Reikan’s Focal System- DSLR Autofocus Microadjustment Calibration

Posted by on May 28, 2012 in Blog | 22 Comments

I need and want all of my images to be spot on tack sharp.  So I spend the necessary time to make sure I get that from my photography.

The higher end DSLR’s today come with some kind of lens microadjustment.   And there are lots of ways to calibrate your lenses using that capability.   I wrote about all of the ones that I knew of about a year ago HERE.

But I found a new one…. and its REALLY Cool.  In fact it’s the only method that I use now.  Its Reikan’s FoCal System.

It’s AUTOMATIC or at least that is one of the options.  There is some set-up required but no more than all of the other systems.  After that just sit back and relax while it snaps away.  You even get a data report at the end.

So let me tell you about the System and why I think its so great.   I will also show/tell you about some tips about how to run the calibration.

So what is the system.   Well there is no hardware….its all software.   You need to have a decent printer and a PC or a Mac (in a week or so a Mac version is coming out).  I think its better to have a laptop so that you can set it up close to the camera.  You will need a tripod.   Oh and your camera and lenses are necessary.

How does it work?   Well this is my understanding of how it works.   They have  a target that has a lot of sharp edges on it.   I think they somehow measure the total whiteness of the target.  If the lens is not focused there will be blurring of some of those sharp edges.  That will reduce the total whiteness.   The software controls the camera by taking several shots at each each AF MA (micro-adjustment) point.   A curve of the points of data is plotted of whiteness value (they call it Focus Quality) versus MA point.  They are very secretive on how this all works so this is only a guess on my part.   There is a peak whiteness point to the plot for a specific MA point…. And there is your final MA setting.

The software keeps track of any light changes that have nothing to do with the MA adjustment.   This is important in that if you are using sunlight or indirect sunlight or CLF lighting.   There can potentially be light shifting.   Also if there is not a good curve fit as it traverses the MA spectrum it will ask you to run more data points.  The program can also run focus points off center.  It can run wide and telescoped calibrations on zoom cameras.   It also has a tool that will run a lens through its different aperture values and tell you the sweet ( aperture) spot.

I am an engineer by profession and I gotta tell you that this software is REALLY Cool.

Can I trust it?

I think yes.  I went back to the Lens AlignPro system the I was previously in love with and reran the cal on that and got the same result.  But I realized then that there is a lot of quasi qualitative judgement going on.    In the visual systems you rely on some judgement on where the front focus area starts and where the back focus ends.  These are not hard lines because depth of field focus continuously improves to a focus point and then diminishes to the back end of the the dof.

The Reikan’s Focal System is taking digital measurements (I assume) and using statistics to determine where that peak focus is.  I suppose that a weakness of this compared to the others is that we do not all want our peak focus point in the center of the DOF field.   Some of us want a little front focus others a little back focus.  This is visualized better on a visual system like lens align.  But, you can get the same results by looking at the Reikan’s Focal System data and just picking an AF MA that will put the peak just forward or behind the optimum.  But you won’t get the visual confidence that you CAN see in lens align.

Another issue with Reikan’s Focal System is the lack of a target to camera alignment system.  Reikan’s Focal System does have a neat software system that verifies the target and makes sure that you have enough of the target in the view, that the center is under the cameras center focus sensor, and that the  target is not rotated too far.  But there is nothing that makes sure that the target plane is parallel to the camera sensor plane.  As I discussed in earlier blog LensAlign Pro has a system that ensures these planes are aligned.   I get around that by putting my target on a wall that is more or less perpendicular to the floor.

Another tip is the paper for the target.   Reikan’s Focal System recommends a thick good quality paper.   I ignored this at first and just used regular paper.   Bad idea.  I found that it is difficult to get regular paper to lay flat on a wall no matter how much you tape it down.   In addition I think the higher quality of the paper allows for sharper edges on the target and this,  I believe improves the test.  The paper that I found works best is Office Depot’s Premium Brochure & Flyer Paper.  Its 45 lb paper with a 96 brightness rating.

