Reikan FoCal doubts

Messages
31
Name
Charles
Edit My Images
Yes
Hi all.
I got this software to perform calibration on my lenses specially for the 85mm f/1.2L II. However, I am finding the results for the 85mm not in line with what I was expecting, that is, the AFMA that I got does not serve my needs. I will explain.
I have performed the calibration as it is suggested with the exception of the target in which the one I used was a printed version of the physical targets that they sell. I have performed the calibration just out of curiosity whilst still waiting for the delivery of the targets but don;t know if it will make a big difference.
I have used a very well lit room with artificial lighting. The shutter speed was around 1/125s.

The recommendation to perform the test for the 85mm is around 3.5m but because on the room I was I didn't have enough space I did it at 2.9m and also tried at 2.1m.
Picking up the 2.9m test the AFMA came out to be +5.
Went out and tried some shots. Being a portrait lens you expect to take picture at 1.5m or so from your subject if you want shoulder or half-chest. Somewhere around that. Not a surprise the images didn't come out in focus. On the other day I went out and did some shots wide open and also at f/2 to test the lens out and the images were not in focus as well. Felt like it just needed a couple of mm to be in focus. At that time the lens had +5 in the AFMA settings but that is because one day, to my eye, when I went out to try the lens for the first time I have put it at that setting because the image looked better in terms of focus. It was strange the software returning that same setting. But that is a question that I have, does the software do anything else besides setting the AFMA value? If so, ok. If not probably putting there +5 would make sense.

Nevertheless, after the calibration like mentioned above did some shots to try the new AFMA (was assuming the software did more that just putting the AFMA setting) and the images were not in focus.

In the past I have had tested the lens manually at a given distance, closer that 3.5m, and the AFMA came out to be +17. I did change the +5 to +17 and repeated the shots and they were bang on.

Can someone tel me what is going on here?

I question the efficacy of this software. Probably for people that use their lenses for targets well above the minimum recommended distance probably this isn't an issue because the far your subject is the wider your DoF is. Probably this might work very well for you.
What about for those that use lenses for distances below the calibration distance?
Is this normal I am experiencing?
Does really FoCal make a difference?

Just as a side note in regards to future calibration test that I will be doing:
1. Increase lighting to have high shutter speed.
2. Perform the test at the recommended distance.
3. Repeat the test at around 1.5m from the target.

Another thing. When I performed the AF consistency several times the results changed a bit. One time it was basically 100%, another time it was on the lower end of the tolerance, on the other test it was bellow the minimum of the range another time it was within the range, etc. Is this normal and expected for some reason?

One more thing. When performing the test the astigmatism was out of the optimal value although below 10%. It was averaging 6%. Probably there are a lot of variables that come in to play from the positioning of the target against the sensor and all that. But one thing I have noticed on the AF consistency check the astigmatism was lower like 3/4%. Is this normal or expected?

Any comments and advice? Thanks in advance.
 
Messages
9,575
Edit My Images
No
As you mention:-

Using Canon EOS 6D with 85mm lens.
The recommended minimum target distance is 3.4m (approx 11 foot 1 inches)

I can only surmise they state a "minimum" target distance because that can have an influence on the accuracy of the methodology. By not using the recommended minimum target distance..............who is to say what the effect might be???
 

Canon Bob

Loves the Enemy
Messages
10,612
Name
Bob
Edit My Images
Yes
As you mention:-

I can only surmise they state a "minimum" target distance because that can have an influence on the accuracy of the methodology.
When Canon introduced AFMA they advised calibrating at a distance of 50x the focal length
 
Messages
16,614
Name
Hugh
Edit My Images
No
Does really FoCal make a difference?
Focal is a bit of software that adjusts the micro AF for you. Adjusting the micro AF makes a big difference. Used well Focal works but although in can be flakey, you also need to work within its abilities.
 
Messages
3,367
Name
Mark
Edit My Images
Yes
I do find it a bit flakey but have best results when:

I use it outside when it’s bright but not direct sunlight (sorta bright/overcast) if there’s such a thing.

I test it at the distance I intend to use it so in this case it would be 1.5m for you and ignore the on screen warnings.

I repeat the tests for consistency.

It doesn’t matter what target you use, you can actually use any thing that stays still e.g a brick wall
or distant castle even! It’s just a contrast test for the software really!

The problem with Nikon and Canon lenses is you can only dial in one camera to subject distance, whereas Sigma and Tamron are considerably better in that you can dial in 4-5 and the software interpolates the inbetweens, they do tend to be more ‘off’ to begin with though.
 
Last edited:
Messages
812
Name
Phil
Edit My Images
No
Hope this isn't a silly question, are you using single point AF when out shooting. It is also worth bearing in mind that the DoF at f/1.2 at 1.5 metres is roughly 2cm (1cm forward and back at that distance) and that the actual AF sensor is larger than the indicator in the viewfinder, even the slightest movement either by the camera or subject will have an effect on the focus.
 
Messages
35,780
Edit My Images
No
I do find it a bit flakey but have best results when:

I use it outside when it’s bright but not direct sunlight (sorta bright/overcast) if there’s such a thing.

