Ad
Ad
Ad
Pages: « 1 [2]   Bottom of Page
Print
Author Topic: Reikan FoCal focus micro-adjustement automation software  (Read 17251 times)
BartvanderWolf
Sr. Member
****
Offline Offline

Posts: 3877


« Reply #20 on: July 07, 2012, 09:19:21 AM »
ReplyReply

I believe that many lenses tend to change the focus slightly when changing aperture, and that Canon are not able to compensate for this in their AF.

Hi,

Most modern lenses do not suffer from this enough to be noticeable, if at all. The fact that the Reikan software suggests so, may have something to do with the software rather than the lens, although it doesn't rule out the lens.

I'm also a bit surprised that the module for finding the best aperture apparently in this case isn't accurate enough to detect the peak, but instead a rather broad range. The example (3rd attachment) from my software tool shows that it is possible to be quite exact (in my chart lower blur is better resolution).

Cheers,
Bart
Logged
Wayne Fox
Sr. Member
****
Offline Offline

Posts: 2930



WWW
« Reply #21 on: July 07, 2012, 02:50:02 PM »
ReplyReply

Even more useful now that the 1DX is reported to have separate MFA for the wide and narrow end of zooms.
This is confirmed, this feature is also in the 5D Mark III.

I hope to test how well it works and how it tracks through various focal lengths.
Logged

Wayne Fox
Sr. Member
****
Offline Offline

Posts: 2930



WWW
« Reply #22 on: July 07, 2012, 03:01:56 PM »
ReplyReply

Not sure why the curve seems to flatten between f/3.2 and f/10. The exposure time creeps from 1s (f/22) to 1/8 s (f/6.3), might be that minor focus misalignement along with camera shake make for an upper limit to achieved sharpenss?

I've done this test using enough light to not need that slow of shutter speeds and seen the same results with every lens.  I've also clicked on the resulting points to see the various images, and to me the image quality is pretty consistent for many lenses once you stop down enough to get it sharp all the way to f/10 or so.  I guess the point would be diffraction, while perhaps having some slight impact, doesn't really seriously degrade image quality until you reach certain point, after which the sharpness falls off dramatically. So far it appears most lenses on the d800e are very similar from about f/4 to about f/8 and show only slightly more softness even at f/10.  At f/11 things are visually degraded and by f/22 every lens I've tested so far is complete mush ... very unusable.  f/16 would only be useful for scenes where no micro detail has any value. 
Logged

Rand47
Sr. Member
****
Offline Offline

Posts: 582


« Reply #23 on: July 09, 2012, 08:18:29 AM »
ReplyReply

Have any of you seen real world improvement in your images?  This is an honest question.  I purchased a lens align device. The more I read about how to use it and the more I read the Q&A on their web forum, the more I began to think that the variables are too numerous for me to think I was going to achieve a practical improvement in AF performance for my lenses that did not have an obvious and "visible in prints" AF problem. The variables were too numerous to allow a single MFA to provide an optimal outcome for all focus distances and apertures (not to mention focal length ranges for zooms,).   For obviously troubled lens/body combos it could certainly improve matters, but other than that it wasn't of value.

I ended up seeing the device as merely a diagnostic tool for identifying the degree and "direction" of visibly mis-focusing body/lens combinations.  I returned the device.  I can do that kind of diagnosis as easily with printed targets taped to the wall & carefully tripod mounted camera & tape measure.

So, I'm honestly interested to know if this new software based system actually produces real world improvements at working apertures with lenses whose performance was not "obviously troubled" prior to testing and adjustment with the software?
Logged
stever
Sr. Member
****
Offline Offline

Posts: 1073


« Reply #24 on: July 09, 2012, 08:51:46 AM »
ReplyReply

the answer, Rand is yes and no.  i just spent yesterday afternoon using LensAlign to microadjust my lenses for new 5D3 (i've used it for the 5D2 and 7D as well).  the Wide/Narrow adjustments intrigued me, but of the 24-105, 70-200 f4 IS, and 100-400, only the 24-105 required the most adjustment (from 40 to 90 as i don't have a long ruler for wider and try not to use the lens at 105) to be noticeable wide open -- -3 at 40, 0 at 90. the 70-200 was +2 at 85 and +3 at 190.  the 100-400 was -2 at 100 and 0 at 400.  these are all adjustments that would noticeable wide open on close inspection and i think useful to me as these lenses often get shot wide open.  since i'm generally unhappy with the 24-105 wide open i'm looking for any help i can get.

in the past i've used LensAlign to identify problems and check results from Canon repair (generally satisfactory).  One exception being my 50 1.4 which has pretty well known focus shift issues and has been to Canon a couple time with minor improvement in performance at f2.8 (it's always been great above f4 and terrible wide open).  it requires -9 at f1.8, -12 at f2.8, and -16 at f4 presenting a dilema.  i'll do some resolution tesing with Imatest soon out of curiosity and for guidance (don't expect another trip to Canon to help and don't think it's worth testing more lenses as i believe this focus shift is a design issue).

