I finally received my RRS L-Bracket in the mail yesterday. With that said, I thought it was time to see what this camera can do (1Ds MkIII). Of course, I had to make sure my lenses were tack sharp. In order to do that, I had to find out what the "AF Microadjustments" was all about.

One word...WOW!!! I thought my lenses were sharp and sweet before. Shoot, based on my results, I should have thrown the lenses away. Of course, I would never throw them away. That's like asking a crack head to give up his supplier. In short, these lenses are far sharper now than ever before (except for the 85mm).

In the end, I made the following adjustments...Tell me what, if any, were your adjustments? I'm curious to see what others are experiencing.

Canon 14mm f/2.8 MkII (at f/8)
+20

Canon 50mm f/1.4 (at f/8)
-10

Canon 85mm f/1.8 (at f/8)
0 (no adjustment needed)

Canon 135mm f/2 (at f/8)
-20

Canon 135mm f/2 (w/ 1.4x MkII) (at f/8)
-10

Canon 24-70mm f/2.8 (at f/8)
+5

Canon 70-200mm f/2.8 (at f/8)
-20

Canon 70-200mm f/2.8 (w/ 1.4x MkII) (at f/8)
-20

Tamron 28-75mm f/2.8 (at f/8)
-20

By the way...Is the Canon 1Ds Mk III the first and only camera to have this feature? Do any of the other Canon models allow you to perform this mod...what about Nikon?

Thanks,
Karl

The micro adjustments are helpful but not what I had hoped for.
What distance did you make these adjustments at? I have found that if I get all the micro adjustments right on at 15 feet and use the lens at 15 feet its great. If I then shoot at 5 feet or 30 feet the adjustment misses the mark.
It seems the only way to get it the way I want it is to send all my glass and my camera to Canon, wait a few weeks then hope they did it right - they don't always get it right.

Okay...You may have just rained on my parade. I based my analysis purely on the test results. I will run some additional tests tomorrow at a shorter and longer distance based on the new settings and let you know what I come up with.

The test distance varied from lens to lens...from 1ft (14mm) to 10ft (70-200mm).

Whoa, mine are generally within ±3 units of neutral at medium-far distance. I don't test them up close. Do those lenses AF accurately on other bodies that do not have AF adjustments?

EBH

+3 or -3 Whoa is right. I will definitely have to take another look and get back to you. The lenses work great on my wife's 20d...Shoot, I thought they worked great on my MkIII, but that was before I saw test shots after making the adjustments. To be continued tomorrow...I will let you know what I come up with.

Thanks.

Is there a published 'best practices' for performing the adjustment? Does the manual lay it out pretty specifically? I've been curious about this (don't own any cameras capable of it).

On my 1D3, I'm between 0 and +8 on all of my lenses and combinations with a 1.4X extender. Half don't need any adjustments. I try to adjust for a normal distance, so the 35 1.4 I adjusted at 6 feet where it usually is when I shoot it wide open while the 70-200 I adjusted at 20 feet. Works decent enough for me.

Should I be on -20 as often as you Karl, I'd send my camera in.

BTW, you state @ f/8.0, but standard practice is to adjust wide open. If it's OK wide open, it stands to reason it's OK at other apertures as the AF system focuses wide open and only closes down aperture the moment you shoot.

Scott, no best practices are available. Lots of people use a focusing chart of some sort but I've resorted to real life situations as I got bored with charts. I tripod mount my camera, use mirror lockup and a remote trigger and shoot a whole series of one subject from -20 to +20 (two shots each) and then compare and rate the results. I usually wind up with a range of perfect shots (say, between +5 and +8) and then fine-tune by shooting another subject at another distance in that specific range. The combined results are usually enough to pick the final adjustment.

The only best practice, as Emile wrote, is to perform these focussing tests wide open, otherwise the DoF at say f8 masks errors.

I did try this last year and I while I don't like charts much, a sheet of newspaper taped to the wall worked well. Tripod, MLU, remote trigger will do fine.

Certainly you must adjust wide open because that's how the AF works. Yes you must adjust for the distance you use. Both Mk III allow Micro AF and also Nikon's D3. This allows the nulling of one AF variable but not all of they so it is an improvement but not the ultimate AF.

Thanks for everyone's comments. It has definitely allowed me to realize my stupidity...Thanks. Just kidding. In all seriousness, everyone's insight has helped me realize I was not performing the adjustments properly. That's what makes FM so useful and helpful...Getting help from the everyday semi-pro to the real deal pro.

SoundHound wrote:
This allows the nulling of one AF variable but not all of they so it is an improvement but not the ultimate AF.

Very true. Ultimate AF is still a dream at this point in time I think. It's an aid, but the photographer is the decision maker. For that it would be nice to have the ultimate viewfinder, but that doesn't exist either. No ultimates in photography...

I wouldn't worry too much unless your subject is flying past you at 200 Mph. For anything else that is static, there is always live view & the MF ring. AF won't always be dead-on wide open anyways. Try using it in the real world rather than the test charts. What may work for test charts can fail miserably when used in the real world, where you are not in controlled situations.

If it works, don't microadjust it. The OP's lenses work fine on his wife's 20D and Mark III, so why bother "testing". Just go out and take exposures dagumit!