I test it at the distance I intend to use it so in this case it would be 1.5m for you and ignore the on screen warnings.

I repeat the tests for consistency.

It doesn’t matter what target you use, you can actually use any thing that stays still e.g a brick wall
or distant castle even! It’s just a contrast test for the software really!

The problem with Nikon and Canon lenses is you can only dial in one camera to subject distance, whereas Sigma and Tamron are considerably better in that you can dial in 4-5 and the software interpolates the inbetweens, they do tend to be more ‘off’ to begin with though.
Personally I would stick to the recommended target and target distances.
 
Messages
3,367
Name
Mark
Edit My Images
Yes
Personally I would stick to the recommended target and target distances.
I wouldn’t because that’s a ‘catch all’ or compromise, I know why it’s recommended but if you do all your shooting close up there’s very little point in calibrating further away, again a deficiency of the nikon/canon lens system that they can’t be calibrated like the third party lenses can.
 
OP
C
Messages
31
Name
Charles
Edit My Images
Yes
Hi all, thanks for your replies. Will try to answer everyone's comments and suggestions.
As you mention:-

Using Canon EOS 6D with 85mm lens.
The recommended minimum target distance is 3.4m (approx 11 foot 1 inches)

I can only surmise they state a "minimum" target distance because that can have an influence on the accuracy of the methodology. By not using the recommended minimum target distance..............who is to say what the effect might be???
I understand why there is a minimum recommended distance if we look at FoCal chart where they explain why. This does not make sense to me otherwise the minimum shooting distance should be precisely that one and not, for example in this case 0.95m. Calibrating a lens for that distance and then when trying to use it at a closer distance and not having things in focus is non sense. So there is something wrong here in regards to this unless the lens or camera has got a problem in regards to the AFMA and the distances you shoot at.
When you calibrate telephoto lenses it is very easy because when you shoot at a very far distance your DoF is bigger.

When Canon introduced AFMA they advised calibrating at a distance of 50x the focal length
Would like to know if anyone in here has done that and does not experience any of these problems when shooting a closer distance than the minimum recommended one. For this lens or other in particular with very big aperture.

Focal is a bit of software that adjusts the micro AF for you. Adjusting the micro AF makes a big difference. Used well Focal works but although in can be flakey, you also need to work within its abilities.
I understand that but if it a bit flaky it is useless because my expectation is high given the advertisement - it should be a software that would work for which is not the case even in regards to being fully automated (6D alows this) as it is always saying that it loses connection or whatever and we are required to go to the AFMA menu and change manually the adjustment to allow the software to proceed. That's obviously a limitation of the OS windows and their software API and communication protocols.
Probably one is better of using David Abbott's procedure to do the lens calibration manually. You can easily see if you have back or front focus then go and adjust your AFMA till you get the same plane of focus as the shot, shot with Live View.

Not time consuming at all.

I have purchased the software to take this doubt out of my head because this lens was/is requirement +18 to shot at 1m to 1.2m of distance. The software didn't do any difference because although not following the minimum distance because wanted to calibrate at working distance of 1.5m the software was not able to conclude anything. Had to stop the calibration at that distance. More on that below.

Wait for the proper target, follow the instructions in regard to light and distance.
That's defensively what I am going to do. Since you have mentioned this can I ask you or anyone in this forum what is the best way to get the camera squared to the target? How can we make sure that we have it square? Any methodology here that anyone uses to get it right that can share?
In regards to the distance will tell more below. Check that. Don't think that the distance of 3.7m recommended is viable because although the test was not done with their target I managed to reach +9 (against +5/+6 at the commended distance) which worked well at 3.7m and 1.5m. Didn't try at 1m (roughly the minimum focusing distance the lens is capable of and quite handy to even increase the bokeh which is what this lens is all about).

I do find it a bit flakey but have best results when:

I use it outside when it’s bright but not direct sunlight (sorta bright/overcast) if there’s such a thing.

I test it at the distance I intend to use it so in this case it would be 1.5m for you and ignore the on screen warnings.

I repeat the tests for consistency.

It doesn’t matter what target you use, you can actually use any thing that stays still e.g a brick wall
or distant castle even! It’s just a contrast test for the software really!

The problem with Nikon and Canon lenses is you can only dial in one camera to subject distance, whereas Sigma and Tamron are considerably better in that you can dial in 4-5 and the software interpolates the in between, they do tend to be more ‘off’ to begin with though.
FoCal told me that the type of lighting isn't a problem. But canon notebook about AFMA calibration say to shoot with the lighting that you normally shoot. Curiously enough that notebook also mentions that you can calibrate your lens at your working distance or if not possible or you don't know where to start then use this x50 rule.

I tried to use FoCal at 1.5m and it was not able to reach any conclusion hence why I said that this software might not make a difference for certain lenses.

I understand the bit of contrast and I quite like their target actually although haven't figured out which one is best for each lens - the 210mm should be used for telephoto lens and the 150mm for short tele to wide?

