My RF 70-200 f2.8 is front focusing

Started Dec 3, 2019 | Discussions
R5D4
R5D4 Contributing Member • Posts: 838
My RF 70-200 f2.8 is front focusing
13

Sounds impossible, right?

I was enjoying my new RF 70-200, acquired over thanksgiving break on a vacation (and away from my test target and tripod) when I saw this thread over at FM Forums:

https://www.fredmiranda.com/forum/topic/1623630

Nonsense, I thought. My lens is awesome. The pictures it takes are awesome. It's impossible for a lens to front or back focus on a mirrorless camera anyhow, right?

Anways, the thread gnawed on me, and when I finally got home I got out the target and the tripod.

Son of a bi____. My lens does it too. At 200mm, near MFD, f2.8, it focuses about 3mm in front of the confirmed focus point:

200mm @ 2.8, near MFD

At 163mm, near MFD, 2.8. Looks to be about 2.5 to 3mm in front of the target:

163mm, @ 2.8 near MFD

At 135mm, the distance is shrinking. Probably 2mm

135mm @ 2.8 near MFD

At 100mm it looks more or less cleared up, but maybe still 1mm? I took a 70mm shot also in my first test and it seemed perfect.

100mm @ 2.8 near MFD

Each shot was 100% repeatable. I took about 10 shots for each of these, with the camera on a tripod, 2 second timer and the focus mode in single point. Every single shot was identical to the one before it. Also I zoomed in to 10x, focused and then was able to get an even sharper image by turning the focus ring.

I went back and looked at all the photos I had taken. There was some slight pop missing from a lot of portraits. Other photos I couldn't tell. And then I found this series of 5 images taken with Eye AF, all with the same miss by about 2-3mm in front of the pupil of the eye:

I am contacting Canon tomorrow and will send the lens in to be repaired. If it comes back not repaired, I plan to get a refund from the seller.

If you read the thread over on FM Forums, mine is something like the 8th lens with this same problem. There is a shop owner in that thread who tested 3 copies on both the R and RP. I really don't want to mess around exchanging this lens for another copy if the problem is universal.

I'm posting this as a PSA for anybody with this lens- you might want to take a closer look, STAT. I'll update here on what happens with Canon.

 R5D4's gear list:R5D4's gear list
Canon EOS R5 Fujifilm X-T3 Canon EOS R Canon EF 35mm F1.4L II USM Fujifilm XF 35mm F2 R WR +5 more
shawnphoto Senior Member • Posts: 1,282
Re: My RF 70-200 f2.8 is front focusing

pokesfan wrote:

Sounds impossible, right?

I was enjoying my new RF 70-200, acquired over thanksgiving break on a vacation (and away from my test target and tripod) when I saw this thread over at FM Forums:

https://www.fredmiranda.com/forum/topic/1623630

Nonsense, I thought. My lens is awesome. The pictures it takes are awesome. It's impossible for a lens to front or back focus on a mirrorless camera anyhow, right?

Anways, the thread gnawed on me, and when I finally got home I got out the target and the tripod.

Son of a bi____. My lens does it too. At 200mm, near MFD, f2.8, it focuses about 3mm in front of the confirmed focus point:

200mm @ 2.8, near MFD

At 163mm, near MFD, 2.8. Looks to be about 2.5 to 3mm in front of the target:

163mm, @ 2.8 near MFD

At 135mm, the distance is shrinking. Probably 2mm

135mm @ 2.8 near MFD

At 100mm it looks more or less cleared up, but maybe still 1mm? I took a 70mm shot also in my first test and it seemed perfect.

100mm @ 2.8 near MFD

Each shot was 100% repeatable. I took about 10 shots for each of these, with the camera on a tripod, 2 second timer and the focus mode in single point. Every single shot was identical to the one before it. Also I zoomed in to 10x, focused and then was able to get an even sharper image by turning the focus ring.

I went back and looked at all the photos I had taken. There was some slight pop missing from a lot of portraits. Other photos I couldn't tell. And then I found this series of 5 images taken with Eye AF, all with the same miss by about 2-3mm in front of the pupil of the eye:

I am contacting Canon tomorrow and will send the lens in to be repaired. If it comes back not repaired, I plan to get a refund from the seller.

