April 21, 2014, 07:27:55 AM

Author Topic: New MFA method  (Read 17525 times)

Mt Spokane Photography

  • 1D X
  • *******
  • Posts: 7714
    • View Profile
Re: New MFA method
« Reply #15 on: February 04, 2013, 06:32:24 PM »
Liveview AF (contrast Detect), at least on the 5D MK II and 7D  has turned out to be inaccurate and inconsistent.  The focus method described was one I used a couple of years back, its nothing new.  Unfortunately, it is unreliable because it assumes thet Contrest Dect AF is accurate, but it has been shown to have errors, so much that FoCal had to change their software to accomodate the errors.
 
The most accurate methods still need about 10 shots setting the lens AF back to infinity or MFD and then throwing out obvious errors in focusing and averaging the others.  Then, most shots will be in focus, but there will always be a few outliers. The 6D seem to have more accurate CD focus, the 5D MK II and 7D tends to be worse.  The figures for the 5D MK III are not in yet.
http://www.reikan.co.uk/focalweb/index.php/2012/12/af-consistency-comparison-nikon-canon-phase-detect-contrast-detect/
 
I'd avoid this method if you have a 5D MK II or a 7D.  FoCal which allows you to take a large number of shots at each AFMA setting is a great tool for ignoring obvious AF errors and picking the correct setting that gives the highest number of in focus shots.
 

canon rumors FORUM

Re: New MFA method
« Reply #15 on: February 04, 2013, 06:32:24 PM »

horshack

  • PowerShot G16
  • **
  • Posts: 43
    • View Profile
Re: New MFA method
« Reply #16 on: February 04, 2013, 06:58:01 PM »
Hi everyone,

This is horshack (dpreview) / snapsy (FM), the guy who developed this AF tune technique. Don't get hung up on Live View - the fact that LV is used to establish critical focus (step 1) is only incidental to the process. You can use whatever focus method you want, including viewfinder AF or MF, LV contrast AF or MF...doesn't matter, as long as you establish and verify critical focus prior to starting the other steps.

What's unique about this technique is that you're using the confirmation dot and scaling the AF tune values to find the range and ultimate midpoint (final) AF tune value, all while keeping the same critical focus you established in step #1 (never refocusing). This avoids the mechanical AF variability that you get with all other MFA methods, since those involve reacquiring PDAF to verify focus, which in turn require multiple PDAF acquisitions/check cycles to normalize the PDAF variability out of the process.
« Last Edit: February 04, 2013, 07:01:03 PM by horshack »

neuroanatomist

  • CR GEEK
  • *******
  • Posts: 12793
    • View Profile
Re: New MFA method
« Reply #17 on: February 04, 2013, 07:26:32 PM »
Thanks, Mt. Spokane - I had read those data, and forgotten about them.

Hi everyone,

This is horshack (dpreview) / snapsy (FM), the guy who developed this AF tune technique. Don't get hung up on Live View - the fact that LV is used to establish critical focus (step 1) is only incidental to the process. You can use whatever focus method you want, including viewfinder AF or MF, LV contrast AF or MF...doesn't matter, as long as you establish and verify critical focus prior to starting the other steps.

I wouldn't call it incidental.  As you point out, establishing 'critical focus' is, well...critical.  If LV is inconsistent, as the data indicate, there's no simple way to establish that baseline, certainly not with a single shot. Have you tested repeatability - several rounds of a single LV focus, resulting in the same selected AFMA value every time?

Have you, for example, used your method and selected an AFMA value 'on the edge' (one that gives confirmation where the next value out does not) then re-checked that value multiple times and seen confirmation every time, then the next value out and observed a lack of confirmation every time?

Thanks!
EOS 1D X, EOS M, and lots of lenses
______________________________
Flickr | TDP Profile/Gear List

Plamen

  • Canon AE-1
  • ***
  • Posts: 78
    • View Profile
    • Math and Photography
Re: New MFA method
« Reply #18 on: February 04, 2013, 08:47:18 PM »
Liveview AF (contrast Detect), at least on the 5D MK II and 7D  has turned out to be inaccurate and inconsistent.

