AF Microadjustment, Lensalign Pro and Repeatability

[HDNitehawk]

I sent of and got the Lensalign Pro last week. I set it up this weekend and began testing my back up camera (which I use as a walk around camera for work) and the 35mm F1.4L. I am using a 5D Mark II on this test.


I set it up at the minimum distance, ran a series of pictures and checked them on the computer. Very accurate results in the set. Then I tried the camera at various other distances. I found only slight variations in results in these three test which could be contributed to distance.


Then later I decided to check again, because I am somewhat of a perfectionist. Every time I have reset up in a different location I get slightly different results. Some may be contributed to lighting, maybe not because I have tried to eliminate this possibility.


My results have varied from Plus 3 with Plus 3 leaning toward the forward focus side and Plus 6 leaning toward the back focus side.


How I am taking the pictures is sets of 8 at each point, from 0 to 10. Renaming the files out of EXIF data and then reviewing full screen. Either at 100 or 200% depending. Each set has been very consistent and accurate in that you can easily establish the DOF that is being generated at each setting and see the gradual change from one point to the next.


The variation in results is from one set to the next.


I understand that the wider angle lens will not adjust as well as the telephoto's.


So I guess what I am trying to figure out is this normal for the 35mm lens? Is this just from the subtle change in lighting from one test set up to the next?


Or does the camera AF system float a little bit and this is the norm?


Or is the simple answer, I am to picky, set it at plus 5 it will back focus more than front focus and go test the next lens.

[andnowimbroke]

I did have trouble at first because of weak light, but even then, I had a lens that varied from distance to distance and actually used the afma enable/disable button to change between working distances ( enable +5 for distant shots and disable for closer work or whatever it was).I can't really speak for your lens though. I also cheated later on when checking after a couple months. I did the 10x live view thing ( zoomed to the scale side) with flapping mirror only I just focused. The ones that didn't focus right the first time, I just refocused again without manually unfocusing and it would usually sneak up and hit it the next time. I guess that meant it was in the margin of error. I wouldn't set up a camera like that, but just from a re-test point of view, I guess it didn't waste too much time, and I got to get out my ruler for only the second time since I bought it

set it at plus 5 it will back focus more than front focus

Which way would be best? Doesn't blur in front of yourfocus point"work faster" for a lack of a better word and the blur behind gives you more room play? Hope that made sense. And just to put things into perspective, with the system, your getting rid of all variables, but once it's in your hands and your swaying backward and foreword or they are..well, afma isn't meaningless, but it's not the end-all-be-all either.

[HDNitehawk]

Which way would be best? Doesn't blur in front of yourfocus point"work faster" for a lack of a better word and the blur behind gives you more room play? Hope that made sense. And just to put things into perspective, with the system, your getting rid of all variables, but once it's in your hands and your swaying backward and foreword or they are..well, afma isn't meaningless, but it's not the end-all-be-all either.

Thanks andnowimbroke


I suppose which way would work best would be the back focus. At 12' the DOF for this lens at 1.4 is 3'. I think its real close that the autofocus is really going to be important, between 3 and 8 feet in a room of people taking pictures. I suppose aim for the nose and you should catch there face very well if your back focused just a little?





Rick

[neuroanatomist]

Hmmm...I haven't done immediate test-retest of my adjustments, but when I went through the AFMA on my 5DII, I also retested 4 lenses (or so I was thinking at the time) that I had previously AFMA'd on my 7D, with 4-6 months between tests. Two of the four (85L, 100L Macro) were the same, one was different by one unit (70-200mm II went from +1 to +2). The fourth lens was the 24-105mm, and that one had a substantial change - then I smacked my head as I realized that it wasn't the same specific lens - I initially bought a used copy, which I sold when I bought the 5DII to get a new lens at the kit price. So, of three lenses retested two had no difference and one changed by one unit.

[HDNitehawk]

Last night I restested again the 35mm. I set up directly under the overhead lighting in the living room so to get good pictures. I left the tripod set up and did several tests. I ended up at plus 5 and they were consistent. I am thinking it is just the subtle variations inlighting.


But. I set up to do the 24mm F1.4L using my main 5d mark II and ran the range from minus 10 to plus 10. I ended up with a very good 9


Then I brought the lens back, same tripod same set up but this time I turned off the microadjustment and did 10 pics on default. Then turned it back on and ran a test betweeen plus 5 and plus 15. The second set, was totaly off.....it was hitting on about plus 6.


So a third test, this time it hit plus 8 to plus 9 like the first.


I am wondering if just doing the test where your continualy changing the settings over and over if the camera might read the setting a little diffrent each time. What I am going to try to do now is do the test, find the setting, and come back later to see if it duplicates by just using only the one setting.


These two lens I really don't need percission and accuracy after about 12 feet. But real good accruacy up close will make some interesting pics at f1.4

[neuroanatomist]

I set up directly under the overhead lighting in the living room so to get good pictures.

Is the overhead light your only light source? If so, is the LensAlign positioned far enough back so the focus target is well-lit? Here's the setup I use for longer focal lengths, where I need the target near a wall so I can set up the camera up to 32' from the target. There's a light shining down onto to the ruler so that's illuminated (the side light on a Tiffany torchire lamp, in this case), but also notice the gooseneck light pointing at the LensAlign from the front for a bright, even illumination of the focus target plate.


[img]/cfs-file.ashx/__key/CommunityServer-Components-UserFiles/00-00-00-35-15/LensAlign-setup.JPG[/img]