Reading mode: Light Dark

Canon 1D Mark III micro adjustments ???

Go to last post
T

tomfil

Member
Messages
11
Reaction score
0
Location
US
I take some shots of a static object and i like you opinion about the best shot. I focus in 2.8 number in the paper.I take -20 setting to a +20 setting using increments of 5 between each shot. I Its normal thar the microadjustments with and without converter 1,4 are not equal???

http://www.pbase.com/tomfil

Thks in advance
 
I would assume they would not be different, as the the manual for the 1d3 notes that the camera will distinguish settings between the lens with and without an extender.
--
----
Thanks!
Jordan Wilberding
http://diginux.net
 
I expect this is normal. My own settings:

70-200 f/4L: -5
with 1.4 TC II: +8

300 f/4 L IS: +3
with 1.4 TC II: -3

I got these values using a 13*19" chart I made, placed 18.5 feet from the camera (a randomly chosen distance that meets Canon's suggestion of 10-20X lens focal length), and exactly parallel to the camera's film plane and in line with the optical axis. Camera was mounted on a very rigid Bogen tripod, ISO 800 to get fairly high shutter speeds. IS was off.

After getting these results, I shot sequences of a number of objects with the camera hand-held and IS on. In all four cases, the objects were in nearly perfect focus (can't blame the camera if I moved just a touch). This was much better result than when I used the default camera settings.

BTW: After you're done with the lens microadjustments, try playing with picture settings sharpness, contrast, and saturation. I always use Neutral picture setting, and the effects of changing the individual settings are pretty interesting. Also, I spoke with Canon's support folks, and they said these adjustments affect both RAW and JPEG (which surprised me). My tests suggest they're right.

Abbott
 
Interesting I thought RAW was RAW and nothing effected it, accpet maybe highlight tone priority which seems to make the exposure different in jpg than RAW. Ok if you use DPP because the program understand the function and the devlops your RAW correctly.

I watched this short video which I thought strange because the man states he shoots RAW then talks quite a bit about WB and picture stlye and other things that I thought would effect RAW images.

http://www.abetterbouncecard.com/

Video #5: How I set up my Canon 1D MKIII

--
Have a look and leave a message
http://www.pbase.com/jeremy_broome_smith
 
Abbott Schindler said:
I expect this is normal. My own settings:

70-200 f/4L: -5
with 1.4 TC II: +8
Click to expand...
I find your numbers kind of remarkable. Have you ever had the 70-200 calibrated by Canon?

Regards,

Joe Kurkjian, Pbase Supporter



SEARCHING FOR A BETTER SELF PORTRAIT
 
At just a couple minutes into the video I stopped watching to reply... he mentioned he set the Picture Style for how he wants the images to look on the rear LCD. I do this, too, but usually process with DPP or ACR with a more saturated look. The LCD looks terrible with too much saturation, i.e. Landscape, for very colorful subjects, like race cars, and then hard to use the LCD to judge focus.

As far as I can tell, the WB and Styles do not affect RAW, i.e. are not locked in, but in DPP uses those settings as defaults when it reads the files. In ACR the Styles are ignored but the in-camera WB setting is the default. I just verified that with Lightroom (I can't verify with CS3 as it's not loaded on the machine I am using today).
 
Yes, I had the lens calibrated shortly after I got it a couple of years ago. With the dialed-in adjustments, the lens provides dead-on focus throughout its range.

I just sent my 100-400L in for calibration and some adjustments. Am looking forward to seeing what sort of adjustments it requires.

But why do you find my numbers remarkable?
J A K said:
Abbott Schindler said:
I expect this is normal. My own settings:

70-200 f/4L: -5
with 1.4 TC II: +8
Click to expand...
I find your numbers kind of remarkable. Have you ever had the 70-200
calibrated by Canon?

Regards,

Joe Kurkjian, Pbase Supporter



SEARCHING FOR A BETTER SELF PORTRAIT
Click to expand...
 
Hi Abbott:

Remarkable ... I guess "shocked" would have also worked. The 100-400 on my 1DmkII (as well as 10D, 30D, 40D) shows virtually zero static AF error AND when I tested the 1.4X plus 100-400/1DmkII the static error was again zero.