The problem with third party lenses is that they are not reliable or used to be. But do you want to take chance and in 10 years time if you change a body the lens might not perform as before or the lens might get broken earlier that it was supposed to given the price you paid? For Sony probably that is not a problem because protocols are not a problem they are open source so anyone can make a lens for that E-mount. And now on the mirrorless era for canon and nikon probably this issue with AF consistency or reliability might be gone because the type of AF these cameras use is different and you will always get in focus pictures without performing any calibration to the lens - in principle.

Hope this isn't a silly question, are you using single point AF when out shooting. It is also worth bearing in mind that the DoF at f/1.2 at 1.5 metres is roughly 2cm (1cm forward and back at that distance) and that the actual AF sensor is larger than the indicator in the viewfinder, even the slightest movement either by the camera or subject will have an effect on the focus.
Yes, I am using the single point AF but one don't need to bother with that because the software allegedly sets that up for you. Either way, I tens to use a sturdy tripod and only on the second attempt to perform the AFMA adjustment I have placed some weight on the tripod to make it even more sturdy given that when calibrating at 1.5m I was on a carpet.

Don't understand that bit of the AF sensor size and the viewfinder. Can you elaborate? Thanks in advance.

Nevertheless, that shouldn't matter because the closest distance this lens can focus is 0.95m so don't know if this is valid. But in terms of calibration the target does not move the only thing can move is the camera/tripod. When the targets arrive from FoCal I will use something like 5kg to 10kg on the tripod to make sure it does not move.

I wouldn’t because that’s a ‘catch all’ or compromise, I know why it’s recommended but if you do all your shooting close up there’s very little point in calibrating further away, again a deficiency of the nikon/canon lens system that they can’t be calibrated like the third party lenses can.
Canon's notebook about AFMA says that one can calibrate at the working distance so to say or if you don't know use this x50 rule. I tried to calibrate at 1.5m and the software was not able to reach any conclusion. Might think that or it was either because of the lighting condition (it was not very even but the S.S was around 1/250s (taking into account they overexposed by 1 stop otherwise it would have been 1/500s), target quality or at that distance the thing was not 100% squared with the target or it is more sensible to any of the variables to perform the calibration.

.

Now a bit more of an update on the second attempt:


I was doing basically everything that focal requires with the exception using their target so I printed out the target on the supermarket close by and did some experiments.

3.7m, 3.5m, 2m and 1.5m.

The length of the line has been introduced on the software.

The lighting they say it really doesn't matter if it is daylight or artificia l light. I have used a room with very good lighting overall although have added an additional lamp close to the target not making it even across the target but don't thing this would cause any problem.

The shutter speed was around 1/250m. The target distance of 2m and 1.5m were done on top of a carpet floor but have used a 2kg weight on the tripod to make it more sturdy.

Conclusion that I have reached after the second attempt.

The software couldn't reach any conclusion at 1.5m.

at 2m came up with AFMA of +9.

at 3.5m +6 and at 3.7m +5. So, at the recommended distance the AFMA setting can be said to be around that +5/+6.

I left this setting and shot another photo at 2m and 1.5m they were out of focus.

At 3.5m and 3.7m the test shots I did were good.

Switched back the AFMA to +9 and did some test shots at the above mentioned distances 3.7m, 3.5m, 2m and 1.5m and all of them were very nice.

Didn't do any AF consistency where the quality of the focus is assessed (the best one was close to QoF=2050 can't remember with which setting) at all those distances with the AFMA setting at +9.

When the FoCal targets arrive I will perform again all these tests with a more even light around the target besides the ones on the ceiling (there are two) and will test the QoF at the given AFMA setting that the software comes up with and see what do I get. Probably will do it before any AFMA calibration just to check and have a baseline although based in previous tests.

The AFMA of +9 seems to fit all distances but need to do some real life outdoor shoots to assess this.

At that setting the QoF on some quick shots inside the house to a person at 1.2m, 1.5m and 2m seemed ok. Will have to try it at 3.7m as well.

In the past I was using +18 because I did it on the fly at the working distance and the QoF was excellent at 1.2m, 1.5m and 2m. Will make a comparison as well on this +9 vs +18. Don't know if it is a false perception or what. Need to check the eyelashes and eye.

Will probably use the software to make this comparison with the target +9 vs +18 and 1.5m.

So the software was not totally useful here so basically like the other user said I had to work within its limitation.

I think that's it. Feel free to comment please.
 
Messages
3,193
Name
Steve
Edit My Images
Yes
I wouldn’t because that’s a ‘catch all’ or compromise, I know why it’s recommended but if you do all your shooting close up there’s very little point in calibrating further away, again a deficiency of the nikon/canon lens system that they can’t be calibrated like the third party lenses can.
Hi Brazo, your thought process is somewhat flawed but I understand why and how that can happen. The calibration distance isn't a catch all, it's a factor in the calculation for the correction and because the lens's calibration isn't linear you can't use just any old distance you like.