If you read the thread over on FM Forums, mine is something like the 8th lens with this same problem. There is a shop owner in that thread who tested 3 copies on both the R and RP. I really don't want to mess around exchanging this lens for another copy if the problem is universal.

I'm posting this as a PSA for anybody with this lens- you might want to take a closer look, STAT. I'll update here on what happens with Canon.

Are we even sure Canon eye AF focuses on the pupil?

Are you using the latest firmware?

I haven’t been able to reproduce the error even shooting test charts.

 shawnphoto's gear list:shawnphoto's gear list
Canon EOS R Canon EOS RP Canon EOS R5 Canon EF 24-105mm F4L IS II USM Canon RF 28-70mm F2L USM +4 more
Battersea Senior Member • Posts: 1,072
Keep us updated
1

Let us know what Canon say and the repair outcome. Thanks.

 Battersea's gear list:Battersea's gear list
Canon PowerShot G16 Canon EOS 7D Canon EOS 6D Canon EF 50mm F1.8 II Canon EF 85mm F1.8 USM +6 more
gium Senior Member • Posts: 1,609
Re: My RF 70-200 f2.8 is front focusing
3

I also just read the thread at fredmiranda as well. Interesting find!

I have the Spyder Lenscal at home, but never thought of using it with my RP. I will check the focus of my RP and my EF 70-200/2.8 II at 200mm MFD, but I've read someone else has already done that and everything was fine. Good news for me, but I just want to be sure.

 gium's gear list:gium's gear list
Canon EOS 5D Mark IV Canon EOS R5 Canon EF 15mm f/2.8 Fisheye Canon EF 85mm F1.2L II USM Canon EF 24-85mm f/3.5-4.5 USM +8 more
Franz Kerschbaum
Franz Kerschbaum Contributing Member • Posts: 876
Re: My RF 70-200 f2.8 is front focusing

For macro work it is quite normal, that one has to work with a quite asymmetric focus distribution (simple geometric optics). Typically the linear range (in e.g. mm) of sharp focus behind the set focus is much larger than the linear range in front of it. That why you often manually frontfocus on purpose to have a symmetric focus distribution in macro work. Maybe canon implemented this in purpose? Maybe they overdid it?

 Franz Kerschbaum's gear list:Franz Kerschbaum's gear list
Canon EOS 5D Mark II Canon EOS 7D Mark II Canon EOS M6 Canon EOS Ra Canon EOS R5 +23 more
lawny13 Veteran Member • Posts: 3,029
Re: My RF 70-200 f2.8 is front focusing
6

Sorry, but you are mistakened.

Just because  you have a MILC system doesn't mean that you won't have focus issues. I ran into this problem on sony as well. Generally you will find it in every MILC. 
Basically mirrorless removes the need for micro focus calibration. This is only with respect to the correction of the image plane of the focus sensor and that of the sensor. However, that is only part of the factor. 
I will use the the eye image as an example. The camera detects a face, then from there an eye. It places an AF box on the eye. As yourself what is in that box. The eye itself, as in the pupil, eye lid, and eye lashes. So that is what it is focusing on. So like when a camera is focusing on full auto it tends to grab onto the closest thing in the fame with good enough contrast. Well the same is done here with what is in the frame that is closest are the eye lashes. 
For your ruler example. You place the box on the 0 for example. But it seems to front focus. Well if I were to be looking at your camera, and the AF box, it would likely be focusing on the button of that AF box with corresponds to the closest elements to you in that box. 
So... it is working perfectly, though I am sure your expectation is that it prioritizes the center most portion of the AF box? I agree that this should be the case actually, but it is obviously not what is done. 
I have had experiences with sony MILC setups where the camera would constantly grab onto the background of the subject I was trying to capture, because the background also happened to be partially in the AF box. It was quite annoying, and I had to revert to manual focus.

The thing is that a DSLR like the 5DIV and the 7DII have the option of that extra small AF point. That was better for precise focusing (minus the focus sensor mirco adjustment thing). Though MILC removes the mirco adjustment problem it introduces other things. Canon's AF box contains a whole like of pixel size AF points It is unclear which are uses, and how they are prioritized though closer to camera seems to be the case based on what you show. It would be nice if the box could be made even smaller for more precise things. Or in the case of eye AF that the pupil itself is also prioritized, or that it does a center of AF weighted AF. 
Sorry for the rant, but I am fairly certain that there is nothing to "repair" as far as your lens is concerned. This AF issue has thing to do with the hardware (lens or camera) and more to do with software. Canon has some room for improvement in that department.

