June 23, 2018, 04:16:06 PM

Author Topic: Never let it be said...  (Read 16945 times)

Eldar

  • Canon EF 400mm f/2.8L IS II
  • *********
  • Posts: 3241
    • Flickr
Re: Never let it be said...
« Reply #15 on: September 15, 2013, 02:48:18 AM »
Apart from the 24-70 f2.8L II, I have had to do AFMA on all my other lenses on all bodies, some more than +/-10. I suspect that you would benefit from buying/borrowing a proper tool, because I cannot believe you have been that lucky. But on the other hand ... some also win the big prices in lotteries ...  :-\

canon rumors FORUM

Re: Never let it be said...
« Reply #15 on: September 15, 2013, 02:48:18 AM »

johnf3f

  • EOS 5DS R
  • ******
  • Posts: 867
  • Canon 1Dx
Re: Never let it be said...
« Reply #16 on: September 22, 2013, 03:55:00 PM »
I think I might be missing something in the AFMA process. So you you focus on some asses with AF targets nearby at a 45 degree angle?  :o
[/quote]


OOPS! Whiskey induced dyslexia or wishful thinking - you decide!
Canon 1DX, 7D2, 16-35 F4 L IS, 24-70 F2.8 V2, 100 F2.8 Macro, 100-400 L IS Mk2, 300 F2.8 L IS, 800 F5.6 L IS, Holga Pinhole lens.

Sporgon

  • Canon EF 400mm f/2.8L IS II
  • *********
  • Posts: 3589
  • 5% of gear used 95% of the time
    • www.buildingpanoramics.com
Re: Never let it be said...
« Reply #17 on: September 22, 2013, 04:13:55 PM »
I thought it was inaccurate to have a target at 45* and it should always be a perpendicular, flat target, otherwise you can't be sure where the AF is picking up on the 'slope'.

johnf3f

  • EOS 5DS R
  • ******
  • Posts: 867
  • Canon 1Dx
Re: Never let it be said...
« Reply #18 on: September 23, 2013, 01:50:57 PM »
I thought it was inaccurate to have a target at 45* and it should always be a perpendicular, flat target, otherwise you can't be sure where the AF is picking up on the 'slope'.

I have not found this to be an issue, as the focus point is very clear and the idea of the slope (and distance scale) is to show up any error. However, if it concerns you, then there are targets available that utilise a vertical target combined with a 45 degree distance scale.
Canon 1DX, 7D2, 16-35 F4 L IS, 24-70 F2.8 V2, 100 F2.8 Macro, 100-400 L IS Mk2, 300 F2.8 L IS, 800 F5.6 L IS, Holga Pinhole lens.

bycostello

  • EOS 5DS R
  • ******
  • Posts: 904
    • London Weddings
Re: Never let it be said...
« Reply #19 on: September 24, 2013, 03:38:57 AM »
I have tried all my 9 Canon AF lenses (ranging from 17 to 800mm) and ended up with micro adjustment at or very near 0 on my1D4, they also focus just fine on my other EOS cameras (5Dc, EOS3, EOS 33V and EOS 50E) am I just lucky?
I am not questioning that micro AF adjustment is not a great confidence builder, just saying that I have not found a use for it - except for confidence building!

If you were using a yard stick and duck tape like the OP then yes you are lucky.




 ;D ;D ;D

johnf3f

  • EOS 5DS R
  • ******
  • Posts: 867
  • Canon 1Dx
Re: Never let it be said...
« Reply #20 on: September 24, 2013, 08:25:25 PM »
I have tried all my 9 Canon AF lenses (ranging from 17 to 800mm) and ended up with micro adjustment at or very near 0 on my1D4, they also focus just fine on my other EOS cameras (5Dc, EOS3, EOS 33V and EOS 50E) am I just lucky?
I am not questioning that micro AF adjustment is not a great confidence builder, just saying that I have not found a use for it - except for confidence building!

If you were using a yard stick and duck tape like the OP then yes you are lucky.




 ;D ;D ;D


Maybe so, but having said that within the small circle of photographers I know well (including 1 pro) none of us have used this facility on any of our lenses to the best of my knowledge. The only Nikon shooter I know well has not needed to use it either. The cameras in question are a 40D, 5Dc and half a dozen 1D Mk4s, lenses are nearly all Canon (mainly L series) from 16mm to 800mm, sorry I can't remember the details of the 2 bodies that our Nikonian friend uses. I have also had no AF accuracy issues with my EOS 3, EOS 33V and EOS 50E - I must be really lucky!
P.S. the only camera I have used the Yardstick method with was my Leica IIIg - still spot on (within 1/2 inch or less at 6 feet) after all these years  with all 3 lenses and has never been serviced.
Canon 1DX, 7D2, 16-35 F4 L IS, 24-70 F2.8 V2, 100 F2.8 Macro, 100-400 L IS Mk2, 300 F2.8 L IS, 800 F5.6 L IS, Holga Pinhole lens.

