• Please use real names.

    Greetings to all who have registered to OPF and those guests taking a look around. Please use real names. Registrations with fictitious names will not be processed. REAL NAMES ONLY will be processed

    Firstname Lastname

    Register

    We are a courteous and supportive community. No need to hide behind an alia. If you have a genuine need for privacy/secrecy then let me know!
  • Welcome to the new site. Here's a thread about the update where you can post your feedback, ask questions or spot those nasty bugs!

FoCal, a new AF Micro-adjustment tool

Bob Latham

New member
I thought I'd bring an update to the Micro-Adjustment discussion as I've been performing some beta testing on an automatic calibration system written by a guy in the UK.

The software was developed using Canon's SDK utility and attempts to find the correct MA value for each lens. The code controls the camera via its USB port and takes a series of upto 40 images of a test chart (at all MA values from -20 to +20) and then performs an analysis of the contrast. The information is presnted in graphical format and the derived MA value is written (if required) to the camera body.

The initial release covered the 7D and 5DMkII with the 50D and 1DsMkIII now being tested. The 1DMkIII and 1DMkIV will follow shortly. I have just completed calibration of my 1DsMkIII with 7 primes (14, 24, 35, 50, 85, 100 and 135mm) with a solitary zoom (70-200) and the results look good. The values differ from the ones I obtained using a Lens Align Pro but not greatly so and all cases bar one have agreed with my previous direction (front or back focus). I'll be trying the telephotos and remaining zooms in the next day or so.

More can be read here at the developer's website. http://www.reikan.co.uk/focal/

Bob
 
Hi folks,

This new thread, is a branch that I've split off from the original AFMA thread that sparked a lot of interest in the subject of Auto Focus Micro-adjustment. This thread is dedicated to a new initiative that Bob Latham brought to our attention. Thanks Bob!

I suggest that all specific discussions about FoCal are concentrated here, and other AFMA solutions are discussed in their own threads to improve readability.

Cheers,
Bart
 

Asher Kelman

OPF Owner/Editor-in-Chief
Bob,

I'm so glad to see that folk are taking advantage of the SDK of the Canon camera to do more than Canon does. I wish people would realize that there's a lot of room for creativity, just as with the iPhone. This tool I hope is going to be very useful. Hopefully, I'll get to try it too!

Asher
 
How can you do it with the 1Ds3? You can't control the focus from the computer with that body...

Hi Ben,

The focus of the 1Ds3 can be controlled via a USB tethered cable. The EOS Utility, in Live View mode, allows to manually focus the camera, so I presume that the SDK also offers that functionality. Maybe you were thinking of the earlier 1Ds2, which didn't have Live View.

A compagnion uility to Helicon Focus called Helicon Remote used to allow automatic calibration with that tool which was officially intended for focus stacking, but the developers dropped that AF calibration functionality due to : "There are several reasons that do not allow the program to get reliable measurements". One reason they mentioned was the Nikon implementation of Live View not having enough resolution.

So it will be interesting to see how FoCal deals with those issues.

Cheers,
Bart
 

Bob Latham

New member
The 1DsMkIII is the oldest of the bodies with MA functionality and the only one where Mirror Lock-up isn't supported in the SDK. The 5DMkII, 7D, 50D tests allow for a selectable (default 1s) delay for improved results but this isn't an option for the 1DsMkIII and so tripod stability becomes a more important factor.

I'm staying a little open minded about the qualities of this system until the 1DMkIV is supported. I have two of these bodies and would hope to see closely matched results on a lens by lens basis with only a global offset to be seen which will be indicative of minor differences in the body's AF calibration.

Bob
 
The 1DsMkIII is the oldest of the bodies with MA functionality and the only one where Mirror Lock-up isn't supported in the SDK.

Hi Bob,

That's odd and unfortunate, because on my 1Ds3 Mirror Lock-up does function in the EOS utility when shooting tethered. When the camera is set to C.Fn III 15, both options 1. Enable, and 2. Enable:Down with Set, allow to take images with the mirror staying up (activated on the camera first because there is no function for that in EOS utiity) between exposures.

Cheers,
Bart
 

Bob Latham

New member
Bart,

If I understand you correctly then I belive you've highlighted the problem. Mirror lock-up needs to be activated on the camera and this would prevent an automatic function that cycles through writing an MA value, refocussing, raising the mirror, waiting for an elapsed time, taking the shot and then dropping the mirror prior prior to starting the cycle again with a new MA value.

There is a more manual approach available in FoCal that allows images to be taken and then analysed by the software to provide a graph of MA value versus contrast (read sharpness).

Bob
 
Bart,

If I understand you correctly then I believe you've highlighted the problem. Mirror lock-up needs to be activated on the camera and this would prevent an automatic function that cycles through writing an MA value, refocussing, raising the mirror, waiting for an elapsed time, taking the shot and then dropping the mirror prior prior to starting the cycle again with a new MA value.

If the camera needs to drop the mirror before the firmware can be written to, then that's indeed not possible (if the the mirror cannot be raised from the software to take the next shot).

Since the AF stepping motor determines the smallest focus step that can be taken, maybe they could use the number of micro steps of focus (which can be operated with the mirror up) and translate that back to an adjustment setting before writing it to firmware. That would probably require a database with lens specific translations.

If FoCal indeed does intermediate writing of the AFMA values, another issue would be hysteresis in the focus mechanism. It can make a difference if an AF operation approaches from the minimum focus distance position, or from infinity. So if the writing cycle goes from -20 to +20, then any subsequent AF starting at infinity might be off a bit, for some lenses more than others. I've seen that behavoir myself when doing close up phototography, where I approached optimal focus step by step, but when switching to the other direction because I went too far the whole image was out-of-focus at the first click.

There is a more manual approach available in FoCal that allows images to be taken and then analysed by the software to provide a graph of MA value versus contrast (read sharpness).

Indeed, and it would be less of a stress to the firmware which probably also has a limit as to how many write cycles it can stand. While a little less convenient than a fully automatic approach, it would theoretically also allow to process a full run in one direction folowed by a full run in the other direction, and average the results to reduce the hysteresis effect, before writing to the camera or proposing a value for manual input.

Cheers,
Bart
 

Bob Latham

New member
I doubt (although I only speculate) that mirror position is a pre-requisite for writing to the firmware. I think your later statement about lens motor increments is more likely to be the key. One unit of MA equates to 1/8th of the single sided depth of focus. This will not translate to lens motor increments across a range of lenses and possibly not even within the focus range of a specific lens (unless the focus motor increments are not linear with respect to distance).

One issue with MA adjustment concerns zoom lenses and my 70-200 ISL MkII highlights this by needing MA values a full 10 units apart (7 units from my Lens Align tests) between the wide and tele end of things. It would seem that the impending 1Dx addresses this problem by allowing multiple MA values for a single lens.

Bob
 

Ben Rubinstein

pro member
Hi Ben,

The focus of the 1Ds3 can be controlled via a USB tethered cable. The EOS Utility, in Live View mode, allows to manually focus the camera, so I presume that the SDK also offers that functionality.

Perhaps they updated EOS Utility because you couldn't control manual focus of the camera from the program (I tried) back when I owned the 1Ds3.
 
Top