Firmware 4.0 with 18-55 focus accuracy is worse than before

[s_illes79]

This kind of mis-focus seems to be happening quite often - more than with 3.1 FW

Settings: Single point focus somewhere in the middle with biggest box. Good light conditions, bad results. Maybe camera got confused by the green foliage this time, noticed similar bad focus elsewhere...

Has anyone experienced similar ?

 

[forpetessake]

Have you ruled out camera shake or OIS? Did the OIS revert to Mode 1 after the FW update?

This kind of mis-focus seems to be happening quite often - more than with 3.1 FW

Settings: Single point focus somewhere in the middle with biggest box. Good light conditions, bad results. Maybe camera got confused by the green foliage this time, noticed similar bad focus elsewhere...

Has anyone experienced similar ?

 

[LightDriver]

Not sure this is in any way related but I have noticed significantly slower focusing, some lens hunting and also pictures out of focus with FW 4.0 and my 50-140.

This kind of mis-focus seems to be happening quite often - more than with 3.1 FW

Settings: Single point focus somewhere in the middle with biggest box. Good light conditions, bad results. Maybe camera got confused by the green foliage this time, noticed similar bad focus elsewhere...

Has anyone experienced similar ?

 

[Acrill]

Sorry, but I have only noticed improvement since 4.0

Your anecdotal test doesn't really mean anything here. Its not like you shot the exact same scene in the exact same light with two different firmware versions.

A few years of shooting contrast-detection cameras has taught me that they will always be a little bit prone to this type of 'mis-focusing'. The camera is not a telepathic being, it cannot read your mind to determine what should be in focus. The best it can do is focus on the highest contrast area within the focus box.

The only remedy is to use your eyes - watch what it is focusing on the viewfinder and if it mis-focuses, re-focus it on what you want.

 

[nixda]

Have you ruled out camera shake or OIS? Did the OIS revert to Mode 1 after the FW update?

Indeed, OIS is mode 1 (continuous).

 

[s_illes79]

Sorry, but I have only noticed improvement since 4.0

Your anecdotal test doesn't really mean anything here. Its not like you shot the exact same scene in the exact same light with two different firmware versions.

A few years of shooting contrast-detection cameras has taught me that they will always be a little bit prone to this type of 'mis-focusing'. The camera is not a telepathic being, it cannot read your mind to determine what should be in focus. The best it can do is focus on the highest contrast area within the focus box.

The only remedy is to use your eyes - watch what it is focusing on the viewfinder and if it mis-focuses, re-focus it on what you want.

who said it was a test ? just an observation based on hundreds of samples, selected one and uploaded.

let me simplify it for you

- same person -> same technique

- same kind of photography -> static landscape

- same camera

- NEW firmware

=> more out of focus images.

one thing i noticed that eye detection was turned on . I'll turn it off and see what happens.

 

[57LowRider]

Tripod with OIS left on? Just a guess.

 

[s_illes79]

Tripod with OIS left on? Just a guess.

no tripod - but face + eye detection got activated after FW update ... might be the culprit.

 

[s_illes79]

Have you ruled out camera shake or OIS? Did the OIS revert to Mode 1 after the FW update?

Indeed, OIS is mode 1 (continuous).

i think it was always set to Mode 1 ... why would that be a problem?

 

[LightDriver]

Interestingly enough, switching OIS back to "shooting only" seems to have helped somewhat with the hunting lens issue with my 50-140.

Tripod with OIS left on? Just a guess.

no tripod - but face + eye detection got activated after FW update ... might be the culprit.

 

[nixda]

What is your setting for "AF-C/AF-S Priority Selection"?

 

[s_illes79]

What is your setting for "AF-C/AF-S Priority Selection"?

both set to focus

 

[nixda]

What is your setting for "AF-C/AF-S Priority Selection"?

both set to focus

Do you always leave OIS on? How is your success rate with OIS off when you use fast shutter speeds?

 

[AlbertTheLazy]

This kind of mis-focus seems to be happening quite often - more than with 3.1 FW

Settings: Single point focus somewhere in the middle with biggest box. Good light conditions, bad results. Maybe camera got confused by the green foliage this time, noticed similar bad focus elsewhere...

Has anyone experienced similar ?

I've just looked at this in 100% view and can't see anything sharp anywhere. 'Somewhere in the middle' would indicate that you focused around about that bird (?) just below the hedge/tree line, but that area doesn't look any sharper than the rest of the field.

Whatever it is, it doesn't look like missed focus to me.