In fact, the AF focus error is so close to perfect I can't equal it manually. Once I was manually testing (using the Angle Finder C) the 100-400 at 400mm and found the resolution of the manual focus was too coarse to actually nail focus at 400mm; it's a matter of gears and on some lenses you might manually adjust perfectly at 400mm but not at 380mm. Anyway, USM has near infinite resolution and the resolution target I was shooting just popped up to perfect when I switched to AF. I first discovered this on my 10D (yeah I know, I was lucky to get a 10D with perfect AF). There have been a few folks that have said the 100-400 lens is soft at the long end and I've always felt they just came up against dead-band of the coarse manual adjustment (naturally most folks testing a lens would choose to use manual focus to eliminate AF errors but in the case of my gear the AF error is zero).

Here is my test using the 100-400/10D with the "taped pins" to fool the AF system when the 1.4X was used. There are four 100 percent crops in one composite; this is what the resolution target looked like with and without the 1.4X, both manually focused and Single-shot AF enabled.



Anyway, sorry for the long winded answer but I had to give you a little background information so you would understand why I thought your lens had a calibration problem. I'm glad to hear you have sent the 100-400 to Canon for calibration. I hope you come back and report what your micro-adjust settings are with and without the 1.4X attached; IMO if everything is right-on you should not see any difference. So please, come back and post your results using a calibrated lens, thanks.

Regards,

Joe Kurkjian, Pbase Supporter



SEARCHING FOR A BETTER SELF PORTRAIT
 
Interesting, Joe.

I have a number of Canon lenses and tested them all. Procedure was to mount my target (which was a pattern of lines printed on an Epson R2400), position the camera/lens combinations as described earlier. Then I'd set the AF microadjustment, defocus the lens, allow the camera to AF with the center point selected. Camera was released with a remote release and shutter speed was always 1/300 or faster. First I ran in increments of 5, then in individual increments between the resulting 2 sharpest 5's. As Canon suggests, I viewed high-res JPEG files at 100% to make my decisions. I believe I was able to resolve to the nearest microadjustment increment. And then I verified my results in the field. Is this the sort of procedure you did as well?

None of my lenses came out dead-on. In fact, the reason I tested so extensively is that the first time I ran tests (much less rigorously), I came out with virtually no correction required on any lens. When I mentioned this to a Canon support person (when I called about my 100-400), she expressed astonishment that nothing needed correction. That's when I became a bit more disciplined (made the target, used my heaviest tripod, was more deliberate about defocusing between frames, etc.). Sure enough, when tested extremely carefully, each lens required a bit of adjustment.

I should add that even without microadjustment, all the lenses seemed to focus just fine in the viewfinder. But when viewed on a monitor or in a print, those "visually in-focus" images were just a tad out. That doesn't seem to happen since I calibrated the camera to each lens.

I'll post back to this thread in a few weeks when I get the 100-400 back (Canon said to expect 3-4 weeks).

Abbott
J A K said:
Hi Abbott:

Remarkable ... I guess "shocked" would have also worked. The 100-400
on my 1DmkII (as well as 10D, 30D, 40D) shows virtually zero static
AF error AND when I tested the 1.4X plus 100-400/1DmkII the static
error was again zero.

In fact, the AF focus error is so close to perfect I can't equal it
manually. Once I was manually testing (using the Angle Finder C) the
100-400 at 400mm and found the resolution of the manual focus was too
coarse to actually nail focus at 400mm; it's a matter of gears and on
some lenses you might manually adjust perfectly at 400mm but not at
380mm. Anyway, USM has near infinite resolution and the resolution
target I was shooting just popped up to perfect when I switched to
AF. I first discovered this on my 10D (yeah I know, I was lucky to
get a 10D with perfect AF). There have been a few folks that have
said the 100-400 lens is soft at the long end and I've always felt
they just came up against dead-band of the coarse manual adjustment
(naturally most folks testing a lens would choose to use manual focus
to eliminate AF errors but in the case of my gear the AF error is
zero).

Here is my test using the 100-400/10D with the "taped pins" to fool
the AF system when the 1.4X was used. There are four 100 percent
crops in one composite; this is what the resolution target looked
like with and without the 1.4X, both manually focused and Single-shot
AF enabled.



Anyway, sorry for the long winded answer but I had to give you a
little background information so you would understand why I thought
your lens had a calibration problem. I'm glad to hear you have sent
the 100-400 to Canon for calibration. I hope you come back and
report what your micro-adjust settings are with and without the 1.4X
attached; IMO if everything is right-on you should not see any
difference. So please, come back and post your results using a
calibrated lens, thanks.

Regards,

Joe Kurkjian, Pbase Supporter



SEARCHING FOR A BETTER SELF PORTRAIT
Click to expand...
 
