Home · Register · Join Upload & Sell

Moderated by: Fred Miranda
Username  

  New fredmiranda.com Mobile Site
  New Feature: SMS Notification alert
  New Feature: Buy & Sell Watchlist
  

FM Forums | Canon Forum | Join Upload & Sell

1       2       3      
4
       5              37       38       end
  

Dot-Tune: Autofocus Fine Tuning in under 5 minutes

  
 
snapsy
Offline
• • • • • •
Upload & Sell: On
p.4 #1 · p.4 #1 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes


Zenon Char wrote:
I guess the old keeping the target parallel to the sensor is still important. I have lens align which does a great job but may be to small to get LV focus accurate.

I may have missed this. What is everyone doing here? I just turned on my camera level and spin it 360 degrees until it was level in all directions. No guarantee my wall is level to that degree of requirement.



Yep, parallel target setup is still important because otherwise we might tune to a biased reference point. A LensAlign target is also still useful for this technique, since it provides the ability for us to decide the precise balance of front/rear DOF - you can decide this balance in step #1 by focusing so that there's more DOF in front or back of the parallel target, suited to your personal preference and needs.



Feb 10, 2013 at 01:33 PM
Zenon Char
Offline
• • • • • •
Upload & Sell: On
p.4 #2 · p.4 #2 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes


I meant using the red target to get the target parallel. I never really liked the ruler and would not use it with your method. Again always different readings. I guess even if the LV AF square is larger than the Lens Align itself you just have to verify that the target face of lens align is in focus. At one time I taped a bill to lens align and used that as a visual reference when using the ruler.

I'll give the lens align target a shot.



Feb 10, 2013 at 02:20 PM
jrsforums
Offline
• • •
Upload & Sell: Off
p.4 #3 · p.4 #3 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes


Tag


Feb 10, 2013 at 03:22 PM
docnlaw24
Offline
• • •
Upload & Sell: Off
p.4 #4 · p.4 #4 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes


tag


Feb 10, 2013 at 07:01 PM
KKFung
Offline
• • •
Upload & Sell: Off
p.4 #5 · p.4 #5 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes


Yes I parallel the sensor for the test also, I think it is very basic set up


Feb 11, 2013 at 08:44 AM
petefromzim
Offline

Upload & Sell: Off
p.4 #6 · p.4 #6 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes


snapsy wrote:
6. Calculate your final AF tune value, which is the midpoint of the range you established in steps 4 and 5. For example if your range was -2 to +8 then the midpoint is +3. If your range was 0 to +12 then the midpoint is +6. Set your AF tune value to your calculated midpoint. You're done.



My first post here on FM.

I may be being a bit dumb here but..... In the example of range -2 to +8 quoted, surely the mid-point is +3.5 because the zero (0) position is not being taken into account.

I have previously been AFMA-ing my lenses using the method here:
http://arihazeghiphotography.com/MA-web/
With snapsy's method, I'm getting totally different results. Very confused - maybe the tropical sun is addling my brain....

Pete



Feb 11, 2013 at 09:50 AM
snapsy
Offline
• • • • • •
Upload & Sell: On
p.4 #7 · p.4 #7 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes


petefromzim wrote:
My first post here on FM.

I may be being a bit dumb here but..... In the example of range -2 to +8 quoted, surely the mid-point is +3.5 because the zero (0) position is not being taken into account.

I have previously been AFMA-ing my lenses using the method here:
http://arihazeghiphotography.com/MA-web/
With snapsy's method, I'm getting totally different results. Very confused - maybe the tropical sun is addling my brain....

Pete


Welcome to FM! The (0) position is being taken into account in my examples. The 3.5 you calculated for -2 to +8 is the average for the number of values added to the start point, but what you want is the midpoint (median), which will always be a whole number when the number of values is odd (-2 to +8 has 11 values, 0 to +6 has 7 values). When the number of values is even then the midpoint will have have a fraction, and you can round either up or down to the next whole AF tune value.