Its also important to have the target well lit.  I started with just using the indirect light from my windows and the overhead room light.   There were lots errors with added points needed.  Next I tried a couple of small LED lights that I use for reading sheet music.  FoCal instructions say to NOT use LED as it is a pulsing light.  I checked it out on the internet and LED’s do pulse, some as slow as 60 hz.   But they did work out fairly well with an occasional error.   Recently I have been using a large bright (high kelvin) CFL that I use with my lightbox set up.   FoCal also warns about this as it too pulses.   But my research showed that the frequency of these lights are between 1200 and 45000 hz.    All the tests that I have run with FoCal have been with a TCP spring light CLF (6500K), with a pulse frequency of 45K hz, work with this light.  Note that I have found that at 45000 cycles per second that, from internet postings, that the phosphorescence in these CFL’s stay bright between pulsing that is that fast and therefore they are a constant light source.    I think focal would like you to do this in direct sunlight, which I think is impractical,  or will some other non pulsing light.   I am guessing that there are studio lights that do that…. but I don’t have any so my set up will do.

Here is a video of the light discussion:

Positioning the target and camera are important as well.   I measure the height to the center of the target.   I then set up my camera on a tripod.  I have a ball head with a level and make sure that the head is more or less pointing at the target and level.

 

 

 

I then put my camera on the head and adjust the camera height so that the center of the lens is the same as the center of the target.  With that done I look into the viewfinder or the live view and move the tripod/camera left or right to put the center focus sensor on the target.  If the height is not right I will adjust the height again.    Sometimes I have to adjust the ball a tiny bit to make sure that the sensor is right in the middle of the target.  But no major adjustments.  And I deem that good enough.   I am tempted to build a target like the lensalign but with the Reikan’s Focal System target.   Note that I may have totally missed the technology associated with the Reikan’s Focal System and maybe target to camera sensor parellel-ness is not that necessary.

Here is how I do my final alignment of my camera to the target:

 

Finally I run the calibration.   Now there are a lot of presets or preferences that you can set depending on whether you want a very thorough and complete cal or a more cursory one that just gets you in the ballbark.  The camera settings need to be preset.   Although the software will do a lot of this.   The instructions that come with this system are long but very thorough and easy to understand.

When you start the software, with the USB connected it should recognize your camera as one of the cameras that you have registered to this software.  You then need to run the target program to make sure the target is set up correctly.   The software will provide a live view along with visual clues as to where you need to move the camera.  The start the program.   Get a cup of tea or coffee.  Relax.   But do it quickly because the calibration is relatively quick depending on your preferences.   You can also watch the readings as the program goes through the MA selections.

About half way through you will see a colored bar (green) that will indicate an initial guess at the correct MA point.   Then the program checks and validates the MA points near that point.  Then you are done.   The statistically optimum point is chosen and set.   You can even download a report, with lots of pages, detailing the findings for that lens.

There are some other goodies on the software as well.   It will tell you where the aperture sweet spot is on the lens.  It will tell you if you have dust on your sensor and where it is.  It will check to see how repeatable the camera lens combination is on AF.

I think this is just amazing.   Here is video of my running a calibration on a Canon 5D Mk II with a 24-70mm f2.8 lens.

Finally,  I did this cal with the 1.4 version of the software.  It is PC only.   They recently included both the Canon 5 D MK III and the new Nikon DSLR’s.  I know that with the 5D Mk III that it will only run in semi-automatic mode.  So everything is the same but I have to manually change the MA between each shot.   Its not the best but I still will use it over the lens align because I trust the results.   I think that Reikan is working on it.   The Canon software necessary to run the auto was only recently release.   And they are not sure if they will be able to make that camera automatic as of late May12.

The other cool thing is that there will be a MAC version coming out very soon.   Its in beta release now.    I do all my photo processing on the Mac side of my macbookpro so I would rather do the calibration on that side.

A final note.  As cool as ALL of these lens microadjustment systems are,  And again I think Reikan’s Focal System is the best,  this does not guarantee an in focus shot.  Could could be on a tripod looking at a stationary target.  You could have the most rigorous Reikan or LensAlign calibration going for you.   And the shot could still be a little out of focus with AF.   Because AF is not 100% repeatable.  The AF systems today are awfully good but not perfect yet.   So I say do these calibrations and then take lots of shots…. One will be perfect!!

One more thing…. please do not look at the little bits of food on my glasses and lower lip…… Its embarrassing.  I will know now to look in the mirror before my next video.   But its too much work to do these things and I just figured…. what the hell.

