Fuji Autofocusing issues?

[Irene McC]

Yes.

 

[Bobby T]

Come on, folks!

This is becoming like the "discussions" where some E-M1 MkI owners dismissed the documented problems with that body. Some were fortunate enough not to have any problems. I had the strap lug rear wheel problem. My brother had two different bodies with the strap lug problem, and the third eventually developed the bad rear wheel problem.

However, some people had neither problem, and some went into vociferous denial mode about both (DPReview ... ).

No need for any of that behaviour. Please.

I’m not dismissing it. I stated that I could not recreate it. That I have not noticed it any point in my time shooting Fuji. And that if there are steps required to follow, it’s not a problem with the gear. If someone takes that personally instead of scientifically as data points then they need to re think their approach to troubleshooting. There are going to be people who don’t have this problem and can’t recreate it.

And John is correct. The behavior which typically occurs on other sites won’t be acceptable here.

 

[sturgeon]

My own experience with X-T5 ...

I don't use the focus scale on the screen. For me it's just screen clutter.
I have an old but solid Gitzo tripod.
For shots where I think focus is critical, I use the tripod and a remote release to trigger the shutter.

Since it's raining here today and I'm bored, I decided to try some tests, despite not having had any auto-focus issues with this camera, other than ones of my own doing. I used my widest lens, a Fuji 16/2.8. I took numerous shots at F/2.8, F/8, and F/16. All at ISO 800. Using both AF-S and AF-C. Camera set to 425 focus points, single-point AF mode, with a 3x3 frame size (box through the points, not outside them). Mechanical shutter, aperture priority.

Not a single image was out of focus at the point on which I focused the camera.

For my purposes, the X-T5 works just fine. I'm well aware of my physical limitations (I'm a geezer), and I know that if I shoot handheld, no amount of IS will enable me to consistently get the sharper shots I can get with a tripod and a remote release, particularly with a higher megapixel camera like the X-T5. If I shoot handheld, I know I can see sharpness issues. Sometimes a bit of fuzz doesn't matter to me, but often it does.

 

[Darmok N Jalad]

Sometimes I wondered if it was me and the lack of IBIS that caused a mis-focus on the X-T3, but it certainly did seem like it was a little behind my intended focal point, even while using single point. I had noticed the distance meter would be a bit erratic, which is sad since it could stand to be a useful feature. I can’t say for certain if I experienced this issue, as I’m not a tripod shooter.

 

[AndyH44]

2 shots taken a while ago using AF-S, single point small box, exact same target.

Both SOOC. One showed about 30 feet on distance scale; the other showed just 6 FEET.

Not hard to see which is which, especially at 100%. Some might feel the OOF image is acceptable, but I certainly don't.

If you don't have the distance scale activated, you would never catch this issue while shooting!

105F in our front yard BTW

Join to see EXIF info for this image (if available)

Join to see EXIF info for this image (if available)

Cropped about 100%

Join to see EXIF info for this image (if available)

Join to see EXIF info for this image (if available)

 

[doobs]

All,

Sorry to post and then disappear.

We're in the homestretch of a new home build and the cats are getting harder to herd...

One of the things I've been doing with the house is to take a full set of pictures, 4 in each room in the corner every time we visit.

We live in Northern Virginia and the house is in Williamsburg. Typically we do a run down once a week.

We went yesterday, and this was the first time that I have shot with the distance scale turned on. For reference this is my X-T2 with the XF10-24, typically at 10mm.

I was gobsmacked how seemingly random the focusing distances were at apertures smaller than f4.0. All over the map. 3 feet, infinity, 20 feet, etc, etc, etc.

So, I switched to manual focus and used the depth of field scale to ensure what I wanted to be in focus actually was.

Obviously I need to do a lot of testing to see where the limits are, and then act accordingly.

Up until this instance, I was a dedicated, dyed in the wool Fujista. Unfortunately, with this development, the rot has set in a bit.

Thanks for the good dialog. Let's keep the discourse going.

Let's all cross our fingers that Fuji will 1) sort this out, and 2) make the solution available across the board.

cheers,

 

[Piano]

I had random problems with my X-H2S and contacted Fujifilm UK. I sent them some raw images and they agreed that the images were soft and noisy and asked me to send it in. Received it back with a new circuit board (sensor) and front panel so I need to get out and test to see if there's any improvement. Fingers crossed.

 

[AndyH44]

I had random problems with my X-H2S and contacted Fujifilm UK. I sent them some raw images and they agreed that the images were soft and noisy and asked me to send it in. Received it back with a new circuit board (sensor) and front panel so I need to get out and test to see if there's any improvement. Fingers crossed.

Please keep us informed!!!

 

[doobs]

H'mmm.

I did some testing this afternoon.

10 exposures each for:

1. Wide open
2. Mid range
3. Completed stopped down.

For zoom lenses:

1. Widest zoom
2. Mid range
3. Farthest zoom

The focus distance indicator was all over the map for the X-T2, X-Pro2 and X100f.

Yet, bringing them into Capture One, the focus mask indicator is in exactly the same position, at least for the first set I've processed for the X-T2.