Feb 11, 2013 at 10:59 AM
Zenon Char
Offline
• • • • • •
Upload & Sell: On
p.4 #8 · p.4 #8 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes


Welcome to FM forums Pete. I tried that method as well. I'm not sure if it is me but I can't work with any method that I have to view and make a decision. It s not that hard but I'm always indecisive, test again and get different results. I want the method to tell me what to do. It is similar to being married


Feb 11, 2013 at 11:10 AM
mttran
Offline
• • • • •
Upload & Sell: Off
p.4 #9 · p.4 #9 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes


Thank you for sharing this Adam. Currently i rely on AF then Eg-S if AF failed, I will try your "DotTune" to minimize the manual process.


Feb 11, 2013 at 11:18 AM
GammyKnee
Offline
• •
Upload & Sell: Off
p.4 #10 · p.4 #10 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes


Finally got round to adjusting another of my existing lenses for my recently acquired 5DII and decided to give this technique a try.

I liked it! As I approached the end of each range the AF did go through a bit of indecision before it finally gave up the ghost and declared OOF but I just dutifully followed the procedure. The result I got appears to work well. At the target distance the PD AF now (mostly) matches CF in liveview.

It was very nice to be freed from all the repetitive shot-taking and comparison. Excluding set up time (which was quick in itself) the procedure only took a couple of minutes vs. half an hour or more of tedium and indecision. So thank you Snapsy - this is now my preferred adjustment technique.



Feb 11, 2013 at 01:09 PM
 


Search in Used Dept. 

Wilba
Offline

Upload & Sell: Off
p.4 #11 · p.4 #11 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes


Hello snapsy. Congratulations on your method! I don't have a camera with which I can test it, but I know from having done a lot of similar testing that it will work as advertised.

I wanted to share a few ideas that I haven't seen mentioned in this thread. PD AF is proven to be a closed-loop process, that ends when the AF sensor sees an in-focus subject. I'm not sure whether you are aware of the Best Focus Correction Value (Canon term). Check out Busted! The Myth of Open-loop Phase-detection Autofocus for a discussion which covers both. (You're welcome to PM me on DPReview if you like.)

The thing you're playing with is what Doug Kerr and I call the Acceptance Zone - the range over which a lens in MF will confirm focus using PD. I've played with that a lot by changing the camera-to-subject distance, and as you found you can also determine it with AF microadjustment.

It gets interesting when the closed-loop process is trying to put the focus within the AZ. In all of my testing (mainly with Canon), PD AF clusters the focus at the boundaries of the AZ, not randomly within it as you might expect. For a lens like my copy of the notoriously unreliable Canon 50/1.8, the AZ is about 3mm deep near the MFD, which can mean that one end is consistently spot-on and the other is outside the f/2.8 DOF (4mm). On some bodies, which end you get depends on the initial condition (focus in-front of or behind the subject), and you usually get the end on the other side, as if focus is confirmed on leaving the AZ.

PD AF is more variable than CD, but if you are able to control for the end-of-the-AZ-effect (for instance, if you have lens+body combination which always finds the same end from the same initial gross misfocus), then you can prove that even with a lens like the 50/1.8, the focus mechanism is capable of tremendous accuracy, like better than 5% of the thinnest DOF, which is 3% of the AZ, and 0.02% of the MFD.

So I'm convinced that the focus process within the lens is accurate enough that it's irrelevant to overall focus accuracy. Therefore, what matters most is how the AF sensor sees the world through a particular lens, and how the AF logic in the camera is able to use that information to determine when the two images align.

Your method uses the middle of the AZ, which will be fine as long as the AZ is significantly narrower than the DOF. Let me know if any of that doesn't make sense, or helps you understand what you're doing.



Feb 16, 2013 at 02:48 AM
snapsy
Offline
• • • • • •
Upload & Sell: On
p.4 #12 · p.4 #12 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes


Wilba wrote:
Hello snapsy. Congratulations on your method! I don't have a camera with which I can test it, but I know from having done a lot of similar testing that it will work as advertised.