Franz Kerschbaum
Franz Kerschbaum Contributing Member • Posts: 876
Re: My RF 70-200 f2.8 is front focusing
1

Nobody (and also the OP) makes focuschecks on tilted surfaces/targets. The AF target is in the middle and the ruler is far autside the AF box that is allowed only to contain perpenticular structures....

We have a different problem (?) here ...

 Franz Kerschbaum's gear list:Franz Kerschbaum's gear list
Canon EOS 5D Mark II Canon EOS 7D Mark II Canon EOS M6 Canon EOS Ra Canon EOS R5 +23 more
R5D4
OP R5D4 Contributing Member • Posts: 838
Re: My RF 70-200 f2.8 is front focusing
10

lawny13 wrote:

Sorry, but you are mistakened.

Just because you have a MILC system doesn't mean that you won't have focus issues. I ran into this problem on sony as well. Generally you will find it in every MILC.
Basically mirrorless removes the need for micro focus calibration. This is only with respect to the correction of the image plane of the focus sensor and that of the sensor. However, that is only part of the factor.
I will use the the eye image as an example. The camera detects a face, then from there an eye. It places an AF box on the eye. As yourself what is in that box. The eye itself, as in the pupil, eye lid, and eye lashes. So that is what it is focusing on. So like when a camera is focusing on full auto it tends to grab onto the closest thing in the fame with good enough contrast. Well the same is done here with what is in the frame that is closest are the eye lashes.
For your ruler example. You place the box on the 0 for example. But it seems to front focus. Well if I were to be looking at your camera, and the AF box, it would likely be focusing on the button of that AF box with corresponds to the closest elements to you in that box.
So... it is working perfectly, though I am sure your expectation is that it prioritizes the center most portion of the AF box? I agree that this should be the case actually, but it is obviously not what is done.
I have had experiences with sony MILC setups where the camera would constantly grab onto the background of the subject I was trying to capture, because the background also happened to be partially in the AF box. It was quite annoying, and I had to revert to manual focus.

The thing is that a DSLR like the 5DIV and the 7DII have the option of that extra small AF point. That was better for precise focusing (minus the focus sensor mirco adjustment thing). Though MILC removes the mirco adjustment problem it introduces other things. Canon's AF box contains a whole like of pixel size AF points It is unclear which are uses, and how they are prioritized though closer to camera seems to be the case based on what you show. It would be nice if the box could be made even smaller for more precise things. Or in the case of eye AF that the pupil itself is also prioritized, or that it does a center of AF weighted AF.
Sorry for the rant, but I am fairly certain that there is nothing to "repair" as far as your lens is concerned. This AF issue has thing to do with the hardware (lens or camera) and more to do with software. Canon has some room for improvement in that department.

Nope. Any of my other lenses focus on the pupil. Including the EF 85 1.4, where the depth of field is extremely shallow. I have over a dozen images where the RF sees the eye (shows the tiny box on the eye during focusing) but every image is front focused.

You misunderstand how focus targets work. The part that looks like a QR code is vertical, and is more than big enough to contain the focus box. The AF box is set to small in the settings. The ruler is at a 45 degree angle to that, but the camera isn’t using the ruler to grab focus.

 R5D4's gear list:R5D4's gear list
Canon EOS R5 Fujifilm X-T3 Canon EOS R Canon EF 35mm F1.4L II USM Fujifilm XF 35mm F2 R WR +5 more
R5D4
OP R5D4 Contributing Member • Posts: 838
Re: My RF 70-200 f2.8 is front focusing

shawnphoto wrote:

pokesfan wrote:

Sounds impossible, right?

I was enjoying my new RF 70-200, acquired over thanksgiving break on a vacation (and away from my test target and tripod) when I saw this thread over at FM Forums:

https://www.fredmiranda.com/forum/topic/1623630

Nonsense, I thought. My lens is awesome. The pictures it takes are awesome. It's impossible for a lens to front or back focus on a mirrorless camera anyhow, right?

Anways, the thread gnawed on me, and when I finally got home I got out the target and the tripod.

