Sponsor Message:
Aviation Photography Forum
My Starred Topics | Profile | New Topic | Forum Index | Help | Search 
AF Microadjustment And Zoom Lenses  
User currently offlineNewark727 From United States of America, joined Dec 2009, 1312 posts, RR: 0
Posted (1 year 2 months 3 weeks 13 hours ago) and read 3976 times:

About three weeks ago I posted a topic here about issues with the autofocus on my Canon EF 24-105mm IS L zoom lens, which were persistent even after a local Canon repair center had looked at the assembly and made changes. One recommendation was that I attempt some microadjustment with it and my 50D body. I did this, but I hadn't gotten a chance to attempt testing it in practice for the kind of shooting that was giving me trouble, until today. I was following the guide linked below:
http://www.northlight-images.co.uk/a...ameras/1ds3_af_micoadjustment.html
Granted, I didn't get around to the very final steps (viewing a series of test images on my computer monitor full-size) that some methods recommended, I mostly used live view and a tripod against the test pattern for as long as I could before my battery started to complain. But my question is a bit simpler than that, and it is: what should I expect when I do the microadjustment at the longest zoom as recommended, then take pictures at a shorter range?

My inconclusive efforts seem to have made some impact, which I would tentatively judge to be an improvement, at 105 mm of zoom, but either had no effect at all or made things worse between around 75 and 100 mm. Below that level focus seemed to be about as good as it had previously been and I would expect it to be, but due to my location and the types/sizes of aircraft I was getting to shoot, relatively few photos took place at lower focal lengths. Other settings on my camera varied as I was using manual mode for both shutter speed and aperture, which I varied between 6.3 and 10.

9 replies: All unread, jump to last
 
User currently offlineckw From UK - England, joined Aug 2010, 713 posts, RR: 16
Reply 1, posted (1 year 2 months 2 weeks 6 days ago) and read 3869 times:

It's a tricky one, as the complexity of modern zoom lens design means that AF calibration can vary throughout the zoom range. The Canon recommendation is logical in so far as accurate focus is more critical at the longest focal length ... at short focal lengths, the greater depth of field will tend to mask any errors.

But I think this is a sort of common sense guideline that may not work for a particular lens - note that Canon's latest models allow for different micro-adjustment settings for the same lens at different focal lengths, which does suggest that a single micro adjustment for a zoom has been found inadequate.

I think this is a matter of trial and error with your lens. I would try microadjustment at longest, shortest and mid focal lengths. You may find that a mid setting gives acceptable results at both wide and tele lengths. Conversely, if you find that you can only get decent results at one or the other zoom extremes, I'd suggest there is a more serious problem with the lens, and get Canon to look at it.

If returning to Canon, it won't hurt to include examples of the tests you've run.

Cheers,

Colin



Colin K. Work, Pixstel
User currently offlinedendrobatid From United Kingdom, joined Nov 2004, 1648 posts, RR: 62
Reply 2, posted (1 year 2 months 2 weeks 5 days 18 hours ago) and read 3833 times:
AIRLINERS.NET CREW
HEAD SCREENER

I think the camera is a far bigger issue since the advent of digital too.
In film days even the cheapest of SLRs was machined to quite a tight tolerance, the film guides were milled and the film was held in a very precise position with the pressure plate. The front of the camera was also milled with a hard lens mount affixed so that the position of the film plane was very constant from body to body.
However I am not convinced that sensors are affixed with the same degree of accuracy though micro adjustments will correct any slight differences.
I too have a 24-105 and was seriously beginning to suspect an issue with it. It sits on a Canon 20d which does not have a way of adjusting it and the results were unspectacular. It works far better on my 40d but I recently acquired a 5D mk1, (used but barely used) and the results are little short of superb. The 20d had had a new shutter fitted and I wonder if the sensor had moved a few microns which would be enough to make a difference.

Before you worry too much about the lens, try a a different body as it could be an issue.

Mick Bajcar


User currently offlineckw From UK - England, joined Aug 2010, 713 posts, RR: 16
Reply 3, posted (1 year 2 months 2 weeks 5 days 10 hours ago) and read 3800 times:

Mick perhaps has the clue to the real issue here - the 24-105 was designed with the 5D in mind ... it has been the kit lens for the 5D from day 1.