Abbott Schindler said:
Interesting, Joe.

I have a number of Canon lenses and tested them all. Procedure was to
mount my target (which was a pattern of lines printed on an Epson
R2400), position the camera/lens combinations as described earlier.
Then I'd set the AF microadjustment, defocus the lens, allow the
camera to AF with the center point selected. Camera was released with
a remote release and shutter speed was always 1/300 or faster. First
I ran in increments of 5, then in individual increments between the
resulting 2 sharpest 5's. As Canon suggests, I viewed high-res JPEG
files at 100% to make my decisions. I believe I was able to resolve
to the nearest microadjustment increment. And then I verified my
results in the field. Is this the sort of procedure you did as well?
Click to expand...
First off, I don't own a 1DmkIII.

Maybe I missed exactly what you own; is it a blue-dot or an older unit with (or without) the AF fix, or what (sorry if you already mentioned this someplace else)?

Regarding your procedure, off the top of my head it sounds very well thought out. One thing I would suggest is to slew from both minimum focus distance and infinity to ensure there is no dead band in the AF loop. I've not run into dead band in any of my tests but make the suggestion for you to consider to eliminate the possibility of dead band.

Just out of curiousity have you considered repeating the micro-adjustment at several magnifications (e.g. the object to sensor distance could be set to 11, 21, and 31 times the focal length)? Of sourse the purpose of this "extended testing" is to confirm the goodness of the micro-adjustment over a range of magnifications.
Member said:
None of my lenses came out dead-on. In fact, the reason I tested so
extensively is that the first time I ran tests (much less
rigorously), I came out with virtually no correction required on any
lens.
Click to expand...
It's hard to imagine you would that much less rigorous to the point of not seeing an AF error. Maybe your camera changed; have you considered this possibility. Would you mind explaining "how" you could have missed an OOF condition?
Member said:
When I mentioned this to a Canon support person (when I called
about my 100-400), she expressed astonishment that nothing needed
correction. That's when I became a bit more disciplined (made the
target, used my heaviest tripod, was more deliberate about defocusing
between frames, etc.). Sure enough, when tested extremely carefully,
each lens required a bit of adjustment.

I should add that even without microadjustment, all the lenses seemed
to focus just fine in the viewfinder. But when viewed on a monitor or
in a print, those "visually in-focus" images were just a tad out.
That doesn't seem to happen since I calibrated the camera to each
lens.

I'll post back to this thread in a few weeks when I get the 100-400
back (Canon said to expect 3-4 weeks).
Click to expand...
Great, I'll be looking forward to that, thanks very much!
Member said:
Abbott
J A K said:
Hi Abbott:

Remarkable ... I guess "shocked" would have also worked. The 100-400
on my 1DmkII (as well as 10D, 30D, 40D) shows virtually zero static
AF error AND when I tested the 1.4X plus 100-400/1DmkII the static
error was again zero.

In fact, the AF focus error is so close to perfect I can't equal it
manually. Once I was manually testing (using the Angle Finder C) the
100-400 at 400mm and found the resolution of the manual focus was too
coarse to actually nail focus at 400mm; it's a matter of gears and on
some lenses you might manually adjust perfectly at 400mm but not at
380mm. Anyway, USM has near infinite resolution and the resolution
target I was shooting just popped up to perfect when I switched to
AF. I first discovered this on my 10D (yeah I know, I was lucky to
get a 10D with perfect AF). There have been a few folks that have
said the 100-400 lens is soft at the long end and I've always felt
they just came up against dead-band of the coarse manual adjustment
(naturally most folks testing a lens would choose to use manual focus
to eliminate AF errors but in the case of my gear the AF error is
zero).

Here is my test using the 100-400/10D with the "taped pins" to fool
the AF system when the 1.4X was used. There are four 100 percent
crops in one composite; this is what the resolution target looked
like with and without the 1.4X, both manually focused and Single-shot
AF enabled.



Anyway, sorry for the long winded answer but I had to give you a
little background information so you would understand why I thought
your lens had a calibration problem. I'm glad to hear you have sent
the 100-400 to Canon for calibration. I hope you come back and
report what your micro-adjust settings are with and without the 1.4X
attached; IMO if everything is right-on you should not see any
difference. So please, come back and post your results using a
calibrated lens, thanks.

Regards,

Joe Kurkjian, Pbase Supporter



SEARCHING FOR A BETTER SELF PORTRAIT
Click to expand...
Click to expand...
 
tell me where/what you are shooting at to set these cameras. Is it something I can print out?