Son of a bi____. My lens does it too. At 200mm, near MFD, f2.8, it focuses about 3mm in front of the confirmed focus point:

200mm @ 2.8, near MFD

At 163mm, near MFD, 2.8. Looks to be about 2.5 to 3mm in front of the target:

163mm, @ 2.8 near MFD

At 135mm, the distance is shrinking. Probably 2mm

135mm @ 2.8 near MFD

At 100mm it looks more or less cleared up, but maybe still 1mm? I took a 70mm shot also in my first test and it seemed perfect.

100mm @ 2.8 near MFD

Each shot was 100% repeatable. I took about 10 shots for each of these, with the camera on a tripod, 2 second timer and the focus mode in single point. Every single shot was identical to the one before it. Also I zoomed in to 10x, focused and then was able to get an even sharper image by turning the focus ring.

I went back and looked at all the photos I had taken. There was some slight pop missing from a lot of portraits. Other photos I couldn't tell. And then I found this series of 5 images taken with Eye AF, all with the same miss by about 2-3mm in front of the pupil of the eye:

I am contacting Canon tomorrow and will send the lens in to be repaired. If it comes back not repaired, I plan to get a refund from the seller.

If you read the thread over on FM Forums, mine is something like the 8th lens with this same problem. There is a shop owner in that thread who tested 3 copies on both the R and RP. I really don't want to mess around exchanging this lens for another copy if the problem is universal.

I'm posting this as a PSA for anybody with this lens- you might want to take a closer look, STAT. I'll update here on what happens with Canon.

Are we even sure Canon eye AF focuses on the pupil?

Are you using the latest firmware?

I haven’t been able to reproduce the error even shooting test charts.

I am using the R w/ 1.4 and the lens is 1.0.3

Thats really good news if you can’t reproduce this. Over at FM anybody who has run the test has found the error. What is your lens firmware version?

 R5D4's gear list:R5D4's gear list
Canon EOS R5 Fujifilm X-T3 Canon EOS R Canon EF 35mm F1.4L II USM Fujifilm XF 35mm F2 R WR +5 more
AkshajN Regular Member • Posts: 360
Re: My RF 70-200 f2.8 is front focusing

pokesfan wrote:

shawnphoto wrote:

pokesfan wrote:

Sounds impossible, right?

I was enjoying my new RF 70-200, acquired over thanksgiving break on a vacation (and away from my test target and tripod) when I saw this thread over at FM Forums:

https://www.fredmiranda.com/forum/topic/1623630

Nonsense, I thought. My lens is awesome. The pictures it takes are awesome. It's impossible for a lens to front or back focus on a mirrorless camera anyhow, right?

Anways, the thread gnawed on me, and when I finally got home I got out the target and the tripod.

Son of a bi____. My lens does it too. At 200mm, near MFD, f2.8, it focuses about 3mm in front of the confirmed focus point:

200mm @ 2.8, near MFD

At 163mm, near MFD, 2.8. Looks to be about 2.5 to 3mm in front of the target:

163mm, @ 2.8 near MFD

At 135mm, the distance is shrinking. Probably 2mm

135mm @ 2.8 near MFD

At 100mm it looks more or less cleared up, but maybe still 1mm? I took a 70mm shot also in my first test and it seemed perfect.

100mm @ 2.8 near MFD

Each shot was 100% repeatable. I took about 10 shots for each of these, with the camera on a tripod, 2 second timer and the focus mode in single point. Every single shot was identical to the one before it. Also I zoomed in to 10x, focused and then was able to get an even sharper image by turning the focus ring.

I went back and looked at all the photos I had taken. There was some slight pop missing from a lot of portraits. Other photos I couldn't tell. And then I found this series of 5 images taken with Eye AF, all with the same miss by about 2-3mm in front of the pupil of the eye:

I am contacting Canon tomorrow and will send the lens in to be repaired. If it comes back not repaired, I plan to get a refund from the seller.

If you read the thread over on FM Forums, mine is something like the 8th lens with this same problem. There is a shop owner in that thread who tested 3 copies on both the R and RP. I really don't want to mess around exchanging this lens for another copy if the problem is universal.

I'm posting this as a PSA for anybody with this lens- you might want to take a closer look, STAT. I'll update here on what happens with Canon.

Are we even sure Canon eye AF focuses on the pupil?

Are you using the latest firmware?