i think that Canon could correct for aperture focus shift in the future if they are willing to do it.
Logged
Scott O.
Sr. Member
****
Offline Offline

Posts: 312


WWW
« Reply #25 on: August 16, 2012, 11:16:32 PM »
ReplyReply

Hi, my name is Scott and I have been a micro-adjustor for several years.   Grin Grin Grin

Actually, there are so many variables that a guess is the best you will do...the variables are:
- focal length of zoom lens
- aperture used
- distance from target
- maybe amount of light on target or paper it is printed on

All this effort for something that may or may not be significant in the final quality of the image...gads I am anal!
Logged

Ellis Vener
Sr. Member
****
Offline Offline

Posts: 1851



WWW
« Reply #26 on: August 17, 2012, 03:25:56 PM »
ReplyReply

Have any of you seen real world improvement in your images?  This is an honest question.  I purchased a lens align device. The more I read about how to use it and the more I read the Q&A on their web forum, the more I began to think that the variables are too numerous for me to think I was going to achieve a practical improvement in AF performance for my lenses that did not have an obvious and "visible in prints" AF problem. The variables were too numerous to allow a single MFA to provide an optimal outcome for all focus distances and apertures (not to mention focal length ranges for zooms,).   For obviously troubled lens/body combos it could certainly improve matters, but other than that it wasn't of value.

I ended up seeing the device as merely a diagnostic tool for identifying the degree and "direction" of visibly mis-focusing body/lens combinations.  I returned the device.  I can do that kind of diagnosis as easily with printed targets taped to the wall & carefully tripod mounted camera & tape measure.

So, I'm honestly interested to know if this new software based system actually produces real world improvements at working apertures with lenses whose performance was not "obviously troubled" prior to testing and adjustment with the software?

Yes I have used the LensAlign, both the Pro and now the Mk II model. I do see a real noticeable difference between an untuned body/lens combination and a tuned one. I have seen this with various Canon and Nikon bodies and lenses.

Your point about the number of real world variables is a sound one however getting a baseline for "best" AF  performance  beats not having a measured baseline at all.  This will hold true whether you use the LensAlign or the Reikan FoCal or any other system to check AF performance.

In situations where your subjects are not moving and neither is your camera (because it is mounted on a tripod or camera stand) focusing using Live View either manually with an external magnifier like the Hoodman Loupe and also increasing the live view magnification or using the contrast detection of AF (as opposed to the phase detection based AF system used when you are shooting and viewing through the viewfinder) will yield best results.


Focus is focus and sharpening is sharpening and it is a wonderful thing when they meet.
Logged

Ellis Vener
http://www.ellisvener.com
Creating photographs for advertising, corporate and industrial clients since 1984.
Scott O.
Sr. Member
****
Offline Offline

Posts: 312


WWW
« Reply #27 on: August 17, 2012, 09:43:43 PM »
ReplyReply

Software version 1.6 was released today...according to FoCal the algorithm has been changed to give more repeatable and accurate results, among other things.
Logged

Rhossydd
Sr. Member
****
Offline Offline

Posts: 1991


WWW
« Reply #28 on: November 09, 2012, 05:07:27 AM »
ReplyReply

Following on from the release and discussion of Michael Tapes new MFA tool here on Lula, plus having recently had a few less than optimally focused shots myself, I decided to have a fresh look at FoCal and see if things have progressed since this thread was started.

The first observation is that FoCal is now a very well documented, easy to use and reliable program. It also works well on hardware below their recommendations, so I've been able to work outdoors with a little 1.4ghz netbook to calibrate longer lenses without being tied to a desktop system.
The target validation routines seem to work well and are simple to use which should ensure good reliable results.

FoCal Pro automatically shoots and evaluates dozens of shots without any intervention. I would expect the greater number of samples should lead to more accurate results.
The analysis information being presented in real time as testing progresses instils a good degree of confidence in the procedures used. The reports generated and saved at the end of analysis add to that confidence by including the sample images for user review.

As the process is so simple once the target has been set up, it's easy to carry out extensive testing of how MFA changes with f stop and how consistent the AF system is. Trying to do this sort of testing manually would take a very long time and might be prone to errors unless it was carried out with meticulous attention to detail recording results.

As previously noted in this thread, one curiosity of the whole MFA testing procedure is that it reveals the differences in AF performance on lenses depending on the selected f stop and shooting distance and, in the case of zoom lenses, focal length too. This is information that people may be unaware of if they just do very basic testing with only a few variables.
Once you do have a full picture of how AF performance is varying with different settings it's possible to consider the optimum value for one's own shooting style and preferences. For some this might be easy, e.g. always using a 85mm f1.2 wide open for portraits at 15ft. For others it might need more consideration e.g. when FoCal shows a wide ranges of values for a zoom that will be used at many different settings.
I'm tending towards using the setting recommended for the most critical conditions, wide open at the longer focal length, so far that's working well for me.

My own initial testing is suggesting that previous chart based manual evaluations had given poorer results than I realised, or that values have shifted over the last two years, possibly a bit of both. It will be interesting to see if these values actually change over time or with different environmental conditions.