Also at what distance are you guys, are you going a certain distance past the minimum focusing distance of the lens?

And last do you have any good documents or web sites on the micro adjustment procedure. You guys seem to have some very good information on the procedure
Thanks
--
Taking life one picture at a time ;-)
http://www.pbase.com/samd12
 
My notes are below:
J A K said:
First off, I don't own a 1DmkIII.
Click to expand...
That could be part of the discussion. My discussion is all about 1DmkIII.
J A K said:
Maybe I missed exactly what you own; is it a blue-dot or an older
unit with (or without) the AF fix, or what (sorry if you already
mentioned this someplace else)?
Click to expand...
Blue dot. S/N 557xxx with firmware 1.1.3.
J A K said:
Regarding your procedure, off the top of my head it sounds very well
thought out. One thing I would suggest is to slew from both minimum
focus distance and infinity to ensure there is no dead band in the AF
loop. I've not run into dead band in any of my tests but make the
suggestion for you to consider to eliminate the possibility of dead
band.
Click to expand...
I slewed about half a turn of the manual focusing ring, always in the same direction. Also, I visually verified that the focusing distance shown (which is a very crude check) always came to about the same place after slewing and then applying AF. I did this with each lens several times. I also looked at where the lens focus ended up during most of the individual shots. I believe you're referring to hysteresis and I think I eliminated that as well as possible.
J A K said:
Just out of curiousity have you considered repeating the
micro-adjustment at several magnifications (e.g. the object to sensor
distance could be set to 11, 21, and 31 times the focal length)? Of
sourse the purpose of this "extended testing" is to confirm the
goodness of the micro-adjustment over a range of magnifications.
Click to expand...
I didn't do this during the rigorous test, but I did do it during my field verification tests. I shot a series of 8-10 images for each lens, and the subject-to-camera distances varied from 8 feet to 250 yards (same series for all lenses). For these tests, I kept the f/stop wide open (Av mode). As I said, all shots came out spot-on as far as I can tell.
J A K said:
J A K said:
None of my lenses came out dead-on. In fact, the reason I tested so
extensively is that the first time I ran tests (much less
rigorously), I came out with virtually no correction required on any
lens.
Click to expand...
It's hard to imagine you would that much less rigorous to the point
of not seeing an AF error. Maybe your camera changed; have you
considered this possibility. Would you mind explaining "how" you
could have missed an OOF condition?
Click to expand...
Simple: first set of tests was done using a procedure I've seen published before: print a letter-size focus chart (I believe it's called "Focus21"), set the cam up on a tripod with the chart at 45 degrees to the camera. Because of the small chart size and fineness of its markings, the lenses have to be relatively close...nowhere near the 10-20X focal length that Canon recommends. I also used a lighter-duty tripod which may have allowed for a bit of camera shake. The resulting test images all looked pretty good around no microadjustment. This was just a sloppy test, I think. the microadjustment is just that: micro. I'm now convinced that you need a pretty decent setup to properly evaluate what's needed.

Abbott
 
Haven't tested this lens yet; it's at Canon for servicing. When I do the test, I'll be at about 20 feet at 400 mm fl and wide open. This is within what I've seen Canon recommend (10-20x fl, which would be 4-8 meters, or about 13-26 feet). I'll then go outdoors and do my usual series of hand-held verification images.
pikeface said:
When you were testing your 100-400 can you say how far away you were
please.
--
Have a look and leave a message
http://www.pbase.com/jeremy_broome_smith
Click to expand...
 