Mike1 wrote:
I wouldn't worry too much unless your subject is flying past you at 200 Mph. For anything else that is static, there is always live view & the MF ring. AF won't always be dead-on wide open anyways. Try using it in the real world rather than the test charts. What may work for test charts can fail miserably when used in the real world, where you are not in controlled situations.

DITTO. Been there done that.

I had the 1DS-mk3 calibration checked when they aligned the viewfinder. It must have been off, the avg was +8 clicks before, now +4.75 clicks

17-40 +3
35 f1.4 +8 (after recent cal)
24-105 +5
70-200 +5
70-200 and 1.4 ext +5
400 f5.6 +1
500 f4 +8
500f4+1.4x +6
500f4 +2x +4
28-135 +6

The 35 f1.4 did not need any MA at real close distance, until I noticed that the center was soft while the off center rulers were prefect. Fooled by filed curvature. I retested longer and also at landscape distances and came up with +8.

My 500 is tested at 34 feet, other lengths at appropriate distances.

xrayvision wrote:
....................... I have found that if I get all the micro adjustments right on at 15 feet and use the lens at 15 feet its great. If I then shoot at 5 feet or 30 feet the adjustment misses the mark.
.......................

This seems to be a point most people don't want to hear.

I have found the same exact problem. Example, the 100mm 2.8 Macro lens. It's a +6 at subjects under a foot or so away, but at infinity, 0- -2 works much better. Annoying. The visual difference can be quite large depending the subject and focusing distances.

-J



dcmiller wrote:
xrayvision wrote:
....................... I have found that if I get all the micro adjustments right on at 15 feet and use the lens at 15 feet its great. If I then shoot at 5 feet or 30 feet the adjustment misses the mark.
.......................

This seems to be a point most people don't want to hear.

I have found the same exact problem. Example, the 100mm 2.8 Macro lens. It's a +6 at subjects under a foot or so away, but at infinity, 0- -2 works much better. Annoying. The visual difference can be quite large depending the subject and focusing distances.

-J



dcmiller wrote:
xrayvision wrote:
....................... I have found that if I get all the micro adjustments right on at 15 feet and use the lens at 15 feet its great. If I then shoot at 5 feet or 30 feet the adjustment misses the mark.
.......................

This seems to be a point most people don't want to hear.

ben egbert wrote:
...
My 500 is tested at 34 feet, other lengths at appropriate distances.


I wonder what "appropriate distance" is generally - I think I read somewhere that focus calibration is done at 40 or 50 times focal length, could that be true?

I am not sure what is appropriate length. I try to keep the image size the same for all lenses, so it is always proportional. I also test the combo on real subjects afterwards.

For example, I take my 500 outside and take pictures of birds at my feeder, they must be sharp wide open with best focus where I chose. I also use this lens out to several hundred feet, so I take a picture of a rooftop across the street. I can see how the shingles lose sharpness on each side of the target. If this fails, I re-check micro adjustment. I am getting correct focus at all distances now except very close with 35 and below (field curvature issue).

For WA lenses, I make sure that they focus properly stopped down and will render an infinity focus properly. Most of my WA lenses don't produce sharp images at infinity if I accept standard Canon calibration. In fact, I suspect that Canon calibrates with a bias to front focus. This is based on my personal experience.

I tweaked my 35L for 8' and now it's way out at 20." So I am not certain that, at big aperture of F1.2-F1.8 that many lenses will not AF properly over the entire distance range.

Don't forget that some lenses actually exhibit what is known as "focus shift" as the aperture is stopped down. Modern lens designers try to avoid or minimize it, but it does exist.

I needed a +15 micro adjustment for my 35L. ALthough I use the lens mostly at about 2-3 meters, it doesn't seem to have any problems focusing at infinity. The same for my 85L II and 135L (both needed +5).

Some lenses perform better at closer distances while other perform better at long distances and some perform good at both closer and longer distances. Maybe the MA exaggerates these design related characteristics in some way?

Here's an interesting finding. The 200 1.8 is dead on at 10 ft. -no adjustment. At 60 ft. it requires -13. Between 10 and 60 feet the microadjustment required follows a linear progression from 0 to -13. But at or near infinity it will miss the focus unless I dial the adjustment back to -10 or less. Now for the catch. When I attach the 1.4x extender it is dead on focus at all distances with no adjustment at all!
I'd love to have a Canon engineer explain this.
Since Canon has taken the ingenious descision to orphan the 200 1.8 I want to know what inside this lens is adjusted for autofocus corrections. I'll get the tools and take it apart myself if I have to - no problem.
The 35L also requires +7 at up to 8 ft. and +12 at 15 ft. keeping in mind I want to shoot at f1.4 to f2 with this lens and still have a subjects eyes in focus and not her ear

I think these "modern" zoom lenses are not parfocal like the older manual focus lenses and the engineering dept have overlooked that maybe due to time constraints. As far as primes, I think the design of the lens in correlation to distance scale has also been overlooked as well... As photographic equipment gets more complex, one would expect the lens designs to be the same, but it's not always what it should be, even though it looks good on paper. Ah... engineering at its worst.