22 Comments

  1. Fredrik Rodhe
    May 29, 2012

    Hi,
    How do you adjust a zoom lens? When runnig FoCal on my zooms I found that the program wanted different adjustments on different focal lengths. And I obviously don’t want to change the MFA every time I zoom in or out.
    Regards
    Fredrik

    Reply
    • bobbyk
      May 30, 2012

      Thanks for the comment. I think for your situation you should pick a focal length that you use the most OR pick one in the middle of the zoom. Hopefully there is not a large difference between the MA adjustments between wide and tele. For me, I just got the Canon 5DMKIII and it has entries for wide and tele so in the in between shots the camera interpolates.

      Reply
  2. Ron Ginsberg
    June 1, 2012

    Greetings and thank you for your review and tech tips.

    A couple of things I tried with some different results:
    A USB over CAT 5 extender with a 100 foot cable did not work. I wanted to run it through a window from my desktop to the camera outdoors to get a long range for my 300mm f/4L and 300mm + 1.4xTCIII. The error was a .NET timing error. Oddly, the remote viewer functions and control of the EOS Utility worked. Since I was thinking about it anyway I acquired a laptop and the standard seven foot 50D USB cable worked well. (I use thee laptop in the field for weather radar over wireless and image previews.)

    Regarding mounting the target I acquired an inexpensive picture frame and mounted the thick matte paper target over the protective plastic sheet so nothing is in front of it. I affixed a small level to the frame horizontal member and attached a stranded stiff electrical wire for hanging it from a work light tripod I have. This allows me to level the target horizontally. Using the level on the tripod head gets me into the ball park. I can also put live view up with a grid to fine tune leveling and perspective as needed.

    My 50D came with a global MFA setting of +7. The 300mm alone resulted in +10 and with the 1.4x TCIII +14.

    I have a walk-around 18-200mm kit lens that I thought I would never really get quality images out of. I ran it a max zoom and it resulted in an MFA of -6. That’s 13 below the default +7 and I now can crop really well getting reasonable sharpness.

    In addition I started using it before it was released for the 50D in manual mode using the fine .jpg images. The resulting settings were not to far off from the auto settings.

    I also like the semi-auto mode where I start at a point and manually step one way and then come around around from the other side. There is till both data and the highly cropped image to view and this provides a visual check.

    One point not mentioned is that the report in .pdf format includes a section for each testing point even those repeated with an image quality data result and an embedded highly cropped image of the center area and immediate surround plus the test parameters and result graph as indicated. This can be stored and printed for future reference.

    As to the secret method I’ll just guess that a software high-pass edge detect filter is used with a random clutter filter to reduce noise effects maybe – in combination with average white level. I’ll throw a dart with that one.

    The support site now allows a user to change on his own licensing data for a single body or up to I believe six lenses. Don’t recall off hand. The user can modify the license parameter up to six times before contacting the developer is necessary.

    The app is very well thought through.

    Reply
  3. Bob Keenan
    June 2, 2012

    Interesting comments, I like the frame idea. I also have the 5d MK 3 and am using the 1.5 beta version of FoCal. It only works in semiautomatic mode but I still like it. I guess the problem is that the newer Canon firmware is not allowing external MA adjustments.

    Reply
  4. Reikan FoCal Review
    June 10, 2012

    […] […]

    Reply
  5. Frank
    June 10, 2012

    Why do you think FocCal has no way of making sure the target is 100% aligned parallel to the sensor plane? Because its not that critical. At 50x FL ( the recommended distance) and f2.8 your DoF is around 30 cm. The 0.5 cm error with a manually aligned targets insignificant.

    Reply
    • bobbyk
      June 10, 2012

      Good point… I guess that explains it. Thanks

      Reply
  6. Einar
    June 17, 2012

    It’s quite simple to parallel the camera to the target by using a mirror. When the the reflected image of your lens is in your viewfinder/screen you are done.

    Reply
    • bobbyk
      June 17, 2012

      Yes, someone else suggested it. It works great. Thanks for the comment.

      Reply
  7. Pete
    June 26, 2012

    I have just started playing with Focal Pro and my Nikon lenses and I am delighted with this software.

    I did notice however in the last movie that the results for your 24-70 are out of this world… at least compared to my Nikon 24-70.
    It would be great if Reikan could implement sort of an online database with results, so users could easily see if their lens is within an acceptable range or not.

    Also the note about zoom lenses, apparently you should test at both ends and pick a value in the middle. Well, at 24mm I should set my AF at +6 and at 70mm I should set it at -16.
    If I pick a value in the middle towards the tele end, say -7, how could that possibly improve anything ?
    At the wide end, I would be 13 off the ideal mark, where as if I just left it at it’s default 0, I would only be 6 points off. At the tele end it would improve by 7 points, but it would still be far from ideal.
    Maybe my lens just needs recalibration by Nikon ?