The company that I work for manufactures visibility sensors and we calibrate them to a distance of 900m. Once that calibration is done the sensor can accurately measure the visibility at anything from 10m to 100km. If you calibrate our sensors at anything besides 900m then it can't measure visibility at any distance because the software has a false basis in the calibration curve. Like FoCal we have a target for the sensor but in our case it simulates 900m visibility. Trying to explain to customers how this works takes up more of my time that I care to think about but in essense it's just the way we do it. Thankfully, our competitors also do it in a similar way but at different calibration distances :)

I would suggest that if you do it the FoCal way then I'm pretty sure that when you're shooting portraits at 1.5m (or any other distance) it will be fine. There are times when you just have to trust someone and go by the results :)
 
OP
C
Messages
31
Name
Charles
Edit My Images
Yes
Hi Brazo, your thought process is somewhat flawed but I understand why and how that can happen. The calibration distance isn't a catch all, it's a factor in the calculation for the correction and because the lens's calibration isn't linear you can't use just any old distance you like.

The company that I work for manufactures visibility sensors and we calibrate them to a distance of 900m. Once that calibration is done the sensor can accurately measure the visibility at anything from 10m to 100km. If you calibrate our sensors at anything besides 900m then it can't measure visibility at any distance because the software has a false basis in the calibration curve. Like FoCal we have a target for the sensor but in our case it simulates 900m visibility. Trying to explain to customers how this works takes up more of my time that I care to think about but in essense it's just the way we do it. Thankfully, our competitors also do it in a similar way but at different calibration distances :)

I would suggest that if you do it the FoCal way then I'm pretty sure that when you're shooting portraits at 1.5m (or any other distance) it will be fine. There are times when you just have to trust someone and go by the results :)
Please read my comment above and comment on it please. Thanks.
 
Messages
3,367
Name
Mark
Edit My Images
Yes
Sometimes I get so fed up of Focal that I simply shoot the target with live view (to remove focus errors) and then again using the cameras pdaf and varying afma figures and compare them to see which is closest to the live view image. Can be surprising quick and easy.

I recall using the auto afma feature on my d850 on a particular lens which gave a figure of -16 and then using focal which gave -15. I was really impressed both gave identical results. When I then dialled it in took a test photo it was completely out of focus! I tried my quick and dirty method above and + 8 seemed to match the live view pic And most importantly was accurate in real world focus tests.
 
Messages
3,367
Name
Mark
Edit My Images
Yes
Hi Brazo, your thought process is somewhat flawed but I understand why and how that can happen. The calibration distance isn't a catch all, it's a factor in the calculation for the correction and because the lens's calibration isn't linear you can't use just any old distance you like.

The company that I work for manufactures visibility sensors and we calibrate them to a distance of 900m. Once that calibration is done the sensor can accurately measure the visibility at anything from 10m to 100km. If you calibrate our sensors at anything besides 900m then it can't measure visibility at any distance because the software has a false basis in the calibration curve. Like FoCal we have a target for the sensor but in our case it simulates 900m visibility. Trying to explain to customers how this works takes up more of my time that I care to think about but in essense it's just the way we do it. Thankfully, our competitors also do it in a similar way but at different calibration distances :)

I would suggest that if you do it the FoCal way then I'm pretty sure that when you're shooting portraits at 1.5m (or any other distance) it will be fine. There are times when you just have to trust someone and go by the results :)
Does your system use an algorithm to fill in the gaps though a bit like the Tamron/Sigma docks?

Im not sure the Nikon version is that complex, I always fine tune for the distance I use the lens at (admittedly) it’s not at 1.5m but I can get it bang on (eventually)! It would after all be a bit silly to fine tune a 100mm macro lens at 50 x 100mm when you use it two inches from the subject.
 
Last edited:
Messages
6,784
Name
Bazza
Edit My Images
No
Can one get over critical on focus ?. I have the Spyderlenscal and it was driving me nuts to be honest. To fine tune on my nikon cameras it involved the right distance from target to focal plane -both being on the same level and faciing each other square on- correct lighting on target- right camera settings - remote shutter release trigger to lessen camera shake by button pressing. Checking every shot and adjusting fine tune, doing the same with each lens and again with each camera (3). In the end it started taking away the fun in photography. No stuff all that I am getting better results by just looking at the rear screen with the ability of the push of a button to check/adjust the focus manually in closer up view on the rear screen.
I don't do portraits of pro shoot but to my mine if focus is so critical editing can sort most of that out anyway. Now I am back to just enjoying taking potos and keeping the ones I like, Don't really care if other like them but it is a bonus. If they want better photos let them go buy their own camera equipment
 
