EF Lenses on R Mount (f/1.2 portrait primes)

Marco Nero wrote:

CESA wrote:

Questions now:

A. Having to adjust the lens this bad, meaning to such a high value is a bad sign?

I'm inclined to think so. My first copy of this lens needed me to shift the calibration to the rear (*due to terrible backfocusing issues) ... all the way to the limits of the MFA settings. I showed the Canon dealer that sold it to me and he agreed there was a problem... because If I later attempted to sell the lens to someone using say an EOS 60D, they wouldn't be able to use the lens because they can't calibrate that specific model of camera since it has no MFA setting. The lens I swapped it for required just one or possibly two notches or so from 0 (zero). I still needed a minor calibration but the different from the first lens was like night-and-day.

Thanks Marco. The thing is that I bought it two months ago or something and it is out of that return time we have.

I bought it from an used shop in mint condition at looks like new. The price was around £750.

I can try to see if I can return it but will then have to wait for another opportunity to get another one.

The shop gives 1 year of warranty and I don't know if I can trigger that and pledge that the camera cannot AFMA properly with this lens. Would this help on anything?

Please see comments below.

B. Would be prudent to return the lens?

I'd be inclined to do so - especially if the only way to gain focus properly is by pushing the MFA settings to the furthest settings.

Yeah, that's what I saw from a crude test. Will re-do again the test. Could this be a problem with the lens itself? Is it like it has passed control but because it was inside the tolerance it was ok?

C. Do you think I should test at another distance? Like 4m or 5m?

I's be inclined to make your calibration again if you were not satisfied with the results. Because if you are not confident in the lens, you won't likely use it again very often. This should be the best lens in your collection so getting the calibration right using flat targets (like batteries) ought to make it possible to set the camera to match the lens. Be wary of using curved AF targets (like tube shaped AA batteries)... try to use flat targets that are not highly reflective. Ensure the surfaces have something with contrast for the AF to lock onto and keep them aligned so that they are not tilting away from the lens. You should put the lens cap (or your hand) in front of the lens between shots to shift the focus to "infinity" (or close enough) . Then take your hand away and allow the camera to focus on the target and take you test shot. This ensures that the focus is forced to travel quite a bit before locking onto the target. If you were to just shift focus between various targets that are close to each other, you may not be able to tell if the calibration is working or not.
.
Focus will be more critical at closer distances where the DOF is shallower so don't go too far. I tend to shoot a lot at minimal focus distance (MFD) a lot with this lens so I had to ensure that my lens was calibrated properly with the MFA feature. At greater distances the AF should be producing good results because the plane of focus ought to be greater. Up close it will be very narrow.
.
However, I had calibrated for this lens a number of times, only to have to re-calibrate the camera again when I took pictures and the AF wasn't nailing the shot perfectly. This issue really does only relate to f/1.2 lenses. Put any of these same lenses onto an EOS R camera and you're nailing the AF every time. It's great that you can use this lens on mirrorless cameras although the 85mm focal length, especially at f/1.2, is not as impressive on an APS-C camera as it is on Full Frame models.

If it happens that after the calibration I get a value that brings the focus to the correct plane of focus using the flat target.

I am using focal target and will be using focal software to perform the calibration as well.

In my experiment I have put the lens to infinity manually and took 3 different shots.

I see, the closest distance I shoot is around 1m or so. Normally 3m so will be doing for both but first when I use the focal will do for the recommended distance.

But Nero two other questions:

1. If the lens turns out to have and adjustment around +17 does it means that there is some mechanical or electric mechanisms that is failing?

2. If it turns out to ave that calibration, does it mean that in another camera DSLR like 5DMKIV will have that same adjustment?

3. Even if this lens has got this adjustment on a DSLR, if I go to EOS R or mirrorless camera the lens will not need any adjustment but the fact that had to be adjusted on the DSLR will it likely mean that it will potentially have a problem in the EOS R or mirrorless camera? Do you know what I mean?

4. Imagine that he lens is sent to Canon- can they bring and fine adjust the lens itself if it is a problem with the lens in the sense that although it has passed the control it was close to the limit spec, they can bring it to the best spec which is supposed to be close to 0% if the tolerance is say +-5%. Do you know what I mean?

Thanks in advance for your time.

, Marco Nero wrote:

CESA wrote:

But Nero two other questions:

1. If the lens turns out to have and adjustment around +17 does it means that there is some mechanical or electric mechanisms that is failing?