The accuracy of LV depends hugely on the light level. Without knowing that, I would take those tests with a grain of salt.

My own experience with the 5D2 and the 100L, LV, is as perfect focus as it gets, even in artificial but not too low light. In low light, it is unreliable or it just cannot focus when PD has no problems.

horshack

  • PowerShot G16
  • **
  • Posts: 43
    • View Profile
Re: New MFA method
« Reply #19 on: February 04, 2013, 08:49:10 PM »

Hi everyone,

This is horshack (dpreview) / snapsy (FM), the guy who developed this AF tune technique. Don't get hung up on Live View - the fact that LV is used to establish critical focus (step 1) is only incidental to the process. You can use whatever focus method you want, including viewfinder AF or MF, LV contrast AF or MF...doesn't matter, as long as you establish and verify critical focus prior to starting the other steps.


I wouldn't call it incidental.  As you point out, establishing 'critical focus' is, well...critical.  If LV is inconsistent, as the data indicate, there's no simple way to establish that baseline, certainly not with a single shot. Have you tested repeatability - several rounds of a single LV focus, resulting in the same selected AFMA value every time?

Have you, for example, used your method and selected an AFMA value 'on the edge' (one that gives confirmation where the next value out does not) then re-checked that value multiple times and seen confirmation every time, then the next value out and observed a lack of confirmation every time?

Thanks!


If by LV being inconsistent you mean its CDAF then sure it's by no means infallible, although it is rather precise, both in my experience and in Roger's LensRental tests. But again you don't have to rely on its consistency because you can verify critical focus visually at 10x and adjust manually if necessary before proceeding with my method. Or focus entirely manually if you prefer. If by inconsistent you mean that LV @ 10x isn't sufficient for visually confirming critical focus, then that I would disagree with that. But you can always take a photo to confirm focus if you don't feel LV can be trusted.

As for the repeatability of my technique, so far it's been very consistent in my tests, to within one or two AF tune units across iterations, when using the same target, focus distance, and lighting conditions between iterations. If you change any of these factors then you may certainly get different results, but those differences will arise from the sensitivity of the PD phase sensing mechanism itself rather than from this specific technique (ie, you'd get the same AF tune deltas using other MFA techniques in those situations).

As for the accuracy of the technique, I developed and tested it first on a D800 with 5 different lenses, all of which produced AF tune values equal or better than what I achieved via LensAlign. I have also tested it on a 5DM3 with two lenses so far, with the same level of accuracy. On my original dpreview for the Nikon version of the technique (link: http://forums.dpreview.com/forums/post/50774257), everyone who has tried it so far has achieved similar levels of accuracy. I think a few have compared it to Focal as well.

TrumpetPower!

  • 1D Mark IV
  • ******
  • Posts: 951
    • View Profile
Re: New MFA method
« Reply #20 on: February 04, 2013, 08:55:23 PM »
If LV is inconsistent, as the data indicate(....)

Is it actual live view that you've found to be inconsistent (that is, not an accurate representation of what the sensor captures), or is it live view slow mode autofocus that's inconsistent?

If it's the latter, I can easily believe that...on more than one occasion when doing tripod work, I've improved upon the autofocus. For most stuff it's "good enough," but it's not the gold standard.

If it's the former, though...well, that'd be a huge problem, actually....

b&

Mt Spokane Photography

  • 1D X
  • *******
  • Posts: 7714
    • View Profile
Re: New MFA method
« Reply #21 on: February 04, 2013, 11:53:59 PM »
CD focus using liveview is what is inconsistent (on some bodies, but not all).  That green dot is part of the CD focus system, so it is suspect as well.
The liveview image you see is right off the sensor, and is the image you are getting, so you can use 10X and focus fairly well using the rear LCD.  Its better if you tether to a large monitor though.

canon rumors FORUM

Re: New MFA method
« Reply #21 on: February 04, 2013, 11:53:59 PM »

