My current method to determine AF-Fine Adjust values

Started Mar 26, 2013 | Discussions
Kerusker
Contributing MemberPosts: 714Gear list
Like?
My current method to determine AF-Fine Adjust values
Mar 26, 2013

My approach is a more practical than scientific one.
Starting point is Live View (LV using contrast detection on sensor = CD-AF) beeing more accurate than Phase Detect AF (PD-AF or AF for short). Using manual focus in LV (LVm) with up to 10 times magnification may be even more accurate than auto focus in LV (LVa). For simplicity, I use LVa for this description.

procedure A:
Auto focus in LV (LVa), turn off LV and observe the lens barrel when auto focussing (PD-AF).
The barrel may not move, turn towards 0 or turn towards infinity (v, ->0, ->oo)
v:      the AF-FA value seems to be ok (at least close to best)
->0:   PD-AF does FF i.e. move the AF-FA value to negative direction (f.ex. +5 ->+1 or +2->-4)
->oo: PD-AF does BF i.e. move AF-FA value to positive direction (f.ex. -2->+3)
Repeat procedure A until the lens barrel does't move anymore.

In procedure A you will find that there may be several AF-FA values without barrel movement.
Therefore we will try to find the AF-FA limits where the barrel starts moving towards 0 (V->0) and towards infinity (V->oo). Usually we'll get a range V->oo,.....,V->0 in the AF-FA scale from -10 to +10, but some times V->oo or V->0 may be out of this range (and can't be determined).

procedure B:
B1) Auto focus in LV, turn off LV and observe the lens barrel when auto focusing (PD-AF).
R1) The barrel may not move, turn towards 0 or turn towards infinity (v, ->0, ->oo)
Repeat B1 with a different AF-FA value until the barrel does not move.

B2) Lower the AF-FA value and repeat B1 until the barrel starts to move towards infinity (V->oo)
B3) Increase the AF-FA value and repeat B1 until the barrel starts to move towards 0 (V->0)

B4) Pick an AF-FA value in the middle between V->oo and V->0 (if you prefer slight FF, take the one to the left of the middle value).

For a zoom lens you should do procedure B at least for the wide end and the long end of the zoom and then pick an AF-FA value which is approximately in the middle of both V->oo,....,V->0 ranges. In bad situations, there may not be a common AF-FA value at all in both ranges.

You can find a video for a similar procedure here:
'Dot-Tune' http://www.youtube.com/watch?v=7zE50jCUPhM

I've tried 'Dot-Tune' on my K-5 but since focus is set only once to determine the AF-FA range, it seems to depend a lot on the chosen focus setting. Also all mechanical variances are excluded since AF is turned off for the lens and only the AF indicator is used.

In procedure B above you should execute B1) several times for each AF-FA value to get good statistics with all AF elements included. (LVa does refocus each time!).

Procedure A and B can be executed without taking any picture (but you can of course).
Test setup is very similar to the one in the video.

I use the following settings on my K-5:

2-sec timer (turns shake reduction off)
Enable AF on AF Button and disable AF on Shutter Button Half Press (Exposure Menu 5 'Button Customization')
Menu C1/4 'Meter Operating Time' = 3sec (minimum)
AF Center Point only
Lens aperture wide open (this is usually the case for auto focusing, but aperture may be closed for LV if target is very bright)
distance from cam to target should be about 50 times the focal length of the lens (or at least 25 times)

If the AF-FA ranges are often outside the -10,...,+10 scale (at one end), you should use Debug to control (and change) the AF µ-Correction value.

If you've read all this, you sure have an interest in AF fine tune.