I don't think the mechanism is failing, but if you require a considerable adjustment, it means that the lens is within spec but needs a calibration to be applied. If you feel that +17 is too much, then you'd be right thinking it's pushing the boundaries of what is acceptable from Canon.

Yes, understood. I took the liberty and called Canon support and asked to speak to a technician and he said precisely that. That I shouldn't worry about that.

I kinda asked him if in the advent, with usage, the camera needs more AFMA points due to the lens usage over the years would it be possible to send it back and have it calibrated to factory standards or within spec and he said that it isn't and that I shouldn't worry about that.

2. If it turns out to ave that calibration, does it mean that in another camera DSLR like 5DMKIV will have that same adjustment?

Each camera sensor will be slightly close or further away from the back of the lens. So each camera will need to be calibrated for the lens. It's probable that the distances will be similar but they won't likely be exactly the same. When you make an MFA adjustment calibration, it has no effect on the lens, only the body of the camera being used. Since each camera has its own limitations and variations, you will have to calibrate each camera with an f/1.2 lens. The MFA adjustment will be similar between cameras using the same lens... but not exactly the same.

Yeah, I understand.

3. Even if this lens has got this adjustment on a DSLR, if I go to EOS R or mirrorless camera the lens will not need any adjustment but the fact that had to be adjusted on the DSLR will it likely mean that it will potentially have a problem in the EOS R or mirrorless camera? Do you know what I mean?

I understand. But with mirrorless, the sensor is used to calibrate the focus. Remember your MFA adjustment with a DSLR + lens requires a shot taken with AF using the Live Preview? Remember how this is your test shot to be used when comparing it to the OVF shot? The LCD shot ought to be in perfect focus. The same applies to the use of an EOS R (or any mirrorless camera with the same lens). All shots taken with mirrorless cameras should be problem-free. The AF will also be more accurate with cameras like the EOS R because of the new DPAF sensor.

Yeah, read that on another places and tried to get some more info over youtube from people that own the EOS R and this lens and they say it works perfectly.

Actually now on my second attempt to perform the test, I have shot with Live view and the image was perfect.

4. Imagine that he lens is sent to Canon- can they bring and fine adjust the lens itself if it is a problem with the lens in the sense that although it has passed the control it was close to the limit spec, they can bring it to the best spec which is supposed to be close to 0% if the tolerance is say +-5%. Do you know what I mean?

When you send in your lens to Canon, they will insist that you send your cameras in with the lens. They calibrate the cameras for each lens. They use the MFA to do this. There's nothing inside the lens that they can modify. There is a microprocessor inside each lens that communicates with the camera and this can be replaced if it fails. The lens elements can also shift if the lens is struck or dropped. Sending any lens into Canon almost inevitably results in it being returned with a different problem... like corner softness or softness at specific focal lengths. Only send a lens in for repair if it has been damaged.
.
Professional photographers often send their cameras and lenses in to Canon to be calibrated using a Calibration tool similar to the FoCal one. Again, the process used is essentially the one you can do yourself at home. It's extremely rare to get a defective lens when you buy one brand new because Canon have a testing process for every camera and lens before they leave the assembly floor. Anything that fails to test correctly is re-purposed or recycled for parts. If you calibrate your lens using the MFA tool in your DSLR - and it still doesn't lock focus accurately - return it. If it does appear to focus correctly afterwards for 96% of your shots, then the lens is probably fine. Put that same lens on an EOS R camera and it will probably be 99.9% to 100% realiable. Bit if there's a fairly high rate of out-of-focus shots still coming from it after FMA on your DSLR, return it.

I see. It ends up always in the same road.

I have done again the test and here are the results:

1st picture: +18 on the left and +17 on the right. This shot here was done at 2.5m aproximately. But it matches with the camera at 1.15m from the target. Bear in mind the following:

Haven't used the software just their chart

the paper was hung on the wall those small things you can see.

Have tried to put as best as possible the camera parallel to the target.

Mirror lock, always defocused to infinity, timer 2sec with remote, tripod using now 3 sections with the neck of the tripod left on the minimum.

Left +18, right +17

Left +18, right +17, center of the frame

Left +18, right +17, center of the frame zoomed in 2:1

No the same thing but on the left I will +19.

Left +18, right +19

Left +18, right +19, center of the frame

Left +18, right +19, center of the frame but zoomed in 2:1

As you can see my perception is that +18 is somewhat the sweet spot.

What do you guys thing? Thanks