I haven’t been able to reproduce the error even shooting test charts.

I am using the R w/ 1.4 and the lens is 1.0.3

Thats really good news if you can’t reproduce this. Over at FM anybody who has run the test has found the error. What is your lens firmware version?

Just to rule out the possibility of EOS R focusing system problem - Did anyone test this with RP?

 AkshajN's gear list:AkshajN's gear list
Panasonic Lumix DC-S1R Canon EOS R5 Sigma 14-24mm F2.8 DG HSM Art Sigma 105mm F1.4 DG HSM Art Panasonic S Pro 50mm F1.4 +7 more
R5D4
OP R5D4 Contributing Member • Posts: 838
Re: My RF 70-200 f2.8 is front focusing
1

AkshajN wrote:

pokesfan wrote:

shawnphoto wrote:

pokesfan wrote:

Sounds impossible, right?

I was enjoying my new RF 70-200, acquired over thanksgiving break on a vacation (and away from my test target and tripod) when I saw this thread over at FM Forums:

https://www.fredmiranda.com/forum/topic/1623630

Nonsense, I thought. My lens is awesome. The pictures it takes are awesome. It's impossible for a lens to front or back focus on a mirrorless camera anyhow, right?

Anways, the thread gnawed on me, and when I finally got home I got out the target and the tripod.

Son of a bi____. My lens does it too. At 200mm, near MFD, f2.8, it focuses about 3mm in front of the confirmed focus point:

200mm @ 2.8, near MFD

At 163mm, near MFD, 2.8. Looks to be about 2.5 to 3mm in front of the target:

163mm, @ 2.8 near MFD

At 135mm, the distance is shrinking. Probably 2mm

135mm @ 2.8 near MFD

At 100mm it looks more or less cleared up, but maybe still 1mm? I took a 70mm shot also in my first test and it seemed perfect.

100mm @ 2.8 near MFD

Each shot was 100% repeatable. I took about 10 shots for each of these, with the camera on a tripod, 2 second timer and the focus mode in single point. Every single shot was identical to the one before it. Also I zoomed in to 10x, focused and then was able to get an even sharper image by turning the focus ring.

I went back and looked at all the photos I had taken. There was some slight pop missing from a lot of portraits. Other photos I couldn't tell. And then I found this series of 5 images taken with Eye AF, all with the same miss by about 2-3mm in front of the pupil of the eye:

I am contacting Canon tomorrow and will send the lens in to be repaired. If it comes back not repaired, I plan to get a refund from the seller.

If you read the thread over on FM Forums, mine is something like the 8th lens with this same problem. There is a shop owner in that thread who tested 3 copies on both the R and RP. I really don't want to mess around exchanging this lens for another copy if the problem is universal.

I'm posting this as a PSA for anybody with this lens- you might want to take a closer look, STAT. I'll update here on what happens with Canon.

Are we even sure Canon eye AF focuses on the pupil?

Are you using the latest firmware?

I haven’t been able to reproduce the error even shooting test charts.

I am using the R w/ 1.4 and the lens is 1.0.3

Thats really good news if you can’t reproduce this. Over at FM anybody who has run the test has found the error. What is your lens firmware version?

Just to rule out the possibility of EOS R focusing system problem - Did anyone test this with RP?

The guy over on the FM thread who works at a camera store says he tested the lens on the RP with the same results.

 R5D4's gear list:R5D4's gear list
Canon EOS R5 Fujifilm X-T3 Canon EOS R Canon EF 35mm F1.4L II USM Fujifilm XF 35mm F2 R WR +5 more
R5D4
OP R5D4 Contributing Member • Posts: 838
Some EF lenses to compare
2

I ran the test with the EF 70-200 f2.8 IS II and the EF 24-70 f2.8 II. They both seem fine.

EF 70-200 @ 200mm f2.8

EF 24-70 @ 70mm f2.8

 R5D4's gear list:R5D4's gear list
Canon EOS R5 Fujifilm X-T3 Canon EOS R Canon EF 35mm F1.4L II USM Fujifilm XF 35mm F2 R WR +5 more
lawny13 Veteran Member • Posts: 3,029
Re: My RF 70-200 f2.8 is front focusing

pokesfan wrote:

lawny13 wrote:

Sorry, but you are mistakened.

