I kid you not though, just the other week back 5 was perfect.
Thoughts? Similar experiences?

Believe or not it may even depend on the subject distance. That's why I suggest people to do MFA for the most often used distance.

Same distance, different lighting

Wow, that's a tough one...

I sometimes wonder if MA is more of a curse than a blessing so to speak.

Mostly it's a blessing. I have a lens that works waaay better on my 7D (and former 50D) than on my 40D due to MA

But it is one more variable that makes poor performing gear now your problem/fault.

Maybe, maybe it was the 5DmkII's fault. All I know is that I had that POS dialed in and working for at least two weeks. Now it's on a different MA setting and working just as well.

It depends on how "different" the lighting is. If might simply be an autofocus issue. AF is most accurate in very bright light such as full sunlight. Late in the day with full shade, AF will not be nearly as accurate. Unlike an exposure where lighting is normalized, AF uses all of the available light that can be gathered with the iris wide open.

50mm 1.4 wide open in what I would guesstimate to be two stops of light different.

Mr. Clean wrote in post #9215557
I'm wondering if anyone else has experienced the same issue. I've spent a good deal of time getting my two primes nailed on my two bodies. The other day I'm shooting with my 5D and 50mm 1.4 and notice every shot is backfocused. MA is set 'back' 5. I put it to 0 and everything is perfect. I kid you not though, just the other week back 5 was perfect.
Thoughts? Similar experiences?

It also depends on how you determined the MFA setting. If you shot at an angled target, the results can be highly suspect. The focus area for each sensor is actually rather large and while you may have been checking for maximum sharpness at a particular point on an angled target, the actual point of focus could easily be displaced from that desired point. A more reliable method which is also very quick and easy is to just set up a target approximately perpendicular to the optical axis. Take a series of shots of the target over a selected range of micro focus adjsutments (you can get an approximate eyeball idea what part of the MFA gives the best results). Refocus the lens before each shot by first manually setting focus to infinity and then performing one-shot AF. Make several separate measurements for each MFA becaue you will find that AF is not perfect and results can vary from one shot tot he next. Do this outdoors in bright daylight for best results. There will probably be no clear-cut best setting, but probably a judgment call between two or three values.

Mr. Clean wrote in post #9217421
50mm 1.4 wide open in what I would guesstimate to be two stops of light different.

Two stops can be significant especially if we are not talking about extremely bright light in the brighter case.

I discovered that if I tried to do the MFA adjustment indoors even with good lighting that AF accuracy could account for about +/- 2 clicks variation. After I got tired of chasing the MFA all over the place indoors, I set up outdoors in bright sunlight and got far better results. However, that does not change the fact that AF will still be somewhat "looser" under lower lighting conditions.

No noob to MA, that's exactly how I do it, using the LensAlign product.

Bill Boehme wrote in post #9217431
It also depends on how you determined the MFA setting. If you shot at an angled target, the results can be highly suspect. The focus area for each sensor is actually rather large and while you may have been checking for maximum sharpness at a particular point on an angled target, the actual point of focus could easily be displaced from that desired point. A more reliable method which is also very quick and easy is to just set up a target approximately perpendicular to the optical axis. Take a series of shots of the target over a selected range of micro focus adjsutments (you can get an approximate eyeball idea what part of the MFA gives the best results). Refocus the lens before each shot by first manually setting focus to infinity and then performing one-shot AF. Make several separate measurements for each MFA becaue you will find that AF is not perfect and results can vary from one shot tot he next. Do this outdoors in bright daylight for best results. There will probably be no clear-cut best setting, but probably a judgment call between two or three values.

Bill Boehme wrote in post #9217494
Two stops can be significant especially if we are not talking about extremely bright light in the brighter case.

I discovered that if I tried to do the MFA adjustment indoors even with good lighting that AF accuracy could account for about +/- 2 clicks variation. After I got tired of chasing the MFA all over the place indoors, I set up outdoors in bright sunlight and got far better results. However, that does not change the fact that AF will still be somewhat "looser" under lower lighting conditions.

Bill in this case it sounds like you think this is acceptable? Let's say two stops, iso 100 to 400. It's okay for the camera the miss focus?
My opinion is that from 100 to, let's say 1600, there should be no variation in the AF ability of the camera and the lens. Your micro adjustment figures should not change based on the ISO you have to use.

Mr. Clean wrote in post #9217710
No noob to MA, that's exactly how I do it, using the LensAlign product.

Okay, you aren't a noob to MA, but just because you adjusted it to super accurate focus using LensAlign doesn't mean that the real world situation will give a perfect subject on which to focus. It is very easy to end up with an image that is not focused on exactly what you intended due to the size of the autofocus sensor. I spent a couple of weeks working with my 50D when I first got it setting MA and then learning just how it focused in the real world. I went from disappointing images to sharp images. I had to learn the camera, I'm still in that process but of course photography is always a learning process.

I tend to work with focus a lot like I do with exposure when I think it is really critical, I "bracket" focus as well for lack of a better term. For example, I was shooting a very small mushroom with the 70-200 f/4 L at f/4. You can imagine that it is a curved surface and small and at f/4 there is little depth of field. I had to shoot multiple images trying to find just the right spot to focus on since the AF sensor could pick up a different spot than I wanted due to the small size of my subject in relation to the size of the sensor, and the fact the lighting was low. That is much different than shooting an image using LensAlign.

It is all about knowing the way your camera behaves under different circumstances and technique. When I perfect that I'll let you know.

Cheers