Thanks

-- hide signature --

][.Kerusker
we don't see that we don't see (eye's blind spot)

 Kerusker's gear list:Kerusker's gear list
Pentax K-5 II Pentax K-5 Pentax smc DA 15mm F4 ED AL Limited Pentax smc DA* 16-50mm F2.8 ED AL (IF) SDM Pentax smc DA 55-300mm F4.0-5.8 ED
Pentax K-5
If you believe there are incorrect tags, please send us this post using our feedback form.
DarylK
Senior MemberPosts: 1,475Gear list
Like?
Re: My current method to determine AF-Fine Adjust values
In reply to Kerusker, Mar 26, 2013

Makes sense to me. Will have to give it a try and compare the result with where I presently have calibration set at.
Thanks,
Daryl

 DarylK's gear list:DarylK's gear list
Pentax K200D Pentax K-5 Pentax smc DA 15mm F4 ED AL Limited Pentax smc DA* 50-135mm F2.8 ED (IF) SDM Pentax smc DA 10-17mm F3.5-4.5 ED (IF) Fisheye +6 more
Reply   Reply with quote   Complain
GabFla
New MemberPosts: 12Gear list
Like?
Re: My current method to determine AF-Fine Adjust values
In reply to Kerusker, Mar 27, 2013

I think that a final check shall be done at +inf after calibration

Reply   Reply with quote   Complain
Roland Karlsson
Forum ProPosts: 19,519Gear list
Like?
Re: My current method to determine AF-Fine Adjust values
In reply to Kerusker, Mar 27, 2013

Clever

This procedure removes all the tedious test images and evaluations.

 Roland Karlsson's gear list:Roland Karlsson's gear list
Canon PowerShot G10 Pentax K-5 Pentax smc DA 15mm F4 ED AL Limited Pentax smc FA 50mm F1.4 Pentax smc DA 21mm F3.2 AL Limited +6 more
Reply   Reply with quote   Complain
DarylK
Senior MemberPosts: 1,475Gear list
Like?
Re: My current method to determine AF-Fine Adjust values
In reply to Kerusker, Mar 30, 2013

Kerusker,

I just gave this a try.  I very much like your general approach/concept.  However, what I did varied a little from the alternative approaches you mentioned.

What I did was do multiple focuses using live view with observation of where the distance scale of the focus ring most consistently stopped relative to a fixed mark on the lens barrel.  Then, I tried various fine focus adjustments using PD-AF until I was satisfied I found the right setting that most consistently set focus to that same point.  This method worked on all the lenses I typically carry around except for the 18-135 which doesn't have a focus ring that has distance settings (or moves when focusing for that matter).

Thanks for the tip!

Daryl

-- hide signature --

There is nothing noble in being superior to your fellow man; true nobility is being superior to your former self. - Ernest Hemingway
Great photography is largely a matter of perspective, both visually and mentally - Daryl Kottwitz (but probably said before )
My gallery: http://darylkottwitzphoto.smugmug.com

 DarylK's gear list:DarylK's gear list
Pentax K200D Pentax K-5 Pentax smc DA 15mm F4 ED AL Limited Pentax smc DA* 50-135mm F2.8 ED (IF) SDM Pentax smc DA 10-17mm F3.5-4.5 ED (IF) Fisheye +6 more
Reply   Reply with quote   Complain
Roland Karlsson
Forum ProPosts: 19,519Gear list
Like?
Re: My current method to determine AF-Fine Adjust values
In reply to DarylK, Mar 30, 2013

Nice that some are thinking

I have just used the tedious solution with test images of a slanted ruler.

Not fun!

 Roland Karlsson's gear list:Roland Karlsson's gear list
Canon PowerShot G10 Pentax K-5 Pentax smc DA 15mm F4 ED AL Limited Pentax smc FA 50mm F1.4 Pentax smc DA 21mm F3.2 AL Limited +6 more
Reply   Reply with quote   Complain
Kerusker
Contributing MemberPosts: 714Gear list
Like?
Re: My current method to determine AF-Fine Adjust values
In reply to DarylK, Mar 30, 2013

DarylK wrote:

Kerusker,

I just gave this a try.  I very much like your general approach/concept.  However, what I did varied a little from the alternative approaches you mentioned.