Last edited:
Messages
35,780
Edit My Images
No
As for 'square-on' to the target, I have always fixed (i.e. double sided sticky 'blobs') my target to a solid object, e.g. the front porch uPVC door and had the camera mounted on a tripod, manually adjusted to give 'square-on' to my eye and then did the first part of the test which allows the software to decide if the target is set in such a way that it can perform the test ... in which case it gives me the green tick.
Being as the target is in the fully glazed porch it has maximum daylight, though even then I would test in the best light of the day.
I would not consider performing the test inside under 'normal' artificial light, if I did it indoors it would be with the target illuminated by something like a 500watt floodlight ... Reikan say that artificial light is okay ... it is ... but it needs to be sufficient to illuminate the target very well.
The viewfinder curtain must be over to eliminate light leak, it should be set to single point focus and VR must be off.
There will invariably be some slight variation, as there will be slight variations in the test (light variation, slight vibration from moving, firing the shutter etc) but it should be easy to establish an average from fairly close results if the test is done properly, (of course there will be different issues with a zoom lens depending on manufacturer body and the ability/non-ability to test different ranges in which case an educated decision must be made as to which end of the range you will need most).
The test chart is critical to good results, the Focal Test Chart has been designed to provide the contrast that the software has been designed to detect and work with ... use a piece of paper and there will inevitably be 'flexibility', use the Focal Test Chart with its solid backing and that 'flexibility' won't be there.
 
Messages
9,619
Name
Jeff
Edit My Images
No
a few years ago I was having a chat to one of the tech guys at a well known lens makers about this and he said the method they used was to take the lens to a car park and aim at a car number plate that was at right angles to you i.e look along a line of cars and aim at one ,use the centre letter of the number plate and you will then see if your front or back focussing easily adjust it till correct ,sometimes the KISS solution works best
 
Messages
812
Name
Phil
Edit My Images
No
Don't understand that bit of the AF sensor size and the viewfinder. Can you elaborate? Thanks in advance.
In the viewfinder there is a square indicating the focus point in use, the AF sensor is in fact larger than that share so may pick up something just outside the square that has better contrast.
Either way, I tens to use a sturdy tripod and only on the second attempt to perform the AFMA adjustment I have placed some weight on the tripod to make it even more sturdy given that when calibrating at 1.5m I was on a carpet.
Not how you use FoCal but how you go about testing the focus after calibrating the MA...
 
OP
C
Messages
31
Name
Charles
Edit My Images
Yes
Thank you all for your comments. I am waiting for the FoCal targets so once that comes I will be re-do the test and come back with news.
 
Messages
6,784
Name
Bazza
Edit My Images
No
Out of interst maybe just now I did this

With my nikon D810 and Nikon 70-200mm lens @200mm I did a test on my garden arch 50ft away with the camera in quiet mode on Aperture set at f2.8 spot metering,, the camera on a sturdy tripod and remote shutter release.
The test was to take 3 photos fine tuned settings at +20 zero and -20.
With the photos now on the computer monitor I viewed all 3 at once using "open with photos" in the drop down list . Not hard to get all three photos on the monitor screen at once so won't go into this simple exercise
.Once opened there is the sliding scale to enlarge . This is what I did and had all three photos up at once. then i checked for clarity and the fine tune on zero suprised me by being the sharpest.

A focus check at that distance and a house roof a few hundred yards away again was really sharp. far better than any gadget one can buy and works with any lens or camera that has fine tujne in the menu setup
 
Last edited:
Messages
3,367
Name
Mark
Edit My Images
Yes
Out of interst maybe just now I did this

With my nikon D810 and Nikon 70-200mm lens @200mm I did a test on my garden arch 50ft away with the camera in quiet mode on Aperture set at f2.8 spot metering,, the camera on a sturdy tripod and remote shutter release.
The test was to take 3 photos fine tuned settings at +20 zero and -20.
With the photos now on the computer monitor I viewed all 3 at once using "open with photos" in the drop down list . Not hard to get all three photos on the monitor screen at once so won't go into this simple exercise
.Once opened there is the sliding scale to enlarge . This is what I did and had all three photos up at once. then i checked for clarity and the fine tune on zero suprised me by being the sharpest.

A focus check at that distance and a house roof a few hundred yards away again was really sharp. far better than any gadget one can buy and works with any lens or camera that has fine tujne in the menu setup
Yup a quick and dirty test can often be a better starting point. In your circumstances If ‘0’ was the best I’d then try +/- 5 and then narrow the parameters even further.

Im not a fan of the square on test because I don’t take square on portraits in the real world, the important thing is the test is consistent.
 
OP
C
Messages
31
Name
Charles
Edit My Images
Yes
Out of interst maybe just now I did this

With my nikon D810 and Nikon 70-200mm lens @200mm I did a test on my garden arch 50ft away with the camera in quiet mode on Aperture set at f2.8 spot metering,, the camera on a sturdy tripod and remote shutter release.
The test was to take 3 photos fine tuned settings at +20 zero and -20.
With the photos now on the computer monitor I viewed all 3 at once using "open with photos" in the drop down list . Not hard to get all three photos on the monitor screen at once so won't go into this simple exercise
.Once opened there is the sliding scale to enlarge . This is what I did and had all three photos up at once. then i checked for clarity and the fine tune on zero surprised me by being the sharpest.

A focus check at that distance and a house roof a few hundred yards away again was really sharp. far better than any gadget one can buy and works with any lens or camera that has fine tune in the menu setup
Would like to see the outcome when you focus on the subject on the minimum focusing distance with you adjustment at 0 and at 200mm. Come back with the results.