East Wind Photography

  • 6D
  • *****
  • Posts: 617
  • EWP
    • View Profile
Re: New MFA method
« Reply #22 on: February 05, 2013, 09:45:21 PM »
But I think what he refers to is the green dot in PD mode not in CD mode.  I agree LV does not always give you critical focus but it's a starting point and you can tweak manually from there.  Here the use of a lenscal target, newsprint, or something like that would help get that LV 10X shot in perfect focus.  So the question is the AF Detect system consistently flakey or just plain flakey?  I tend to think it's consistently flakey and should work well in this case since you just trying to get the range and not an absolute.

We'll see, once I get over this flu I plan to try it out on my 300 2.8L and 600 F4L.  It cant be any worse than focal in it's current state and if I dont have to carry my test setup out to a football field then it's worth it.

CD focus using liveview is what is inconsistent (on some bodies, but not all).  That green dot is part of the CD focus system, so it is suspect as well.
The liveview image you see is right off the sensor, and is the image you are getting, so you can use 10X and focus fairly well using the rear LCD.  Its better if you tether to a large monitor though.

digital paradise

  • EOS M2
  • ****
  • Posts: 284
    • View Profile
    • Zenon Char Photography
Re: New MFA method
« Reply #23 on: February 06, 2013, 01:40:13 AM »
That is the way understand it. Establish focus in LV, shut it and the AF off. In the viewfinder look for the green AF confirmation dot in the bottom right corner when half pressing the shutter. Located there on my 5D3 anyway.

I just added to my post. The AF grid on the focus screen which is red lights up as well.   
« Last Edit: February 06, 2013, 01:42:33 AM by digital paradise »

Mehmetski

  • PowerShot G16
  • **
  • Posts: 20
    • View Profile
Re: New MFA method
« Reply #24 on: February 06, 2013, 04:30:39 AM »
I will try this for sure

junkwerks

  • PowerShot G16
  • **
  • Posts: 22
    • View Profile
Re: New MFA method
« Reply #25 on: February 06, 2013, 11:57:19 AM »
So has anybody done the comparisons with the OP's suggested method and FoCal or similar?  Can it be quantified in useful numbers as well as subjective picture examinations? 

digital paradise

  • EOS M2
  • ****
  • Posts: 284
    • View Profile
    • Zenon Char Photography
Re: New MFA method
« Reply #26 on: February 10, 2013, 01:02:59 PM »

digital paradise

  • EOS M2
  • ****
  • Posts: 284
    • View Profile
    • Zenon Char Photography
Re: New MFA method
« Reply #27 on: February 10, 2013, 01:06:47 PM »
Read post #15

http://www.fredmiranda.com/forum/topic/1187247/2


Post #14 but that is just someones opinion. Not saying it is not accurate but there is no test data. Fred Miranda seems to think it is pretty good. It is not like the author or FM are making any money on it.

http://www.fredmiranda.com/forum/topic/1187247/2

canon rumors FORUM

Re: New MFA method
« Reply #27 on: February 10, 2013, 01:06:47 PM »

horshack

  • PowerShot G16
  • **
  • Posts: 43
    • View Profile
Re: New MFA method
« Reply #28 on: February 10, 2013, 01:50:39 PM »
Btw I've decided to put a name to the technique so that everyone has an easy way to refer to it and to make it web searchable as it evolves with any improvements. I'm calling it "DotTune", since the focus confirmation dot is central to the technique. I'm also working on a youtube video to bring it to a wider audience.

caruser

  • Rebel T5i
  • ****
  • Posts: 115
    • View Profile
Re: New MFA method
« Reply #29 on: February 10, 2013, 02:41:07 PM »
Tried this yesterday evening, and except for being a bit shocked for how much adjustment my shiny new 1DX needed (only the 24-70@24 came out at 0) it actually worked pretty well, that is when I double-checked some of the obtained values they were spot on...

canon rumors FORUM

Re: New MFA method
« Reply #29 on: February 10, 2013, 02:41:07 PM »