What I did was do multiple focuses using live view with observation of where the distance scale of the focus ring most consistently stopped relative to a fixed mark on the lens barrel.

Good idea. I left the problem how to focus most accurate in LV to the reader (AF in LV or manual focus in LV). Some lenses may not auto focus very well in LV - then you approach may not be the best. But that's easy to find out.

Then, I tried various fine focus adjustments using PD-AF until I was satisfied I found the right setting that most consistently set focus to that same point.  This method worked on all the lenses I typically carry around except for the 18-135 which doesn't have a focus ring that has distance settings (or moves when focusing for that matter).

Why not use a little adhesive sticker on the focus ring?

Thanks for the tip!

Glad that you tried it and gave your feedback.

I think it's a good method to learn about the AF behavior of a lens easily. And maybe the 'focus shift' and 'fine adjust irritation' problems are getting down to what really happens - at least a bit ;).

-- hide signature --

][.Kerusker
we don't see that we don't see (eye's blind spot)

 Kerusker's gear list:Kerusker's gear list
Pentax K-5 II Pentax K-5 Pentax smc DA 15mm F4 ED AL Limited Pentax smc DA* 16-50mm F2.8 ED AL (IF) SDM Pentax smc DA 55-300mm F4.0-5.8 ED
Reply   Reply with quote   Complain
DarylK
Senior MemberPosts: 1,475Gear list
Like?
Re: My current method to determine AF-Fine Adjust values
In reply to Kerusker, Mar 30, 2013

The focus ring doesn't turn at all so no visual feedback for the 18-135.

-- hide signature --

There is nothing noble in being superior to your fellow man; true nobility is being superior to your former self. - Ernest Hemingway

Great photography is largely a matter of perspective, both visually and mentally - Daryl Kottwitz (but probably said before )

My gallery: http://darylkottwitzphoto.smugmug.com

 DarylK's gear list:DarylK's gear list
Pentax K200D Pentax K-5 Pentax smc DA 15mm F4 ED AL Limited Pentax smc DA* 50-135mm F2.8 ED (IF) SDM Pentax smc DA 10-17mm F3.5-4.5 ED (IF) Fisheye +6 more
Reply   Reply with quote   Complain
Col K10d
Veteran MemberPosts: 3,198Gear list
Like?
Re: My current method to determine AF-Fine Adjust values
In reply to Kerusker, Mar 30, 2013

Very similar to the following method with a test image that takes advantage of moire at critical focus. Always found this accurate & fast.  Needs a little working distance for the 300mm..  Basically manual focus with live view and then see if PD AF moves you off critical focus.

http://www.northlight-images.co.uk/article_pages/cameras/1ds3_af_micoadjustment.h

 Col K10d's gear list:Col K10d's gear list
Pentax K-3 Pentax smc DA* 300mm F4.0 ED (IF) SDM Pentax smc FA 43mm F1.9 Limited Carl Zeiss Distagon T* 2,8/25 Voigtlander 90mm F3.5 APO-Lanthar SL II +9 more
Reply   Reply with quote   Complain
afterswish1
Contributing MemberPosts: 608
Like?
Re: My current method to determine AF-Fine Adjust values
In reply to Kerusker, Apr 1, 2013

Brilliantly simple.

Thanks very much! I'm kicking myself for not having thought of it earlier

-- hide signature --

Gravity will make you crazy until you get the hang of it.