--
Albert
Every photograph is an abstraction from reality.
Most people are more interested in the picture than the image.

 

[AdHoc007]

Was a lens filter used?

 

[oscarvdvelde]

I've had very similar focus problems with both a XF 27mm and XF 18-135mm. The green box would appear, but focus is most likely beyond infinity. In case of the 27mm even camera on/off did not solve the problem, maybe remounting lens or battery did. This happened with FW 3 and I have not seen it yet since FW 4.

That said, on my last shooting my X-T1 FW 4.0 with XF 27mm took pretty long each time (3-4 seconds) to provide an image in the viewfinder upon turning on the camera (in high performance mode and no dust shake at start, fresh batteries) which I thought was unusual. Maybe I should check the lens contacts.

 

[s_illes79]

This kind of mis-focus seems to be happening quite often - more than with 3.1 FW

Settings: Single point focus somewhere in the middle with biggest box. Good light conditions, bad results. Maybe camera got confused by the green foliage this time, noticed similar bad focus elsewhere...

Has anyone experienced similar ?

I've just looked at this in 100% view and can't see anything sharp anywhere. 'Somewhere in the middle' would indicate that you focused around about that bird (?) just below the hedge/tree line, but that area doesn't look any sharper than the rest of the field.

Whatever it is, it doesn't look like missed focus to me.

--
Albert
Every photograph is an abstraction from reality.
Most people are more interested in the picture than the image.

as pointed out by someone below focus beyond infinity produces images like that

 

[nixda]

This kind of mis-focus seems to be happening quite often - more than with 3.1 FW

Settings: Single point focus somewhere in the middle with biggest box. Good light conditions, bad results. Maybe camera got confused by the green foliage this time, noticed similar bad focus elsewhere...

Has anyone experienced similar ?

I've just looked at this in 100% view and can't see anything sharp anywhere. 'Somewhere in the middle' would indicate that you focused around about that bird (?) just below the hedge/tree line, but that area doesn't look any sharper than the rest of the field.

Whatever it is, it doesn't look like missed focus to me.

--
Albert
Every photograph is an abstraction from reality.
Most people are more interested in the picture than the image.

as pointed out by someone below focus beyond infinity produces images like that

At f/7.1 and a FL of 18mm, the hyperfocal distance is 2.3m. That means that any focus point that is beyond a couple of meters should render everything sharp to infinity. For example, when the camera is focused at 10,000m or 100,000,000m, the DoF extends from 2.3m to infinity.

IMO, there is something else at the bottom of the behavior you are seeing.

The picture isn't sharp anywhere, which to me indicates that there is motion blur. It is quite possible that the OIS cannot cope with your personal type of tremor and that at the shutter speed you used (1/455s) some sort of image shake results.

You are saying that you are observing this type of behavior a lot, so it shouldn't be too difficult to reproduce it. At that time, also put the camera on some solid support, and take some shots both with the OIS on an off. The more shots the better, to even out statistical fluctuations.

 

[oscarvdvelde]

as pointed out by someone below focus beyond infinity produces images like that

At f/7.1 and a FL of 18mm, the hyperfocal distance is 2.3m. That means that any focus point that is beyond a couple of meters should render everything sharp to infinity. For example, when the camera is focused at 10,000m or 100,000,000m, the DoF extends from 2.3m to infinity.

Beyond infinity does not mean approaching infinity in distance (that's before infinity!). It means the lens elements focus light from any object in the image, near or far, behind the sensor, giving a blur circle at the sensor.

--
Weather & Photography
http://www.lightningwizard.com
http://www.flickr.com/photos/lightningwizard

 

[nixda]

as pointed out by someone below focus beyond infinity produces images like that

At f/7.1 and a FL of 18mm, the hyperfocal distance is 2.3m. That means that any focus point that is beyond a couple of meters should render everything sharp to infinity. For example, when the camera is focused at 10,000m or 100,000,000m, the DoF extends from 2.3m to infinity.

Beyond infinity does not mean approaching infinity in distance (that's before infinity!). It means the lens elements focus light from any object in the image, near or far, in front of the sensor, giving a blur circle at the sensor.

Alright then, the image plane doesn't line up with the sensor plane. I think that can happen with autofocus mechanisms when they are moved to the extremes. For the AF mechanism to hone in on the focus point, it must usually reset itself by moving to an extreme. Perhaps, it never got away from that extreme when these images were taken. That means that the mechanism that moves the lens elements may have gotten whacked somehow.