The 5D not only was full frame, but had much larger pixel sites than crop-sensors like the 20D and 40D. I wonder if the optical characteristics of the lens were optimised for the 5D - at the expense of other bodies.

I'm trying to recall - we have seen a few complaints about the 24-105 here. Were they all on crop bodies?

Cheers,

Colin



Colin K. Work, Pixstel
User currently offlineNewark727 From United States of America, joined Dec 2009, 1312 posts, RR: 0
Reply 4, posted (1 year 2 months 2 weeks 5 days 4 hours ago) and read 3782 times:

Quoting ckw (Reply 3):
I'm trying to recall - we have seen a few complaints about the 24-105 here. Were they all on crop bodies?

Well mine certainly is (a 50D as I said.) Some point when I have the time this weekend I'll test the microadjustment some more and see if I can approach a conclusive solution. Annoyingly, there isn't an option to adjust different amounts at different focal lengths though, that must be a new thing. Thank you for the advice.


User currently offlineSilver1SWA From United States of America, joined Mar 2004, 4738 posts, RR: 26
Reply 5, posted (1 year 2 months 2 weeks 5 days 3 hours ago) and read 3779 times:
Support Airliners.net - become a First Class Member!

Quoting Newark727 (Reply 4):
Well mine certainly is (a 50D as I said.) Some point when I have the time this weekend I'll test the microadjustment some more and see if I can approach a conclusive solution

It might be a good idea to send both your camera and the lens in for adjustment so they can be calibrated for each other.

Quoting ckw (Reply 3):
Mick perhaps has the clue to the real issue here - the 24-105 was designed with the 5D in mind ... it has been the kit lens for the 5D from day 1.

The 5D not only was full frame, but had much larger pixel sites than crop-sensors like the 20D and 40D. I wonder if the optical characteristics of the lens were optimised for the 5D - at the expense of other bodies.

I'm trying to recall - we have seen a few complaints about the 24-105 here. Were they all on crop bodies?

For what it's worth my 24-105 has been great on my 400D, 40D, 7D and 5D2. I have absolutely loved it since the day I bought it. Of course, it's like a whole new lens on my 5D, but never had any issues aside from the crop factor and focal length.

I do think its time I send my lens to Canon for a cleaning and calibration. It's built up a lot of internal dust over the years and I've noticed lately that I'm starting to get a lot of fringing on the outer regions of the photos.

[Edited 2013-01-31 23:15:25]


ALL views, opinions expressed are mine ONLY and are NOT representative of those shared by Southwest Airlines Co.
User currently offlinedendrobatid From United Kingdom, joined Nov 2004, 1648 posts, RR: 62
Reply 6, posted (1 year 2 months 2 weeks 5 days 1 hour ago) and read 3770 times:
AIRLINERS.NET CREW
HEAD SCREENER

Quoting ckw (Reply 3):
I wonder if the optical characteristics of the lens were optimised for the 5D - at the expense of other bodies.

Colin

I am not convinced by that argument, it does work very well on the 40d but not on the 20d.
What I am saying is that a sensor is so small and its positioning is so critical relative to the lens that only a tiny movement will affect the performance but the lens is not at fault, the camera body is. On a film camera that gap was machined to a very high precision, the actual body of the camera but I wonder if it is possible to always position the sensor so precisely on a prosumer camera. More importantly perhaps, will it stay that way?

Yes there have been a lot of complaints about this lens (Karl Nixon being one) but I wonder if the lenses he tried were all pretty similar, the one he is finally happy with just matched the camera body better - I added an image of his recently taken with it and it was lovely!

Vignetting at 24mm on the 5d is another matter but the software deals with that nicely. I have used my 24-105 for air to air and it is a remarkable lens, tiny for what it is and my 17-40 will spend a lot more time in my bag now ! Colin, you have always been a big advocate of the 5D Mk1 and I now totally understand (and it cost me peanuts too)

I have always been rather suspicious of the 'bad copy' theory of lenses, especially when we are talking about top-end lenses from top end manufacturers but thinking it through, I wonder if there is sometimes a bit of a mismatch, opposite ends of a tolerance spectrum.

