Fuji The DOF Scale Is Wrong

So if you set the focus to infinity and then turn to a very distant object, kilometres way or many thousands of kilometres, e.g. the moon, is it in focus or not?

If not, then and only then can you claim infinty is not infinity.

Likewise, if you place an object at a measured 3 metres and set the focus ring to 3 metres, is it in focus or not?

When working with extremes of depth of field (for example relying on an assumed hyperfocal setting), you need to take into consideration what you intend to do with the images. As the Cambridge calculator shows, there is more to take into consideration than just the focal length and aperture of the lens.

With most lenses these days not having depth of field markings (or so few that they are useless) there is no substitute for experience and applying that same experience in practice.
 
I suppose the good thing is that Fuji erred on the "good" side then? Good side meaning that the DOF is actually wider (or is it deeper) than what Fuji is representing on the camera screen.

True, BUT, the nice thing about the automatic DOF scale for hyperfocal shooting is that if you trust the scale, you don't have to calculate or memorize a bunch of hyperfocal distances for different combinations of focal length and aperture. You should just be able to adjust focus such that the right edge of the highlighted area on the scale just barely touches infinity and you should be good to go. But if the scale is conservative and you do that, you should still have everything out to infinity in focus, bit you cheat yourself at the short end and you'll have less in focus up close than you should (if the scale was correct). This is not an issue for landscapes, but could be a huge issue for street shooting where you really want those closer distances in focus. So then you have to go back to the charts and calculators to figure out the true hyperfocal distance and the benefit of the automatic scale is lost.

But, as noted the key is to understand the concepts and work with the limitations of your equipment.

-Ray
 
So let's look at the worst case scenario.
The is off on the short end, we all agree on that.
What if the scale was off equally on both ends? Just a thought.
That could explain many complaints about infinity focus being off.

It appears that the near end is off by 1/2 or double.
Look at what BB posted for me. It clearly shows how Fuji misinterpreted the inverse square law.

Ray, as a good streetshooter, I'm sure you see the issue at hand.
Peter, for your landscapes and portraits... This could be critical.
Armando.... Ya gotta see this, right?
 
That could explain many complaints about infinity focus being off.
...
It appears that the near end is off by 1/2 or double.

Both of those scenarios are very easy to test as I suggested in my last post. It has nothing to do with depth of field. It would be best to test it rather than reiterate other's comments and perpetuate misconceptions between depth of field and correct focus.
Just focus on something close enough to infinity and check the focus scale. I would expect the distance would also be in the EXIF file if using AF.

Then focus on something closer, say 1m or whatever is convenient compared to the markings on the focus scale. Again you should be able to check against the EXIF file on most cameras.

I know people that manually mark their desired depth of field markings on lenses with scales but no depth of field markings. Not hard.
 
Well, if you set the dog scale to hyperfocal distance... Infinity is not in focus...
I talked to someone at Fuji who verified my findings and hopefully Fuji will make a correction.
There's some shooters that think this is a non issue and that's fine but there are a few I respect highly that agree with me so I'm content in knowing I'm not being OCD or crazy.
Later...
 
Even if hyper focal distance is based on subjective perception of sharpness, there's no point in Fuji doing differently than how every other DOF calculator does it. Plus the fact that Fuji's DOF indications seem to match those for a 35mm lens, I think that indicates a mistake was made by a programmer who forgot the actual focal length of the x100 lens...

Fortunately this is a easy fix, thanks Don for contacting them.
 
Not to seem to disagree with any of the conclusions here, but there is no substitute for experience with a piece of equipment. When the instinct takes over, not the technical knowledge, or the posted readings, etc. IE: "Beware the man that only has one gun and uses it well."
 
Verified: Camera's CoC diameter is 0.007mm

IMHO, Fuji's DoF scale isn't plain wrong, but Fuji is using a rather unconventional CoC value for its DoF calculation. Obviously, Fuji assumes that an average digital print of an X100 image file is likely to be heavily cropped, thus requiring the use of a much smaller CoC value.