Reply   Reply with quote   Complain
DarylK
Senior MemberPosts: 1,475Gear list
Like?
Re: My current method to determine AF-Fine Adjust values
In reply to afterswish1, Apr 1, 2013

Brilliantly simple indeed! So simple that you would think that a menu command could be added to have the camera self calibrate. Mount camera on tripod, point camera to a good contrast target at appropriate distance short of infinity, select "self calibrate lens fine focus" menu command, the camera does the rest with several fast focus iterations in each mode.
Seems doable to me...better patent the idea really quick...never mind, just thank me when you see it implemented.
Thanks,
Daryl
--
There is nothing noble in being superior to your fellow man; true nobility is being superior to your former self. - Ernest Hemingway

Great photography is largely a matter of perspective, both visually and mentally - Daryl Kottwitz (but probably said before )

My gallery: http://darylkottwitzphoto.smugmug.com

 DarylK's gear list:DarylK's gear list
Pentax K200D Pentax K-5 Pentax smc DA 15mm F4 ED AL Limited Pentax smc DA* 50-135mm F2.8 ED (IF) SDM Pentax smc DA 10-17mm F3.5-4.5 ED (IF) Fisheye +6 more
Reply   Reply with quote   Complain
Wanganuilad
Senior MemberPosts: 2,131
Like?
Re: My current method to determine AF-Fine Adjust values
In reply to DarylK, 10 months ago

Is there a choice to use phase detect or contrast detect AF ? I can't see how to select them

-- hide signature --

K-5 K-01 some Len

Reply   Reply with quote   Complain
Ivan Gordeli
Forum MemberPosts: 70Gear list
Like?
Re: My current method to determine AF-Fine Adjust values
In reply to Wanganuilad, 10 months ago

Wanganuilad wrote:

Is there a choice to use phase detect or contrast detect AF ? I can't see how to select them

In Live View it will be always contrast detection, without Live View - it will be always phase detection AF AFAIK

-- hide signature --

"Notes on Photography" blog http://gordeli.wordpress.com/

 Ivan Gordeli's gear list:Ivan Gordeli's gear list
Sony Cyber-shot DSC-F717 Panasonic Lumix DMC-ZS3 Pentax K-5 Tamron SP AF 17-50mm F/2.8 XR Di II LD Aspherical (IF)
Reply   Reply with quote   Complain
Wanganuilad
Senior MemberPosts: 2,131
Like?
Re: My current method to determine AF-Fine Adjust values
In reply to Ivan Gordeli, 10 months ago

Thankyou Ivan

-- hide signature --

K-5 K-01 some Len

Reply   Reply with quote   Complain
falconeyes
Contributing MemberPosts: 713
Like?
Re: My current method to determine AF-Fine Adjust values
In reply to Kerusker, 10 months ago

A more accurate method for people who do already own FoCal for their Canon or Nikon ...

1. Have the Pentax on a tripod take JPG shots of some test chart using normal PDAF autofocus. But use AF fine tune values of e.g., -20, -12, -4, 4, 12, 20 respectively (6 or 12 shots if doing 1 or 2 shots for each). Move focus manually, e.g. to infinity, in between.

2. Write the AFMA (AF manual adjust aka fine tune) values into the images EXIF header, using exiftool. I am sorry but I don't remember ATM the exact name of the exif field which are to be written. It may be necessary to fake more fields to make FoCal detect a Nikon or Canon.

3. Read the test shots into FoCal's manual mode. You need to connect your other camera to validate the license in order to enter manual mode. FoCal will automatically determine and display the optimum AF fine tune value. More on manual mode: http://www.reikan.co.uk/focalweb/index.php/tests/manual-mode/

The optimum fine tune values depend on many variables, like focus distance, color temperature, zoom focal length, which AF point to be used etc. This more accurate method gives a better feeling about the variances and let's you choose a compromise. A more heuristic try-and-error method may lead to a setting which works worse in the field.

Reply   Reply with quote   Complain
falconeyes
Contributing MemberPosts: 713
Like?
Erratum
In reply to falconeyes, 10 months ago

Step 2 in my procedure above is not necessary anymore. I've learned that Focal 1.9 allows to edit AFMA values within the manual mode tool, and that FoCal may deduce AFMA values from file names if missing in the exif header.