    Reply
  8. Allan Carter
    July 1, 2012

    Thanks for the info bob, it will be most useful. I have multiple bodies and lenses and this could be a great tool.
    I too have a Canon 24-70L that I have never had total confidence in. I will run the MA calibration at both ends of zoom and if the variation is too much, I will send it in for calibrating. I just need to sort out the lighting and I will be good to go.

    Reply
  9. Jason
    September 1, 2012

    Hi Bob,

    Have you tried with 1.5 or thre 1.6 release and the new targets? Did you blow up the target to a larger size. I am not sure if I should zoom in on the print out so that only the QR type image is the only printed piece on the matte paper.

    Reply
    • bobbyk
      September 2, 2012

      I have been using the 1.5 with the target at the normal printed size and it works fine. I you are calibrating a large focal length lens then maybe you want to make the target larger,

      Reply
  10. FredG
    September 13, 2012

    Hi Bob
    Thanks for the painstaking effort in putting up your findings. If I may ask, how would you test a 25-105 on a 5D3 with FoCal? Would you test it once at the 105 end X50fl, and once again at the wide 24 end X50fl, and then dial in those TWO recommended microadjust values into the 5D3 respectively?
    Rgds
    Fred

    Reply
    • bobbyk
      September 14, 2012

      That is exactly what I would do. One of the great new features on the 5D3 is the ability to put in two zoom values. BTW it my opinion that if 50xfl is too far away the 25xfl will also work.

      Reply
  11. Jose
    January 5, 2013

    Hello Bob,

    First, Happy New Year. Let me also say you thak you very much for your article, is really nice and helpfull, specially for me that I am having some problems with the focus of my Nikon D7000 and your words let me find a way to test if there is any inconsistency on the focusing system of my camera.

    I would like to ask you about the lights that you describe up here in your article, “TCP spring light CLF (6500K) with a pulse frequency of 1,2-45Khz.”. Well, here in Europe (Spain), I can’t find nothing even close to that pulse frequency. Maybe is due to the voltage… we have 220Volts here. I have also tried to find them on Internet but without any success. Could you please tell me if know some place where I could buy 2 ir 3 of them by the Net? Thanks in advance!

    Jose

    Reply
    • bobbyk
      January 5, 2013

      I wish I could help. But, I am not sure where you can find light bulbs like that which operated at 220V. When I got these lights I was looking for 6500F high watt light bulbs. At the time I did not even know about the cycling problem. After I got the FoCal system I got worried about using those lights. So I went to the Manf. site and they had a spec sheet that showed the high frequency. So I just happened to be lucky.

      Reply
  12. David
    February 23, 2013

    Great review. Im very intrigued, but doing some quick math in my head multpying lenses, bodies and teleconverters yields a lot of combinations. Even taking out a couple of lenses or lens/body combinations I don’t use very much, I’m still way, way above 6. Do I understand you correctly that once i hit 6 combinations I would need to contact the company to de-register one set of bodies or lenses and re-register the next batch, but would that require me to pay an additional licensing fee every time I do it? Thanks!

    Reply
  13. bobbyk
    February 23, 2013

    I think it was one of the posters who said that and I think it was a typo. I checked their site and its allowable up to 5 camera bodies and however many lenses you have.

    Reply
  14. Mitch
    February 25, 2013

    Where do you get the target? I assume you get a graphic with the software and simply print it? Seems if you only have an inkjet, you may not be getting the sharpest edges? A laser printer might seem better – or are each good enough?

    Reply
    • bobbyk
      February 28, 2013

      You print your own target. The quality seems to be good enough on a good inkjet. I set it for the highest quality. I also had a little problem with flatness of the paper. Thin paper does not want to stay perfectly flat and parallel to the camera sensor. So I got some thicker stock, bright white paper. That did the trick for me. Maybe that was not necessary but…. it made me feel better about my set-up.

      Reply
  15. Links | GFC Photography Blog
    February 8, 2017

    […] Well worth a look! Ron Martinsen’s Photography Blog Great site for articles and reviews The Reikan MA System Review A great review of a fantastic piece of software. Brilliant Isle of Skye photography A quite […]

    Reply

Leave a Reply