Mick


User currently offlineSilver1SWA From United States of America, joined Mar 2004, 4738 posts, RR: 26
Reply 7, posted (1 year 2 months 2 weeks 4 days 18 hours ago) and read 3753 times:
Support Airliners.net - become a First Class Member!

Quoting dendrobatid (Reply 6):
I wonder if there is sometimes a bit of a mismatch, opposite ends of a tolerance spectrum.

   I've read a couple articles that say that is precisely the issue with today's digital cameras. That's why it's a good idea to have both the camera and lens sent in for calibration.



ALL views, opinions expressed are mine ONLY and are NOT representative of those shared by Southwest Airlines Co.
User currently offlineckw From UK - England, joined Aug 2010, 713 posts, RR: 16
Reply 8, posted (1 year 2 months 2 weeks 4 days 17 hours ago) and read 3746 times:

Quoting Silver1SWA (Reply 7):
I've read a couple articles that say that is precisely the issue with today's digital cameras. That's why it's a good idea to have both the camera and lens sent in for calibration.

This is certainly the case with "soft" issues (eg, exposure and AF). I've been told by Canon that their tolerance on exposure calibration is 1/3rd of a stop. So get unlucky and you could have a lens body combo which is 2/3rds out and still be considered OK.

But this is not new - I remember having quite a bit of trouble getting a pair of EOS 3 film bodies to give me the same exposure with the same lens.

But I think Mick is talking about hardware tolerances. It is entirely possible he is correct but I would also say that manufacturing processes have come on a long way. Intuitively, I would think (as Mick does) that sensor alignment is hypercritical but yet if this were a problem, surely it would be well reported on the various websites that just love that kind of thing. After all, a sensor which is not exactly parallel would surely show fairly obvious OOF on one side or the other of the frame.

Cheers,

Colin



Colin K. Work, Pixstel
User currently offlineSilver1SWA From United States of America, joined Mar 2004, 4738 posts, RR: 26
Reply 9, posted (1 year 2 months 2 weeks 4 days 16 hours ago) and read 3739 times:
Support Airliners.net - become a First Class Member!

Quoting ckw (Reply 8):
But I think Mick is talking about hardware tolerances. It is entirely possible he is correct but I would also say that manufacturing processes have come on a long way. Intuitively, I would think (as Mick does) that sensor alignment is hypercritical but yet if this were a problem, surely it would be well reported on the various websites that just love that kind of thing. After all, a sensor which is not exactly parallel would surely show fairly obvious OOF on one side or the other of the frame.

I'll have to find the articles, but I'm fairly certain they were talking about hardware tolerances and how one can get unlucky and have both camera and lens on opposite side of the tolerance spectrum making your combo a horrible match. If I recall, the point being made in the article was that having micoradjustment capabilities in camera helped alleviate the need to send both camera and lens to the manufacturer, a win for both the photographer and the manufacturer.



ALL views, opinions expressed are mine ONLY and are NOT representative of those shared by Southwest Airlines Co.
Top Of Page
Forum Index

This topic is archived and can not be replied to any more.

Printer friendly format

Similar topics:More similar topics...
Good Non-AF Zoom Lenses For Canon T-60s posted Wed Aug 1 2001 20:56:05 by Western727
7D Vs 5D2 And Corresponding Lenses posted Sun Nov 21 2010 13:04:40 by Buyantukhaa
Canon Rebel XS Zoom Lenses? posted Mon May 25 2009 15:35:58 by Lbox12
New Sigma Telephoto Zoom Lenses With OS posted Wed Jan 30 2008 21:20:43 by F14
Fun Applications Of Ultra Zoom Lenses posted Fri Jan 25 2008 16:57:20 by Jawed
Canon 350D And D70 Lenses posted Sun Nov 5 2006 00:03:08 by Jspitfire
What Wide Angle/standard Zoom Lenses Do You Use? posted Sat Feb 18 2006 23:03:44 by LGW
Thoughts On The Nikon 80-400VR And Other Lenses? posted Tue Mar 22 2005 23:10:29 by AndyHunt
Converters Vs Zoom Lenses posted Sat Jan 8 2005 09:29:39 by Bell407
Good Quality Zoom Lenses Canon posted Thu Jul 1 2004 00:17:14 by MNeo