Hi Wilba, I read your article when it was first posted. It was a tremendous contribution to the photographic community in furthering our understanding of PDAF. It was the idea that PDAF is closed-loop which partially lead me to conclude that AF tune has be an input bias (bias to the PDAF sensing mechanism) rather than an output bias (bias to the lens movement command), since any output bias would be lost and rendered inconsequential between iterations of a closed-loop system. Here are some links of my posts on dpreview which discuss the technical basis of DotTune (I'm "horshack" on dpreview):

Reverse Engineering AF Tune
Why AF mistune looks like inconsistency
What AF tune addresses



Feb 16, 2013 at 12:10 PM
RonMJ
Offline

Upload & Sell: Off
p.4 #13 · p.4 #13 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes


Has anyone tried this multiple times with the same lens, WITHOUT removing the lens from the tripod? same target, same lighting, same everything. I just did that with my Canon 85 1.8. The results? 0 to -17 (18 setting spread); -3 to -19 (17 settings); +1 to -16 (18 settings) and -5 to -20 (ran out of settings). Four different MA values. Again, nothing changed between runs. Screwy lens? Or is DotTune just as variable as any of the other methods that seem so variable?

I used the Live (CDAF setting) in Live View for step #1.



Feb 16, 2013 at 05:36 PM
snapsy
Offline
• • • • • •
Upload & Sell: On
p.4 #14 · p.4 #14 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes


RonMJ wrote:
Has anyone tried this multiple times with the same lens, WITHOUT removing the lens from the tripod? same target, same lighting, same everything. I just did that with my Canon 85 1.8. The results? 0 to -17 (18 setting spread); -3 to -19 (17 settings); +1 to -16 (18 settings) and -5 to -20 (ran out of settings). Four different MA values. Again, nothing changed between runs. Screwy lens? Or is DotTune just as variable as any of the other methods that seem so variable?

I used the Live (CDAF setting) in Live View for step #1.


Were you refocusing for each of the four ranges you arrived at or using the same initial focus for all? If the former then the difference might be tiny differences in focus - the midpoints for your ranges are relatively close (-8.5, -11, -7.5, -12.5). An another possibility is slight differences in how you're interpreting the feedback in the viewfinder for each cycle; try holding the shutter for 5 seconds at each AF tune value (for more concrete confirmation); you may also want to do a few of these 5-second holds at each value. Please report back if this changes your results. Thanks!



Feb 16, 2013 at 05:47 PM
RonMJ
Offline

Upload & Sell: Off
p.4 #15 · p.4 #15 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes


snapsy wrote:
Were you refocusing for each of the four ranges you arrived at or using the same initial focus for all? If the former then the difference might be tiny differences in focus - the midpoints for your ranges are relatively close (-8.5, -11, -7.5, -12.5). An another possibility is slight differences in how you're interpreting the feedback in the viewfinder for each cycle; try holding the shutter for 5 seconds at each AF tune value (for more concrete confirmation); you may also want to do a few of these 5-second holds at each value. Please report back if this changes
...Show more

1) I 'unfocused' the lens between each run.

2) I did not hold the shutter button for 5 seconds waiting for a confirmation or not. You wrote earlier in this post (page 3, #7, first paragraph of your reply) that if there was a delay then that meant it was likely not a confirmed autofocus value. I would press the shutter and if I did not get a quick focus confirm, then I would use the previous value.

3) I just tried this on my 135L. Results? +9 to -7, +8 to -7, and +6 to -9. The FoCal result was -9.

4) Intuitively, such a large spread in values when focusing a fixed focal length at a fixed distance implies a significant lack of precision. I could understand say -3 to -7, but +8 to -7? I dunno.




Feb 16, 2013 at 06:06 PM
snapsy
Offline
• • • • • •
Upload & Sell: On
p.4 #16 · p.4 #16 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes


RonMJ wrote:
2) I did not hold the shutter button for 5 seconds waiting for a confirmation or not. You wrote earlier in this post (page 3, #7, first paragraph of your reply) that if there was a delay then that meant it was likely not a confirmed autofocus value. I would press the shutter and if I did not get a quick focus confirm, then I would use the previous value.

