OVF Framelines

Broke

Broke

Established
Local time
6:31 AM
Joined
Apr 11, 2007
Messages
99
View My Gallery
Hi all,
I've had my X100 about a month now, and have enjoyed it a great deal. I've tended toward use of the OVF for framing, and for the most part I've had a reasonably good experience with this. In my opinion, this really is this camera's raison d'etre.
I must say that I'm puzzled by a few choices of the Fuji engineers, however. They've chosen to make the lines very conservative in terms of coverage (like Leica) to ensure accuracy at shorter distances. My question is, given the opportunity to design this camera, wouldn't you vary the size AND position of the framelines for any given focus distance. And while you're at that, wouldn't you move the focus patch to correspond with the image centre (where focus is actually obtained) after focusing? These are images from a projector after all... It really would be nice if, after having had focused, one could determine what he had focused on, and what was in frame.
Sigh. This camera is really impressive in some ways, however, I wish they'd kept it in development a couple months longer.
Cheers,
Jim
 
Well, I'd rather have too much in my frame than not enough. With an optical VF, it can only be accurate at one distance really. However, consider yourself lucky to have an EVF as well for the times when you need 100%.

To me, the camera works fine as is... ideally, yes, I would like someone to make a OVF that is accurate at all distances, but it probably will never happen.
 
"With an optical VF, it can only be accurate at one distance really. "

This is my argument -- if you can continuously vary the projected size of the framelines, why does this need be?
Cheers,
Jim
 
This OVF works like every other OVF. Decades ago people migrated away from cameras with OVFs to reflex cameras. One reason why is the OVF view can only be an approximate.

Fortunately the EVF give you a precise view when needed.
 
"This OVF works like every other OVF"

I don't believe this to be true -- this is the first ever OVF with framelines projected by an electronic projector, they can do whatever they wish with what's projected. This is in contrast to the Leica framelines which are simply created by frameline masks. I think that not trying to correct for size of field and for parallax is just lazy...
 
It may be intellectually lazy to call an engineering solution lazy without attempting to understand why something was done and what the alternatives might have been.

The easy answer is that the camera doesn't commit to focusing and frameline size/position until you press the button. Your focusing spot would be jumping all over the place as you successively adjusted the framing and refocused, and it would be worse in AF-C mode. Something like this happens with Leicas, too, where the focusing spot moves with the framelines as you focus closer - but it is much slower and controllable.

In reality, if you fill up the focusing spot and take the split second to look at the distance scale, you rarely get problems - and you can at least see when the system has missed. And the difference between intended and actual focus point is pretty insignificant until you get closer than 1m - which is pretty much the usable limit of an OVF.

Dante

Broke said:
"This OVF works like every other OVF"

I don't believe this to be true -- this is the first ever OVF with framelines projected by an electronic projector, they can do whatever they wish with what's projected. This is in contrast to the Leica framelines which are simply created by frameline masks. I think that not trying to correct for size of field and for parallax is just lazy...
Click to expand...
 
My surmise is that Fuji picked a frameline size that corresponds to 1m-1.5m (i.e., already "shrunk") to keep the finder magnification higher. It's a pretty significant size change from ∞ to 1m.

Dante

ferider said:
The focus spot is one thing, Dante.

But why not allow sizeable frame-lines as in Hexar AF or Contax G ?
Click to expand...
 
Ferider, I checked this in the VF - you can see the shrinkage. Put the camera in MF, and dial the focus down to 0.6m. Now tap the shutter button and you will see the left and top lines move more than the right and bottom ones do. Voilà. Field correction. I stand by my prior point. Fuji had this same limited range with the LCD framelines in the GA645zi.
 
There's yet another reason why the frame lines are tighter than the lense's actual field of view: Eye-to VF axis alignment.

If you look through the optical viefinder and move the optical axis of your eye in relation to that of the camera's viewfinder, you'll notice that the projected VF frame doesn't move, but the actual optical viewfinder's field-of view will.

You can see this most dramatically in the relative change of the lense's edge in the bottom right of the VF.
 
Broke said:
"With an optical VF, it can only be accurate at one distance really. "

This is my argument -- if you can continuously vary the projected size of the framelines, why does this need be?
Cheers,
Jim
Click to expand...


Because you can have shrinking framelines, but you still can't account for parallax error, because the VF is in a different position to the lens. Hence the need for conservatism. An OVF is always a compromise.
 
This thought was mentioned by Nick Devlin in his review also...
http://www.luminous-landscape.com/reviews/cameras/.shtml

"Frame Lines
This was a surprising disappointment: the frames lines projected into the OVF are just not very accurate.* Like the M9, the framelines are significantly under-inclusive, meaning you get a lot more into the image than the lines indicate.* This*is annoying.* The Leica M series are limited by the fact that the framelines are mechanically projected. The X-Pro1, however, uses electronically displayed lines. These could be projected anywhere, can be *changed* in size and, one would have thought, been made wickedly accurate.* Or at least better than they are.* Since the frame grows and shrinks with subject-distance, the projections could perhaps re-size after focus is locked? The projected "guess-points" for focus are pretty accurate. I can think of no obvious reason why the framelines can't be made to be as well. Perhaps an ambitious Fuji tech can reprogram these, but I won't hold my breath. In the meantime, I will do what I already do with the M9 and mentally expand the frame."
 
Dante_Stella said:
Ferider, I checked this in the VF - you can see the shrinkage. Put the camera in MF, and dial the focus down to 0.6m. Now tap the shutter button and you will see the left and top lines move more than the right and bottom ones do. Voilà. Field correction. I stand by my prior point. Fuji had this same limited range with the LCD framelines in the GA645zi.
Click to expand...

Yep, the frame lines certainly do shrink at shorter focussing distances.
 
Okay, here's how I now think about it. Even with the electronically-adjusted frame lines, the camera can still never know the exact image borders. The one thing the camera can accurately do is adjust the focus point for parallax, because it knows the exact distance of the focus point. It doesn't know anything about the rest of the frame!

Let's say you focus on a brick wall 2 feet in front of you. The x100 knows your center point focus is 2 ft., and tries to approximate the frame lines. But it doesn't know the distance of what else is in the frame (in this case everything is up close, there is parallax error.)

Let's say you focus on a flower 2 feet in front of you, with nothing but open horizon behind the flower. The camera has the same AF distance info of 2 ft., and will probably project the same frame lines. But I am thinking that the actual image borders will be different compared to the brick wall example right? Regardless of the close subject distance, most of the frame is at infinity so there is essentially no parallax error in this case.

The best Fuji can do is assume that if AF point is close, it can adjust for parallax correction so your subject will be about where you want it to be in the final image. And if the AF point is far, then the most important parts of the final image are far so little to no parallax correction should be applied. But no matter what it is only a guess, a guideline- it cannot know the actual resultant image borders.
 
You must log in or register to reply here.
Back
Top Bottom