Abbott Schindler said:
J A K said:
First off, I don't own a 1DmkIII.
Click to expand...
That could be part of the discussion. My discussion is all about
1DmkIII.
Click to expand...
I'm following the discussion just fine. :-) You have been talking about the 1DmkIII and micro-adjustment experience with two lenses. My "surprise" is about the micro-adjustment difference between 1.4X on/off results and I'm merely "suggesting" this may be an issue with the lens (and not the camera).
Member said:
Member said:
Maybe I missed exactly what you own; is it a blue-dot or an older
unit with (or without) the AF fix, or what (sorry if you already
mentioned this someplace else)?
Click to expand...
Blue dot. S/N 557xxx with firmware 1.1.3.
Click to expand...
Thanks very much for that information. I'm trying to categorize how blue dots are performing compared to non-blue dots that were subjected to the AF fix. It's way too early to arrive at a conclusion regarding AI Servo AF performance regarding old units with the AF fix. However, at this point in time, I'm "almost" convinced the static AF accuracy seems to be better than that preceding the AF fix, and this may be due to an extremely careful camera calibration post AF fix.
Member said:
Member said:
Regarding your procedure, off the top of my head it sounds very well
thought out. One thing I would suggest is to slew from both minimum
focus distance and infinity to ensure there is no dead band in the AF
loop. I've not run into dead band in any of my tests but make the
suggestion for you to consider to eliminate the possibility of dead
band.
Click to expand...
I slewed about half a turn of the manual focusing ring, always in the
same direction. Also, I visually verified that the focusing distance
shown (which is a very crude check) always came to about the same
place after slewing and then applying AF. I did this with each lens
several times. I also looked at where the lens focus ended up during
most of the individual shots. I believe you're referring to
hysteresis and I think I eliminated that as well as possible.
Click to expand...
No, I know what hysteresis is. I'm talking about a "dead band", a physical focus point that cannot be achieved at a specific focal length via manual control (a gear mesh issue) but can be achieved via AF because of the higher resolution associated with USM. I don't believe for one minute you have a dead band issue using the AF system on a true USM lens; but I was suggesting you "might" want to test for it just to eliminate the possibility. It's entirely possible that some hysteresis is designed into Canon's AF servo loop design but I've not seen any evidence of that in the 10D, 30D, 40D, and 1DmkII designs, and would seriously doubt Canon added hysteresis to the 1DmkIII design (but I always reserve the right to be wrong).
Member said:
Member said:
Just out of curiousity have you considered repeating the
micro-adjustment at several magnifications (e.g. the object to sensor
distance could be set to 11, 21, and 31 times the focal length)? Of
sourse the purpose of this "extended testing" is to confirm the
goodness of the micro-adjustment over a range of magnifications.
Click to expand...
I didn't do this during the rigorous test, but I did do it during my
field verification tests. I shot a series of 8-10 images for each
lens, and the subject-to-camera distances varied from 8 feet to 250
yards (same series for all lenses). For these tests, I kept the
f/stop wide open (Av mode). As I said, all shots came out spot-on as
far as I can tell.
Click to expand...
That's great, sounds pretty much like you are in hog heaven with your camera; which is exactly where you should be.
Member said:
Member said:
Member said:
None of my lenses came out dead-on. In fact, the reason I tested so
extensively is that the first time I ran tests (much less
rigorously), I came out with virtually no correction required on any
lens.
Click to expand...
It's hard to imagine you would that much less rigorous to the point
of not seeing an AF error. Maybe your camera changed; have you
considered this possibility. Would you mind explaining "how" you
could have missed an OOF condition?
Click to expand...
Simple: first set of tests was done using a procedure I've seen
published before: print a letter-size focus chart (I believe it's
called "Focus21"), set the cam up on a tripod with the chart at 45
degrees to the camera. Because of the small chart size and fineness
of its markings, the lenses have to be relatively close...nowhere
near the 10-20X focal length that Canon recommends. I also used a
lighter-duty tripod which may have allowed for a bit of camera shake.
The resulting test images all looked pretty good around no
microadjustment. This was just a sloppy test, I think. the
microadjustment is just that: micro. I'm now convinced that you need
a pretty decent setup to properly evaluate what's needed.
Click to expand...
Okay, thanks for explaining that and taking time to write the comprehensive reply.

BTW, my passing interest in all this nit picky stuff is simply that I'm "interested", nothing more complex than that. My "real" interest is what happens during real world shooting for my specific application. The animals I shoot have a specific depth that I'm trying to get in focus and "usually" if something is 20 feet from me I'm at f/9 or so and when they are four or five feet from me I'll stop down to f/14 or f/16. My primary interest in AI Servo AF is when I'm shooting ducks; I need fairly fast "and" accurate acquisition; once acquired accurate tracking is all I ask for.

Thanks again for sharing your experience(s).

Regards,

Joe Kurkjian, Pbase Supporter



SEARCHING FOR A BETTER SELF PORTRAIT
 
You must log in or register to reply here.

Keyboard shortcuts

f
Forum
m
My threads
Mobile site
About
Editorial content
Cameras & Lenses
Community
All content, design, and layout are Copyright © 1998–2025 Digital Photography Review All Rights Reserved.
Reproduction in whole or part in any form or medium without specific written permission is prohibited.
When you use DPReview links to buy products, the site may earn a commission.
©GPS Media - Guides, Products, Services.
Back
Top