takesome1

  • EOS 5DS R
  • ******
  • Posts: 1190
Re: Never let it be said...
« Reply #21 on: September 24, 2013, 10:58:40 PM »
I have tried all my 9 Canon AF lenses (ranging from 17 to 800mm) and ended up with micro adjustment at or very near 0 on my1D4, they also focus just fine on my other EOS cameras (5Dc, EOS3, EOS 33V and EOS 50E) am I just lucky?
I am not questioning that micro AF adjustment is not a great confidence builder, just saying that I have not found a use for it - except for confidence building!

If you were using a yard stick and duck tape like the OP then yes you are lucky.




 ;D ;D ;D


Maybe so, but having said that within the small circle of photographers I know well (including 1 pro) none of us have used this facility on any of our lenses to the best of my knowledge. The only Nikon shooter I know well has not needed to use it either. The cameras in question are a 40D, 5Dc and half a dozen 1D Mk4s, lenses are nearly all Canon (mainly L series) from 16mm to 800mm, sorry I can't remember the details of the 2 bodies that our Nikonian friend uses. I have also had no AF accuracy issues with my EOS 3, EOS 33V and EOS 50E - I must be really lucky!
P.S. the only camera I have used the Yardstick method with was my Leica IIIg - still spot on (within 1/2 inch or less at 6 feet) after all these years  with all 3 lenses and has never been serviced.

Of course this quote was just a rehash of an earlier post you responded to.

The thing is that the ruler method is not very accurate, it will probably get most people close enough.

Even the other methods like Lens Align and Reiken Focal are only accurate to a certain point.
Both will give you a value at 25x or 50x the distance from the target (or the distance you pick).
You will know what your values are at the light level you use.

With lens align if you just use it once, and do not repeat your tests several times and get the exact same results you will not know if you are getting accurate results.
What you find out if you do enough testing with Lens Align (Not Reiken because it will just fail the test at low light levels) is if you repeat the exact same test and vary the lighting at different levels you will get different results. Do this and you are developing an understanding of how your AF system is performing.

One thing you find is that often a lens can be of +/- quit a few points before you may notice. For instance I have a 300mm f/2.8 IS that is off by +5 on every body I have. Without any adjustment I can get a 80% keeper rate with my 1D IV. Adjust the AFMA it gets about a 97% keeper rate.

I could pick up any of my lenses and put them on my 1D IV or 5D with no adjustments and get a respectable keeper rate. A few of the L lenses I have I can tweak the adjustment a bit and get the keeper rate up.

I find the reason for doing the AFMA is just increasing the keeper rate up by a few percent. It is not correcting a lens that is so far off it doesn't take a decent picture. If a lens is that bad I am shipping it to Canon.


canon rumors FORUM

Re: Never let it be said...
« Reply #21 on: September 24, 2013, 10:58:40 PM »

johnf3f

  • EOS 5DS R
  • ******
  • Posts: 867
  • Canon 1Dx
Re: Never let it be said...
« Reply #22 on: September 25, 2013, 09:40:17 PM »
I have tried all my 9 Canon AF lenses (ranging from 17 to 800mm) and ended up with micro adjustment at or very near 0 on my1D4, they also focus just fine on my other EOS cameras (5Dc, EOS3, EOS 33V and EOS 50E) am I just lucky?
I am not questioning that micro AF adjustment is not a great confidence builder, just saying that I have not found a use for it - except for confidence building!

If you were using a yard stick and duck tape like the OP then yes you are lucky.




 ;D ;D ;D


Maybe so, but having said that within the small circle of photographers I know well (including 1 pro) none of us have used this facility on any of our lenses to the best of my knowledge. The only Nikon shooter I know well has not needed to use it either. The cameras in question are a 40D, 5Dc and half a dozen 1D Mk4s, lenses are nearly all Canon (mainly L series) from 16mm to 800mm, sorry I can't remember the details of the 2 bodies that our Nikonian friend uses. I have also had no AF accuracy issues with my EOS 3, EOS 33V and EOS 50E - I must be really lucky!
P.S. the only camera I have used the Yardstick method with was my Leica IIIg - still spot on (within 1/2 inch or less at 6 feet) after all these years  with all 3 lenses and has never been serviced.

Of course this quote was just a rehash of an earlier post you responded to.

The thing is that the ruler method is not very accurate, it will probably get most people close enough.

Even the other methods like Lens Align and Reiken Focal are only accurate to a certain point.
Both will give you a value at 25x or 50x the distance from the target (or the distance you pick).
You will know what your values are at the light level you use.