I reverse-engineeered which Circle-of-Confusion (CoC) value Fuji is using for the X100's DoF scale: It's 0.007mm - a very unconventional value for an APS-C sensor.

I managed to find a camera/lens state in which the DoF scale's displayed DoF range extended from just below 1.5m to just above 3.0m. I achieved this with an aperture of f16 and a focus distance of about 2.1m. Pretty lucky that the distances of 1.5m, 2.0m and 3.0m are actually marked on the scale (there aren't any others in this distance range)...

I then opened the Dofmaster online DoF calculator, selected 23mm as focal length and f16 as aperture, entered 2,1m as focus distance and then opened the "Camera, film format, or circle of confusion" drop-down list and selected 0.02mm as CoC value (very far down at the bottom of the list, below all the camera models). The calculator delivered a near limit of 0.93m and infinity as far limit - so, the 0.02mm CoC diameter cannot be correct! I then proceeded to try out other CoC values until I found a CoC of 0.007mm which delivered 1.46m and 3.75m respectively - close enough for the readings on my camera's non-linear DoF scale.

But the problem with the DoF scales in the X100 isn't just an unconventional calculation, but also a rather clumsy display implementation: The extremely coarse numbering of the camera's highly non-linear distance/DoF scale makes using it difficult if not impossible.

My suggestion for a solution in a future firmware update is a redesign of the camera's distance and DoF scale:

The camera's DoF scale should dynamically display a changing distance range which depends on the current values of

  1. lens focal length
  2. working aperture, and
  3. measured/set focus distance

In this way, the scale will make much better use of the real-estate in the VF/monitor display, so that numbers can be displayed with a much higher resolution. The displayed distance range of the scale should be updated after a shutter button half-press (or in real time when in manual focusing mode combined with manual exposure /aperture priority mode), and should always be a tad wider than the current DoF range so that the near/far limits are clearly readable. It should be possible to just display the distance values of the focus distance and the near and far DoF limits (plus the infinity mark). Anything else would be overkill.
 
Arjay, I was lost about everything in the top part of your post (my personal circle of confusion), but your last 2 paragraphs are dead on the money. Why it's not working like that now seems ridiculous. Is Fuji working on straightening out the DoF scale?
 
Beating a dead thread here, but anyways-

IMHO, Fuji's DoF scale isn't plain wrong, but Fuji is using a rather unconventional CoC value for its DoF calculation. Obviously, Fuji assumes that an average digital print of an X100 image file is likely to be heavily cropped, thus requiring the use of a much smaller CoC value.

I reverse-engineeered which Circle-of-Confusion (CoC) value Fuji is using for the X100's DoF scale: It's 0.007mm - a very unconventional value for an APS-C sensor.

Don 'reverse-engineered' Fuji's DOF calculations the same way, and came up with 35mm focal length and .02mm COC. Both yours and his scenario match up, so we have no idea what Fuji's reasoning was (both are reasonable to me.) In the end it doesn't matter- the only people who care seem to want a more conventional DOF scale, people who don't care, well, don't care. And regardless of how and why, the DOF is wildly off compared to generally agreed upon standards.

My suggestion for a solution in a future firmware update is a redesign of the camera's distance and DoF scale:

The camera's DoF scale should dynamically display a changing distance range which depends on the current values of

  1. lens focal length
  2. working aperture, and
  3. measured/set focus distance

In this way, the scale will make much better use of the real-estate in the VF/monitor display, so that numbers can be displayed with a much higher resolution. The displayed distance range of the scale should be updated after a shutter button half-press (or in real time when in manual focusing mode combined with manual exposure /aperture priority mode), and should always be a tad wider than the current DoF range so that the near/far limits are clearly readable. It should be possible to just display the distance values of the focus distance and the near and far DoF limits (plus the infinity mark). Anything else would be overkill.

This is a very interesting idea, I had to read twice before I understood! I think having a shifting distance scale adds some cognitive burden, but if there's some latitude far and near it could be easy to work with.
 
Back
Top