This makes the method I described above much more accessible and probably faster than try-and-error approaches. However, you still need another camera with a FoCal license for it.

Reply   Reply with quote   Complain
sleepwalker400
Contributing MemberPosts: 901Gear list
Like?
Re: My current method to determine AF-Fine Adjust values
In reply to Kerusker, 2 months ago

THX I dont find live view to be too accurate so I guess I would really need to  use liveview and manual focusl or just manual in eye view to set it. BUT THE REST OF THIS I DONT GET

 sleepwalker400's gear list:sleepwalker400's gear list
Sony Cyber-shot DSC-R1 Pentax *ist DL Pentax K10D Pentax K-30
Reply   Reply with quote   Complain
MJSfoto1956
Regular MemberPosts: 168Gear list
Like?
Re: My current method to determine AF-Fine Adjust values
In reply to sleepwalker400, 2 months ago

all lenses exhibit some degree of focus creep (see: http://www.dpreview.com/forums/thread/3521291?page=4#forum-post-51859169). Note: there are many other articles describing this phenomenon.

As such, a test done at f/1.8 can be totally wrong for the same lens and the same test at f/5.6!! (Note: I have verified this "focus creep" behavior by careful testing of my own Nikon lenses)

My recommendation (which is completely different than what nearly everyone else on the web is advising) is that you should test and validate and calibrate at your most "critical" aperture setting. I.e. if you shoot a lot of paid events at f/2.8 then you should calibrate your lenses at f/2.8, recognizing that your AF settings "may" not be ideal at f/8 (especially close up).

And likewise, if you tend to shoot landscapes outdoors at f/5.8 to f/8, then calibrate at f/5.6 and recognize that your AF settings "may" not be ideal at f/2.8.

I believe this behavior accounts for a large percentage of the described ambiguous behaviors being described on this and many other forums.

YMMV

Michael

 MJSfoto1956's gear list:MJSfoto1956's gear list
Pentax K-3
Reply   Reply with quote   Complain
sleepwalker400
Contributing MemberPosts: 901Gear list
Like?
Re: My current method to determine AF-Fine Adjust values
In reply to MJSfoto1956, 2 months ago

I was also thinking something similar. I think if you have a zoom you need to see what focal length and how far away you normally shoot and then adjust accordingly

 sleepwalker400's gear list:sleepwalker400's gear list
Sony Cyber-shot DSC-R1 Pentax *ist DL Pentax K10D Pentax K-30
Reply   Reply with quote   Complain
awaldram
Forum ProPosts: 10,464Gear list
Like?
Re: My current method to determine AF-Fine Adjust values
In reply to MJSfoto1956, 2 months ago

IMO you should always calibrate at the widest aperture and focal creep will be compensated by the increase DoF at smaller aperture.

The mistake most people make when attempting to calibrate their lens is they don't' bother/realize that that peak sharpness for a lens will not be 50-50 within the DoF but will vary depending on focal length and aperture.

Some example to show just how far people are miss-calibrating their equipment if they do not look up the correct peak placement in the DoF

16mm @f2.8 10ft target 17% in front 83% behind

16mm @f4 10ft target 3% in-front 97% behind

50mm @f2.8 10ft target 47% in-front 53% behind

Without this knowledge its impossible to calibrate a lens correctly whatever tool you use and IMO leads to many people thinking they have faulty equipment.

Hence why I'm so  cynical when you see 'everyone of my lens requires calibration'

For the average shooter not willing to read up on the subject its safer to let be than fiddle without sufficient information to complete the job correctly.

 awaldram's gear list:awaldram's gear list
Pentax Q Pentax K-3 Pentax smc DA* 55mm F1.4 SDM Pentax smc DA* 16-50mm F2.8 ED AL (IF) SDM Pentax smc DA* 50-135mm F2.8 ED (IF) SDM +11 more
Reply   Reply with quote   Complain
Keyboard shortcuts:
FForum MMy threads