I think I will have to open a new thread to ask in this forum how has got the same lens I Am testing out to exchange ideas. I am finding very difficult to reach a conclusion using FoCal. Today did the 5 times the calibration and got different results following all the recommendation. However, jJust after cleaning everything up I realised that I forgot to cover the eye piece. Damn. Will re-do the test in a couple of minutes but this time full artificial lighting just to check if that inconsistency still maintains despite the fact that will be doing the test again with artificial light.

I don't want to believe that it is a problem with the lens since I bought it from the used shop (camera jungle) where the items was marked as mint. These lenses are built like a tank but anything can happen.
 
Messages
6,784
Name
Bazza
Edit My Images
No
Whatever way you focus as long as you like the result is the correct way for youj
 
Last edited:
OP
C
Messages
31
Name
Charles
Edit My Images
Yes
Whatever way you focus as long as you like the result is the correct way for youj
The problem is that if I calibrate the lens by the book I don't get the results I want. If I follow FoCals recommendation to calibrate the lens at 50x the focusing distance, when I go close to the minimum focusing distance the thing isn't in focus pretty much or say it is a bit off. It shouldn't because it was calibrated. I think that all this thing with calibration is a bit of a lie at least for this lens. Don't know if it is due to the nature of the lens or what. This lens is for portraits. You don't take portraits at 3.7m - I mean you might but probable not most of the times. It depends on your style essentially ok. On my case I would be using it at 1.5m to 2m. Sometimes 3m. So you calibrate at 3.7m and if you go and shoot at 1.5m the focus is out. How come? Something is wrong here.
The problem is that I don't know anyone who has got this lens that can share their experience.
 
Messages
812
Name
Phil
Edit My Images
No
I don't want this to sound wrong, but, you have not answered how you are confirming the results of your FoCal calibration... I don't own the 85 f/1.2 but do own the 135 f/2 which is a very similar beast. The DoF is minuscule at a wide open aperture, so unless you are on a tripod with a totally static subject any slight variation is going to result in a less that satisfactory result.
I use continuous AF and have at least a 90% rate of keepers.
 
OP
C
Messages
31
Name
Charles
Edit My Images
Yes
I don't want this to sound wrong, but, you have not answered how you are confirming the results of your FoCal calibration... I don't own the 85 f/1.2 but do own the 135 f/2 which is a very similar beast. The DoF is minuscule at a wide open aperture, so unless you are on a tripod with a totally static subject any slight variation is going to result in a less that satisfactory result.
I use continuous AF and have at least a 90% rate of keepers.
Are you asking about how did I do the test with FoCal? If so, it was basically how the checklists mandates.

Tripod (vanguard SBH-100 with Alta pro 263T) with 6kg on the bottom to make sure it was still. Tested at 3.7m and 3.5m also 2m and 1.5m. Good lighting since the S.S was around 1/320s although artificial (the test from today using daylight forgot to cover the eye cap. Have used FoCal foam targets but the results are invalid because of the eye cap). The target was a printed target from a supermarket printer. Target parallel and as squared as possible to the camera. The eye piece on this second attempt was covered. That's it.

If you are asking about how I am confirm the results after calibration the answer would be random shots object and people.

'at 2m came up with AFMA of +9.

at 3.5m +6 and at 3.7m +5. So, at the recommended distance the AFMA setting can be said to be around that +5/+6.

I left this setting and shot another photo at 2m and 1.5m they were out of focus.

At 3.5m and 3.7m the test shots I did were good.

Switched back the AFMA to +9 and did some test shots at the above mentioned distances 3.7m, 3.5m, 2m and 1.5m and all of them were very nice.

Didn't do any AF consistency where the quality of the focus is assessed (the best one was close to QoF=2050 can't remember with which setting) at all those distances with the AFMA setting at +9.

When the FoCal targets arrive I will perform again all these tests with a more even light around the target besides the ones on the ceiling (there are two) and will test the QoF at the given AFMA setting that the software comes up with and see what do I get. Probably will do it before any AFMA calibration just to check and have a baseline although based in previous tests.

The AFMA of +9 seems to fit all distances but need to do some real life outdoor shoots to assess this.'

The take away here is. If I was to use +5, result that came from testing at 3.7m by the book, at 2m/1.5m the images were off. Not much details or in another words you could see more details if using +9. This testing manually against the target and checking the pictures on the computer LR 1:1 and 2:1.

So what's the point of calibrating the lens at 3.7m if when I go and shoot at 2m or 1.5m, which isn't the minimum focusing distance, the images are off? Probably FoCal is more suitable for telephoto lenses. Nothing else. Dustin Abbott said in one of his tutorials that some of the results from FoCal he didn't agree.
This also leads me to believe that people that are happy withthe results are people that are calibrating medium to long telephoto lenses where the DoF is inherently higher. The likes of 200mm f/2.8 or 500mm f/4, etc.

Do you get my point? The ultimate test when I put the AFMA setting that came up with is to focus in someone's eye and see that it is acceptably sharp although the sigma 85 1.4 art that I have tested was razor sharp on my camera. Probably the internal calibration matched my AF system on the 6D. Even babies eye lashes were razor sharp.