The next question is "Does AFMA make a difference in real world shooting ?"
There's no easy answer to that. Although unlikely, you may have a set of lenses that are so close to perfect AFMA isn't necessary or your technique may be so poor that you wouldn't see any difference anyway.
More probable, is that one or more of your lenses will benefit from applying an AFMA correction that will deliver sharper results. Anyone doubting this can try the simple test of doing some shots with large changes in AFMA settings and studying the results. It does make a difference.
With the back garden tests I've done so far, I can see significant improvements after getting the correct AFMA and now AF performs better than I can achieve manually through the OVF.
I've also done some compassions using different ISO values. FoCal recommend using the base ISO for best performance, but I doubt there's much difference in the resolution of JPGs out of my 5Dii up to 400iso. So in the UK winter gloom with long lenses I can see improvements in consistency of results by using a higher ISO to use faster shutter speeds.

Given that getting AFMA correct is worthwhile, FoCal is a worthwhile investment as it seems to deliver the results you need to make an informed decision about settings without needing the large amounts of effort needed to check settings manually.

Paul


Logged
hjulenissen
Sr. Member
****
Offline Offline

Posts: 1713


« Reply #29 on: November 09, 2012, 05:36:23 AM »
ReplyReply

Thanks for your update. Did you experiment with the different "reset" methods? ("move the focus to infinity between measurements" vs "move the focus to 0 between measurements")

-h
Logged
BartvanderWolf
Sr. Member
****
Offline Offline

Posts: 3877


« Reply #30 on: November 09, 2012, 05:55:59 AM »
ReplyReply

Thanks for your update. Did you experiment with the different "reset" methods? ("move the focus to infinity between measurements" vs "move the focus to 0 between measurements")

Hi,

That is indeed, in my experience with tethered manual focusing, the main variable. Phase controlled AF cannot really be user adjusted for that, other than averaging the hysteresis bias that will exist. As Michael Tapes suggests with his more manually driven software solution, be consistent when calibrating and always defocus in the same direction between shots. I'm not sure what the automated Reikan software does.

Cheers,
Bart
Logged
hjulenissen
Sr. Member
****
Offline Offline

Posts: 1713


« Reply #31 on: November 09, 2012, 06:43:44 AM »
ReplyReply

...be consistent when calibrating and always defocus in the same direction between shots. I'm not sure what the automated Reikan software does.
There is a user setting, somewhat well-hidden with 4-5 choices. "reset towards infinity" and "reset to 0" are two. I dont remember the default.

I wonder if AFMA settings could be accessed by magic lantern software? If so, For those willing to spend time and money to get a large number of camera/lense focus error data might wish for more extensive correction (i.e. a function of focus distance, focus movement direction, focal length, aperture, etc).

http://www.magiclantern.fm/
-h
Logged
Rhossydd
Sr. Member
****
Offline Offline

Posts: 1991


WWW
« Reply #32 on: November 09, 2012, 07:09:37 AM »
ReplyReply

Did you experiment with the different "reset" methods?
Not yet. It would be interesting to see if AFMA changed with which direction you approach focus from, but I'm not sure knowing the result would actually be much help in real world shooting.
Logged
BartvanderWolf
Sr. Member
****
Offline Offline

Posts: 3877


« Reply #33 on: November 09, 2012, 10:19:56 AM »
ReplyReply

Not yet. It would be interesting to see if AFMA changed with which direction you approach focus from, but I'm not sure knowing the result would actually be much help in real world shooting.

If approaching from one direction repeatedly gives an adjustment of +5, and approaching from the other direction +7, then on average a +6 setting will give the overall lowest front/back focus bias.

Cheers,
Bart
Logged
Rhossydd
Sr. Member
****
Offline Offline

Posts: 1991


WWW
« Reply #34 on: November 09, 2012, 10:42:39 AM »
ReplyReply

If approaching from one direction ......
... average a +6 setting will give the overall lowest front/back focus bias.
Sure, I understand that, but once you start looking at the full range of possibilities direction to focus is one of the variables we're least able to control in normal AF shooting.
Logged
hjulenissen
Sr. Member
****
Offline Offline

Posts: 1713


« Reply #35 on: January 16, 2013, 08:46:19 AM »
ReplyReply

They are now publishing (anonymous) info about focus/sharpness from their users. I think this adds valuable information about real-world mean/variance performance of a particular lense/cam combo in addition to what lense reviews/measurements of a single lense, manually focused optimally can tell us:
http://www.reikan.co.uk/focalweb/index.php/online-tools/lenscamera-information/


Sadly, the y-axis is not a well-defined function of e.g. mtf50, but it seems safe to assume that it is a more or less monotonous function of sharpness/focus.

Unsurprisingly, 50mm f/1.8 II tends to degrade its performance more than the 70-200 f/4 IS as aperture is moved towards their respective maximums:
« Last Edit: January 16, 2013, 08:52:04 AM by hjulenissen » Logged
Pages: « 1 [2]   Top of Page
Print
Jump to:  

Ad
Ad
Ad