Just because you have a MILC system doesn't mean that you won't have focus issues. I ran into this problem on sony as well. Generally you will find it in every MILC.
Basically mirrorless removes the need for micro focus calibration. This is only with respect to the correction of the image plane of the focus sensor and that of the sensor. However, that is only part of the factor.
I will use the the eye image as an example. The camera detects a face, then from there an eye. It places an AF box on the eye. As yourself what is in that box. The eye itself, as in the pupil, eye lid, and eye lashes. So that is what it is focusing on. So like when a camera is focusing on full auto it tends to grab onto the closest thing in the fame with good enough contrast. Well the same is done here with what is in the frame that is closest are the eye lashes.
For your ruler example. You place the box on the 0 for example. But it seems to front focus. Well if I were to be looking at your camera, and the AF box, it would likely be focusing on the button of that AF box with corresponds to the closest elements to you in that box.
So... it is working perfectly, though I am sure your expectation is that it prioritizes the center most portion of the AF box? I agree that this should be the case actually, but it is obviously not what is done.
I have had experiences with sony MILC setups where the camera would constantly grab onto the background of the subject I was trying to capture, because the background also happened to be partially in the AF box. It was quite annoying, and I had to revert to manual focus.

The thing is that a DSLR like the 5DIV and the 7DII have the option of that extra small AF point. That was better for precise focusing (minus the focus sensor mirco adjustment thing). Though MILC removes the mirco adjustment problem it introduces other things. Canon's AF box contains a whole like of pixel size AF points It is unclear which are uses, and how they are prioritized though closer to camera seems to be the case based on what you show. It would be nice if the box could be made even smaller for more precise things. Or in the case of eye AF that the pupil itself is also prioritized, or that it does a center of AF weighted AF.
Sorry for the rant, but I am fairly certain that there is nothing to "repair" as far as your lens is concerned. This AF issue has thing to do with the hardware (lens or camera) and more to do with software. Canon has some room for improvement in that department.

Nope. Any of my other lenses focus on the pupil. Including the EF 85 1.4, where the depth of field is extremely shallow. I have over a dozen images where the RF sees the eye (shows the tiny box on the eye during focusing) but every image is front focused.

You misunderstand how focus targets work. The part that looks like a QR code is vertical, and is more than big enough to contain the focus box. The AF box is set to small in the settings. The ruler is at a 45 degree angle to that, but the camera isn’t using the ruler to grab focus.

Ah... ok. My bad and thanks for the info

Perhaps you should look up in the menu if there is a micro adjust option. I know that the sony bodies actually have it. The reasoning to this is that a lens does need to be calibrated. I assume maybe there is a calibration issue with it. Hence they manufacturers still include a micro adjust option.
Does it also do this in continuous focus?

ZX11
ZX11 Veteran Member • Posts: 5,428
Construction design
1

pokesfan wrote:

I ran the test with the EF 70-200 f2.8 IS II and the EF 24-70 f2.8 II. They both seem fine.

Interesting.

The center is the focus point and the edge of the glass/frame is the area used to measure focus error. At that close range, could the edge be slightly closer than the center due to the construction? Does the focus tighten up on the target if you manually focus from the AF focus point or is the focus as good as it can get after the AF does it?

Imagining the difficulties of an on sensor focus system deciding to go slightly off focus instead of on focus, makes me think the edge of the lens is closer focus than the center due to the construction.

So,... the focus plane isn't dead flat but follows the curve of the front element when focused that close. Which would let the focus point be dead on while the ruler, to the side of the focus point, shows slight focus front.

Since the construction of the lens is so different from the EF version I wonder if it has different strengths and weaknesses. A front element that moves such a distance is weird to me. Though it is cool that there are less correction elements in the design. The designers allowed a curved focus plane?

-- hide signature --

"Very funny, Scotty! Now beam me down my clothes."
"He's dead, Jim! You grab his tri-corder. I'll get his wallet."

 ZX11's gear list:ZX11's gear list
Canon EF 50mm F1.4 USM Canon EF 85mm F1.8 USM Canon RF 24-105mm F4L IS USM
Photomonkey Senior Member • Posts: 2,794
Re: My RF 70-200 f2.8 is front focusing
3

lawny13 wrote:

pokesfan wrote:

lawny13 wrote:

Perhaps you should look up in the menu if there is a micro adjust option. I know that the sony bodies actually have it. The reasoning to this is that a lens does need to be calibrated. I assume maybe there is a calibration issue with it. Hence they manufacturers still include a micro adjust option.
Does it also do this in continuous focus?