With lens align if you just use it once, and do not repeat your tests several times and get the exact same results you will not know if you are getting accurate results.
What you find out if you do enough testing with Lens Align (Not Reiken because it will just fail the test at low light levels) is if you repeat the exact same test and vary the lighting at different levels you will get different results. Do this and you are developing an understanding of how your AF system is performing.

One thing you find is that often a lens can be of +/- quit a few points before you may notice. For instance I have a 300mm f/2.8 IS that is off by +5 on every body I have. Without any adjustment I can get a 80% keeper rate with my 1D IV. Adjust the AFMA it gets about a 97% keeper rate.

I could pick up any of my lenses and put them on my 1D IV or 5D with no adjustments and get a respectable keeper rate. A few of the L lenses I have I can tweak the adjustment a bit and get the keeper rate up.

I find the reason for doing the AFMA is just increasing the keeper rate up by a few percent. It is not correcting a lens that is so far off it doesn't take a decent picture. If a lens is that bad I am shipping it to Canon.

Thanks for your insight! To date I have not had problems with AF accuracy, but I do have problems with AF acquisition and speed - probably just like the rest of us.
However I will have a look at some of your suggestions - I will be more than happy to eat my hat if I get better performance!
Canon 1DX, 7D2, 16-35 F4 L IS, 24-70 F2.8 V2, 100 F2.8 Macro, 100-400 L IS Mk2, 300 F2.8 L IS, 800 F5.6 L IS, Holga Pinhole lens.

BozillaNZ

  • EOS M5
  • ****
  • Posts: 198
Re: Never let it be said...
« Reply #23 on: September 25, 2013, 11:56:16 PM »
Well all I can say is the standard of judging focus is different to each individual.

When I test a lens, I want it to be spot on wide open (f1.4 for my primes, f2.8 for zooms), using those focus points on my 1Ds3:

top most, bottom most, left most, right most, top-left corner, top-right corner, bottom-right corner, bottom-left corner, center

Also it has to be spot on at near MFD, 1m, 10m, 50m, infinity.

If any of those focus 'use cases' results in a consistent front/back focus, I will have to find ways to fix that <- which includes using hex wrench to adjust the mirror box of the camera in two points + AFMA, some times even a lens disassembly.

And believe me, the chance of this state of adjustment happens on all your camera body and lens without any human intervention is about 0.

Now my 24 1.4L, S50 1.4 and 70-200 2.8 II are all tuned up like that, so when I shoot a event, I can just chose any AF points at any distance and get a spot on image for about 85% of time. Anything off will be my own technique or subject motion to blame.
« Last Edit: September 26, 2013, 12:01:51 AM by BozillaNZ »
Look ma, me cameraz can push shadow 10 stoops w/o noizes, OMGWTFBBQChickenwingHaxorz!!!11
www.flickr.com/photos/bozillanz/

bratkinson

  • PowerShot G7 X Mark II
  • **
  • Posts: 14
Re: Never let it be said...
« Reply #24 on: October 11, 2013, 07:37:15 AM »
One of the lesser reasons I upgraded from a 60D to 5D3 was to get the MFA capability.  Although I thought 4 of the 5 L lenses I had were 'right on' mounted on the 60D and 5D3, 3 of them were +3 or so.  The real surprise was the used 80-200 f2.8L 'magic drainpipe'  I found in ebay with a 1993 build date was right on!!  As for the 'other' L, my do-everything 24-105, it was noticebly soft on the 60D, and barely soft on the 5D3, but it took +3 (24) and +5 (105) to really make it knock-out sharp!

chauncey

  • EOS 7D Mark II
  • *****
  • Posts: 564
Re: Never let it be said...
« Reply #25 on: October 11, 2013, 12:21:51 PM »
I doesn't take a genius to use the ruler method if you're adept at Photoshop focus stacking.
Simply take a series of images at incremented adjustments>import them into PS as layers and stack them...the masks will tell you where you ought to be.       ;)
The things you do for yourself die with you, the things you do for others live forever.
 A man's worth should be judged, not when he basks in the sun, but how he faces the storm.
http://1x.com/member/chauncey43

Eldar

  • Canon EF 400mm f/2.8L IS II
  • *********
  • Posts: 3241
    • Flickr
Re: Never let it be said...
« Reply #26 on: October 11, 2013, 02:13:57 PM »
I doesn't take a genius to use the ruler method if you're adept at Photoshop focus stacking.
Simply take a series of images at incremented adjustments>import them into PS as layers and stack them...the masks will tell you where you ought to be.       ;)
+1

canon rumors FORUM

Re: Never let it be said...
« Reply #26 on: October 11, 2013, 02:13:57 PM »