EDIT: I use continuous AF then switched to one shot. Also I don't recompose at f/1.2.
 
Last edited:
Messages
812
Name
Phil
Edit My Images
No
Are you asking about how did I do the test with FoCal? If so, it was basically how the checklists mandates.
Nope
If you are asking about how I am confirm the results after calibration the answer would be random shots object and people.
Yes, I thought that was clear, but in your explanation of your confirmation you are still talking about altering the MA. Put simply when trying out your MA results are you handholding? Is your subject totally stationary?
This also leads me to believe that people that are happy withthe results are people that are calibrating medium to long telephoto lenses where the DoF is inherently higher. The likes of 200mm f/2.8 or 500mm f/4, etc.
Erm...
 
Messages
812
Name
Phil
Edit My Images
No
To add, from a review and what I was saying in a previous comment...
"Even though it excels at photographing test charts even at f/1.2, in real photography the depth of field is paper-thin, so at f/1.2 almost nothing is in focus. Your camera's autofocus system will be taxed to its limit and you'll have a tough time getting your intended subject in focus if you're worrying about lab-grade sharpness. It won't happen unless you're subject is flat, which is only test charts"
 
OP
C
Messages
31
Name
Charles
Edit My Images
Yes
Nope

Yes, I thought that was clear, but in your explanation of your confirmation you are still talking about altering the MA. Put simply when trying out your MA results are you handholding? Is your subject totally stationary?

Erm...
Pardon my English, It is not my first language.

Wait, when I mentioned about altering the MA I was doing it before getting the FoCal software. So that was more or less on the fly. For example: Subject at a certain distance - 2m, Then manually, adjusting the AFMA on the fly to get what I wanted in focus, in this case the eye. The starting point after fine tuning it was a cheapo chart from ebay that costed £5.

After I got the software I left the setting as it was determined by FoCal. It is not bad if I was +9 which works well at 3.7m and 2m however it is not at its full potential and probably my expectations were very high about this software. Having been told that it was the standard on the industry lead me to buy it just to test this lens.
Attention that +9 was the setting I got from testing the lens at 2m from the printed target with aritficial light aka ceiling lights and portable lights close to the target.

Today during daytime with ceiling light only I ran the test again but this time forgot to put a cover on the eye piece. Focal came out with +12 for 3.5m (didn't have more room in the division I was testing on.

From +9 to +12. Inconsistency. Light changed though. So this means that the QoF is affected by the light? It shouldn't. Could it be because of the lens? Maybe but can't compare since don't know anyone else that has got this lens. Then this time during daytime the AFMA settings were changing - did 5 times. +12 then +15 then +17 then +12 had to increase the lighting in the room because it was getting dark. Could light affect the calibration? If so, then artificial light is really the way to go since it is constant. Either way can't rely on the results since the eye piece was not covered. Ah, this time was using FoCal foam target.

To add, from a review and what I was saying in a previous comment...
"Even though it excels at photographing test charts even at f/1.2, in real photography the depth of field is paper-thin, so at f/1.2 almost nothing is in focus. Your camera's autofocus system will be taxed to its limit and you'll have a tough time getting your intended subject in focus if you're worrying about lab-grade sharpness. It won't happen unless you're subject is flat, which is only test charts"
If it is like that I am more relieved since it might not be a problem with the lens but how can I know. Hopefully when I get a mirrorless all these problems will go away but will probably never know if the lens has a problem.

Find attached an example of a picture taken at around 1m with +17 done manually and on the fly. I know things can't be done like this hence why I got the software. At longer distances like 2.5m was good the overall focus and shrpness.
 

Attachments

OP
C
Messages
31
Name
Charles
Edit My Images
Yes
Just an update from you guys. I have contacted FoCal and explained the problem and they said that this is a known issue:

The problem you see is a known issue and it's to do with the lens and camera focus system design.

One way to look at it, forget any focus adjustment (pretend the camera cannon be focus adjusted) - pick any lens, now use the camera auto focus near to the minimum focus distance, notice if there is front or back focus. Now take an image using auto focus at say 5x the minimum focus distance, notice if there is front or back focus.

What you will find is the camera focus system is designed to work within a 'sweet spot' - it is not accurate for all distances.

The recommended minimum calibration distance is based on information from Nikon and Canon, they assert (and we find it generally correct) that once you get out past this 'minimum calibration distance' the focusing is pretty consistent. Check the link (...)

As an aside, both Sigma and Tamron have tried to address this issue with the Sigma Dock and Tamron TAP based lenses which allow 20 different adjustment points for a zoom lens based on focal length and distance to subject.

Using a normal Canon or Nikon lens it's not possible to create adjustments based on subject distance, but knowing there is likely to be a difference depending on the distance (especially with very close distances) at least gives the chance to mitigate (using live-view focus for example)
So basically he is saying that we have to be around that minimum calibrating distance of 50x (strangely enough canon says minimum working distance or 50x if we don't know where to start).. Once we past that distance it is inconsistent. I wonder what about macro lenses. Also don't get why they have as a minimum focusing distance, a distance that is smaller than that minimum calibrating distance. This does not make any sense?