Sony has focus adjust because of their adapters for A mount lenses that have mirrors.

Mirrorless does not need focus adjust as the focus is calculated off the actual sensor.

No one else has focus adjust for mirrorless because the concept makes it meaningless.

I noted in another thread that Canon reps claimed that focus on long lenses were less reliable with mirrorless. It seemed like smoke to cover their absence in the high end MILC segment.

I use my 70-200 2.8L IS with an adapter on my R and have better results than with my mk4. So clearly there is no real hint of an answer there.

I think DPAF is a culprit here as it calculates the focus and drives the lens to the correct position without a check to confirm (except, possibly in AF servo mode). CDAF on the other hand is faulted for its iterative process and relative slowness. No one challenges its accuracy though.

IMO it is a run of miscalibrated lenses that escaped the factory.

 Photomonkey's gear list:Photomonkey's gear list
Kodak Pixpro S-1
NoTicket Regular Member • Posts: 209
Re: Some EF lenses to compare

pokesfan wrote:

I ran the test with the EF 70-200 f2.8 IS II and the EF 24-70 f2.8 II. They both seem fine.

EF 70-200 @ 200mm f2.8

EF 24-70 @ 70mm f2.8

Thank you for taking the time to do this. It is encouraging that this does not happe on the EF versions.

To me this indicates that there is some inbuilt focusing offsets programmed in to the RF 70-200 firmware that can probably be adjusted by Canon.

The reproducibility of practically the exact same offset by everybody that has posted their results leads me to believe this will be resolved soon.

NoTicket Regular Member • Posts: 209
Re: Construction design

ZX11 wrote:

pokesfan wrote:

I ran the test with the EF 70-200 f2.8 IS II and the EF 24-70 f2.8 II. They both seem fine.

Interesting.

The center is the focus point and the edge of the glass/frame is the area used to measure focus error. At that close range, could the edge be slightly closer than the center due to the construction? Does the focus tighten up on the target if you manually focus from the AF focus point or is the focus as good as it can get after the AF does it?

Imagining the difficulties of an on sensor focus system deciding to go slightly off focus instead of on focus, makes me think the edge of the lens is closer focus than the center due to the construction.

So,... the focus plane isn't dead flat but follows the curve of the front element when focused that close. Which would let the focus point be dead on while the ruler, to the side of the focus point, shows slight focus front.

Since the construction of the lens is so different from the EF version I wonder if it has different strengths and weaknesses. A front element that moves such a distance is weird to me. Though it is cool that there are less correction elements in the design. The designers allowed a curved focus plane?

In the fredmiranda thread people also compared the results with manual focus on the focus target and you can see the point of focus becomes much sharper as well.

I don’t believe this has anything to do with the focal plane being skewed or meaningfully non flat.

J A C S
J A C S Forum Pro • Posts: 17,823
Re: My RF 70-200 f2.8 is front focusing

lawny13 wrote:

For your ruler example. You place the box on the 0 for example. But it seems to front focus. Well if I were to be looking at your camera, and the AF box, it would likely be focusing on the button of that AF box with corresponds to the closest elements to you in that box.

He places the AF box in a flat target perpendicular to the axis. This is the whole point of this type of targets.

nordilux
nordilux Regular Member • Posts: 148
The issue seems real...

...let's hope it is something Canon can address with a firmware update (vs taking back and manually re-calibrate the lenses somehow). This shouldn't happen on a pro-level lens in this price range.

herion
herion Contributing Member • Posts: 753
Re: Some EF lenses to compare

"To me this indicates that there is some inbuilt focusing offsets programmed in to the RF 70-200 firmware that can probably be adjusted by Canon."

Sounds like Canon may be coming out with a RF dock a la Tamron's Tap-In...

 herion's gear list:herion's gear list
Canon EOS 30D Canon EOS 40D Canon EF-S 10-18mm F4.5–5.6 IS STM Tamron SP 45mm F1.8 Di VC USD Tamron SP 85mm F1.8 Di VC USD +2 more
Keyboard shortcuts:
FForum MMy threads