Any delay between the shutter half-press and confirmation dot is an immediate disqualification of confirmation for that tune value. It's when there's no delay that I then suggest you keep holding the shutter for 5 seconds to see if the confirmation dot remains solidly lit. I also suggest additional 5-second half-presses at the same value when you're near either end of the tune range, to better confirm the feedback.



Feb 16, 2013 at 10:00 PM
Wilba
Offline

Upload & Sell: Off
p.4 #17 · p.4 #17 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes


snapsy wrote:
Hi Wilba, I read your article when it was first posted. It was a tremendous contribution to the photographic community in furthering our understanding of PDAF.


Thanks very much. Still is!

I don't agree with, "the need for AF tune stems entirely from static optical path/machined tolerances and not from any mechanical element in the system" (What AF tune addresses), since it seems to refer only to body calibration and seems to ignore Best Focus Correction Value (see page VI-7 of Canon's EF 50mm 1.8 Service Manual). How do you incorporate BFCV, and the idea of lens calibration in general, into your thinking?



Feb 17, 2013 at 02:35 AM
snapsy
Offline
• • • • • •
Upload & Sell: On
p.4 #18 · p.4 #18 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes


Wilba wrote:
Thanks very much. Still is!

I don't agree with, "the need for AF tune stems entirely from static optical path/machined tolerances and not from any mechanical element in the system" (What AF tune addresses), since it seems to refer only to body calibration and seems to ignore Best Focus Correction Value (see page VI-7 of Canon's EF 50mm 1.8 Service Manual). How do you incorporate BFCV, and the idea of lens calibration in general, into your thinking?


My assertion first is based on the fact that DotTune arrives at the correct AF tune value without ever engaging the AF motor of the lens. Since PDAF is closed loop, by definition a fixed in-body AF tune bias cannot correct for AF motor miscalibration, since the magnitude of the final motor movement varies between PDAF acquisitions, based on where the focusable lens element is relative to where it will be direct to. Any external bias correction would need to be scalable based on the mechanical distance traveled for the final lens movement, and a single AF tune value is not scalable. Again since PDAF is closed loop, most individual AF motor movement errors that occur during a single PDAF acquisition will be implicitly accounted for during the loop's multiple iterations of sensing/movement operations.

That said, I believe the Canon/Nikon PDAF system was designed to rely on a base-level of accuracy within the lens's AF motor/servo; in fact it's not clear that fixing lens movement errors was even the primary motivation for the system to be closed loop; that may be only incidental to its design. That a lens can be consistently out-of-spec in regards to its mechanical positioning relative to where its directed, and that this condition can only be fixed through manufacturer shimming/adjustment as the 50mm manual you referenced indicates, is actually further support for the notion that AF tune cannot correct for such conditions, and that it wasn't designed to.

One final note..I don't believe the static tolerance issues necessitating AF tune are a body-only calibration issue. I believe the body is part of the equation, but I believe the centering quality of the lens optics plays a significant role as well.



Feb 17, 2013 at 10:03 AM
RonMJ
Offline

Upload & Sell: Off
p.4 #19 · p.4 #19 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes


snapsy wrote:
Any delay between the shutter half-press and confirmation dot is an immediate disqualification of confirmation for that tune value. It's when there's no delay that I then suggest you keep holding the shutter for 5 seconds to see if the confirmation dot remains solidly lit. I also suggest additional 5-second half-presses at the same value when you're near either end of the tune range, to better confirm the feedback.

One-Shot AF, right?



Feb 17, 2013 at 10:54 AM
schlotz
Offline
• • • • •
Upload & Sell: On
p.4 #20 · p.4 #20 · Dot-Tune: Autofocus Fine Tuning in under 5 minutes


Process as described has the lens set to MF using the half shutter to activate the green AF dot.


Feb 17, 2013 at 11:04 AM
1       2       3      
4
       5              37       38       end






FM Forums | Canon Forum | Join Upload & Sell

1       2       3      
4
       5              37       38       end
    
 

You are not logged in. Login or Register

Username       Or Reset password



This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.