Here is an example:

Every shot I look at has the splotch on the bucket, which was my focus point.

I've got the X-Pro2 files importing now, and will do the X100f files tomorrow.

Stay tuned

 

[Darmok N Jalad]

H'mmm.

I did some testing this afternoon.

10 exposures each for:

1. Wide open
2. Mid range
3. Completed stopped down.

For zoom lenses:

1. Widest zoom
2. Mid range
3. Farthest zoom

The focus distance indicator was all over the map for the X-T2, X-Pro2 and X100f.

Yet, bringing them into Capture One, the focus mask indicator is in exactly the same position, at least for the first set I've processed for the X-T2.

Here is an example:

View attachment 401255

Every shot I look at has the splotch on the bucket, which was my focus point.

I've got the X-Pro2 files importing now, and will do the X100f files tomorrow.

Stay tuned

I guess if nothing else, it kinda suggests the focus distance indicator doesn’t work. I know some cameras will calculate the distance and put it in the EXIF, but not all manufacturers do it.

 

[AndyH44]

I guess if nothing else, it kinda suggests the focus distance indicator doesn’t work. I know some cameras will calculate the distance and put it in the EXIF, but not all manufacturers do it.

I'm sorry but that image looks unacceptably OOF to me. My guess is that camera actually focused well beyond the subject distance.
The distance indicator is very accurate on my X-S20, and was very accurate on my now sold X-E4, X-H2, and X-T5. When it indicates let's say 10ft at a subject that is 10ft away your image WILL be in critical focus. Tried this many, many times without fail.

 

[AndyH44]

I guess if nothing else, it kinda suggests the focus distance indicator doesn’t work. I know some cameras will calculate the distance and put it in the EXIF, but not all manufacturers do it.

Easy enough to test accuracy. Enable distance indicator (set DOF at "pixel level"), select MF then dial focus ring until the white line is on the distance your subject is actually at. Turn off shutter AF and take a pic and it will be in focus unless you're a Klutz.
BTW- the distance indicator is just as accurate in AF-S and AF-C modes (may "jump around" in AF-C). Been there, done that literally hundreds of times with all the Fuji cameras I've recently owned.

 

[Piano]

I finally had the chance to get out with the X-H2S with new sensor and front panel. A ladies football match. I have to say that I'm scratching my head with this. I have a number of image sequences where the first image is in focus and the subsequent ones aren't. Here's an example. The second image hasn't kept track of the subject at all despite being shot in a single burst of 5 images with good follow on my part.

Join to see EXIF info for this image (if available)

Join to see EXIF info for this image (if available)

 

[Bobby T]

I finally had the chance to get out with the X-H2S with new sensor and front panel. A ladies football match. I have to say that I'm scratching my head with this. I have a number of image sequences where the first image is in focus and the subsequent ones aren't. Here's an example. The second image hasn't kept track of the subject at all despite being shot in a single burst of 5 images with good follow on my part.

View attachment 401443View attachment 401444

For reference, can you tell us what your autofocus settings were?

 

[Piano]

Electronic Shutter - 15fps
AF-C
Single Point Focus
Custom settings 5 or 1
EDIT: I should add that it didn't seem to make any difference whether I used Zone or single point but it "hopped around" a lot less on single point.

 

[Bobby T]

If you get to shoot this again soon. Try it with af-c setting 2. Then maybe setting 3 unless setting 2 works flawlessly.

 

[Piano]

I'll try anything at the moment Bobby thanks. However I still can't understand how consecutive frames can be so different from each other.

 

[Bobby T]

I'll try anything at the moment Bobby thanks. However I still can't understand how consecutive frames can be so different from each other.

It’s both experimenting to see which, if any af-c works best. As well as a process of elimination. I suggested setting 2 partly because of the background in this shot. People in dark clothes with bright white cars. That heavy contrast is easy for the autofocus to lock onto. It’s not supposed to do that, but it happens.

 

[Piano]

I tried "2" last night, together with 3 different lenses, 50-140, 55-200 and 80mm prime.
50-140 gave same results as previous. 55-200 gave slightly softer better images but less variation in oof sequences. 80mm showed no discernible differences. Still no wiser . I asked my friend to take some images on his Sony A1 on the same memory card for comparison. Much better sharper images but his worse shots were comparable to mine. I have another attempt next Monday where I'll try something different. It's getting like a serial isn't it? Stay tuned!

 

[AndyH44]

I tried "2" last night, together with 3 different lenses, 50-140, 55-200 and 80mm prime.
50-140 gave same results as previous. 55-200 gave slightly softer better images but less variation in oof sequences. 80mm showed no discernible differences. Still no wiser . I asked my friend to take some images on his Sony A1 on the same memory card for comparison. Much better sharper images but his worse shots were comparable to mine. I have another attempt next Monday where I'll try something different. It's getting like a serial isn't it? Stay tuned!

Please try AF-S with a wide angle lens (<27mm) @ f5.6 using the distance scale with DOF set at "pixel" for reference. Choose a subject at a known distance, say 15ft. See if the distance marker (white line) chooses correct subject distance consistently with repeated presses of AF-ON button or shutter (whichever you use).