Any comments? Thanks.
 
Last edited:
Messages
354
Edit My Images
No
Hello,
TBH I'm not a fan of focal. I've seen it produce spurious results even when used by commercial operators. I might be wrong but I don't believe any of the major equipment manufacturers have recommended it either.

I would suggest taking some images, making an adjustment then repeat as required untill happy.
 

Canon Bob

Loves the Enemy
Messages
10,612
Name
Bob
Edit My Images
Yes
........Also don't get why they have as a minimum focusing distance, a distance that is smaller than that minimum calibrating distance. ......
You need to understand 'focus shift' (as I alluded to in your 85/1.2 thread) to make sense of the apparent disparity between MFD and recommended calibration distance.

Focus shift in wide aperture lenses creeps in as the aperture becomes larger and the magnification increases and is caused by uncorrected spherical aberrations....a shortcoming of the optical formula. These are the very same spherical aberrations that give the pleasing bokeh that make the lenses attractive to many users.

I'm sure that your lens will perform excellently for tight portaits at f/5.6 (typical studio shots) or at f/1.4 for 1/2 or full body shots (ie, lower magnification). You're pushing both limitations at the same time and something has to give.

This does not make any sense?
In your application that may be true but for other users, it is possibly not an issue. It's also worth remembering that the 85/1.2 was introduced 20 years before AFMA became available.
 
Last edited:
OP
C
Messages
31
Name
Charles
Edit My Images
Yes
You need to understand 'focus shift' (as I alluded to in your 85/1.2 thread) to make sense of the apparent disparity between MFD and recommended calibration distance.

Focus shift in wide aperture lenses creeps in as the aperture becomes larger and the magnification increases and is caused by uncorrected spherical aberrations....a shortcoming of the optical formula. These are the very same spherical aberrations that give the pleasing bokeh that make the lenses attractive to many users.

I'm sure that your lens will perform excellently for tight portaits at f/5.6 (typical studio shots) or at f/1.4 for 1/2 or full body shots (ie, lower magnification). You're pushing both limitations at the same time and something has to give.
Yes, you are right but at the same time what I don't get is why is that when you use live view the image is tack sharp then? The lens design has nothing to do with it. In here the mirror and in this the AF system/module of the camera is out of the way and everything is happening in the sensor. Unless the camera is really forcing the lens to whatever it has to do in order to the sensor get a focused imaged? Don't know. The spherical aberrations I understand that is the key thing that allows this lens to render how it renders but in terms of its performance close to the limitations like you are saying are not a problem in the mirrorless cameras apparently.

In your application that may be true but for other users, it is possibly not an issue. It's also worth remembering that the 85/1.2 was introduced 20 years before AFMA became available.
Yes, true. 14 years more precisely, when it was first introduced. Yes, the AFMA was introduced one year later on the 1DMK3, that is, in 2007.
 

Canon Bob

Loves the Enemy
Messages
10,612
Name
Bob
Edit My Images
Yes
Yes, you are right but at the same time what I don't get is why is that when you use live view the image is tack sharp then?
Using Liveview with sensor based contrast detect focusing (or mirrorless) eliminates focus shift because focus is achieved at the selected aperture and there is no 'post focus' aperture reduction. The second advantage is that focus is absolute and not simply to within 1/3 DoF. Using lenses with a max aperture < f/2.8 will see focus confirmation when within DoF. The AF sensors only receive 40-45% of the light passing through the lens as the remainder is needed for the viewfinder. This limited amount of light is the main reason that phase detect AF is performed wide open.

The lens design has nothing to do with it.
We'll have to differ on this point.

Yes, true. 14 years more precisely, when it was first introduced. Yes, the AFMA was introduced one year later on the 1DMK3, that is, in 2007.
Your dates are based on the release of the MkII version of the 85/1.2. The MkI appeared in 1989 (18 years before AFMA) and the MKII uses the same optical formula but with improved lens coatings and uprated electronics to achieve faster focusing (yes, the MKI was slower to focus).
 
OP
C
Messages
31
Name
Charles
Edit My Images
Yes
Using Liveview with sensor based contrast detect focusing (or mirrorless) eliminates focus shift because focus is achieved at the selected aperture and there is no 'post focus' aperture reduction. The second advantage is that focus is absolute and not simply to within 1/3 DoF. Using lenses with a max aperture < f/2.8 will see focus confirmation when within DoF. The AF sensors only receive 40-45% of the light passing through the lens as the remainder is needed for the viewfinder. This limited amount of light is the main reason that phase detect AF is performed wide open.


We'll have to differ on this point.


Your dates are based on the release of the MkII version of the 85/1.2. The MkI appeared in 1989 (18 years before AFMA) and the MKII uses the same optical formula but with improved lens coatings and uprated electronics to achieve faster focusing (yes, the MKI was slower to focus).
Thanks for the clear explanation. Can't wait to move to mirror-less just to get rid of this issues with the AF.
Yes Bob, was referring to the MKII.Can't remember if the optics are the same to the MKI version.

Cheers.
 
Top