EVF Focusing - why not overlapping images like an RF?

[kxl]

With the the brilliant engineers at Sony, Fuji and other companies, why hasn't anyone come up with an EVF that shows overlapping images like an RF for manual focusing (including legacy lenses using adapters)?

I'm no engineer but it seems to me that EVF is just a presentation layer, so engineers should be able to develop an interface that displays overlapping images like an RF.

Is it a matter of financial feasibility?

I for one would be interested. I would think that many buyers of such cameras would be interested as well.

 

[Matus]

You would need another sensor & optics (like with true RF - there you have 2 viewfinders) to get different perspective and to have image to overly onto the image from the main sensor. The precision would not be better than with rangefinders either.

Of course it would be possible, but I guess there must be better way with digital cameras.

 

[ruby.monkey]

It seems like a pointless complication in a system that doesn't use a triangulating rangefinder; and I suspect that the vast majority of buyers of EVF-equipped cameras won't miss it one bit.

 

[Joosep]

Yeah, the EVF rangefinder doesnt make alot of sense.

You do have splitscreens in SLRs, maybe would be possible for a digital splitscreen ?

 

[kxl]

It seems like a pointless complication in a system that doesn't use a triangulating rangefinder; and I suspect that the vast majority of buyers of EVF-equipped cameras won't miss it one bit.

Not sure it I'd consider it pointless. I would think that there are many RF users who buy EVF-equipped cameras - just not sure if those numbers justify the cost of building in the overlapping image, but if the numbers justify it, I would think that such a feature would be a selling point for that share of the market.

 

[kennylovrin]

I think it makes perfect sense - it's not about a rangefinder per se, but rather about a rangefinderesque focusing aid. That is, exactly the same kind of feature as focus peaking, just a MF aid.

I actually thought alot about it when I had my X100, as potentially fuji could overlay a sensor feed over the OVF, and theoretically get the same effect as a rangefinder.

There are other issues though, optical + evf might not give optimal sense of when in focus. Also, the lens would have to stop down a lot to be able to act as anormal rangefinder where there is no "DOF effect".

Perhaps a wide open feed of the sensor on top of the OVF would work though, kind like a blurry ghost image. 😉

 

[sig]

Are we discussing a mechanical rangefinder or focus peaking that looks like a rangefinder?

If it is the latest it is prob not made because it is better to have focus peaking all over th frame.

 

[rogue_designer]

In order to make it work that way, it would have to either have two sensors for triangulation, or some way to tell which elements were in front of others (not just out of focus, they could be behind instead). At the very least, it would have to be running two separate sets of calculations and then comparing them and generating a new image set - which is not insignificant from a processing standpoint.

Focus peaking and other contrast based focussing works on a different principle, and as far as I can tell, cannot easily be extrapolated so.

 

[sig]

But you should be able to fake the look by software?

 

[geotrupede]

two sensors or a wide one and considering the shift of the view by adjusting the crop of the small image. possible.
but to be fair, if you are trying to get aligned two images you need a very high resolution.
i suspect that the current EVF are stone age compared to the resolution you would need to precisely overlap images...
but why not, if it is cheaper than the real thing... (can it be cheaper?)
G

 

[rogue_designer]

fake the look? sure - at least with electronically connected lenses to get focus distance information. But surely the whole point is that it would be functional?

 

[sig]

I had the impression that focus peaking worked on manual lenses. Is it not so? And based on that you could give focus confirmation by a yellow diamond.....

 

[rogue_designer]

Yes, focus peaking works on manual lenses. It shows what parts of the image are in focus under most non-challenging conditions. When would it show the diamond? How does it know which subject/part of the composition is your focus point?

I think you're looking for a solution to a non-existent problem.

 

[msbarnes]

i do not see how this focusing feature is stupid; especially not after smile detection.

 

[sevo]

I had the impression that focus peaking worked on manual lenses. Is it not so? And based on that you could give focus confirmation by a yellow diamond.....

"Focus peaking" is really contrast peaking - it is up to the user how to interpret it as a focus confirmation. While this allows to display the location of highest contrast (which usually will be any structured part of the subject which is in focus), it cannot tell anything about unfocused or low contrast bits of the image area. Moreover there is no way for a camera to determine the current setting of a uncoupled manual focus lens (short of impractical solutions like the barrel OCR recently patented by Epson or a secondary AF camera to compare the image with), so it cannot show which way to focus. We'd expect a rangefinder to be accurate even when the camera passes a pattern impossible to focus on, and we'd expect it to give the right direction - neither would be possible, making such a rangefinder imitation more frustrating than the existing EVIL focus aids.

 

[bwcolor]

It could be done. You would lose off center info. available with focus peaking, but you could have both in one system. I suspect that most rangefinder people prefer ovf to evf, so the Fujifilm could appeal to ovf folks. One of the things missing at Fuji is focus peaking.

 

[anerjee]

I'm not sure if this can be done. The EVF view is through the lens, so it will not always be sharp unlike an RF VF. How can this image be used for co-incident type RF focusing?

What may be interesting is a quicker way of focus confirmation than zooming in.

In my experience, focus peaking is useful but I could never be confident without zooming in 4x. The peaking/zooming process is fast, but very unlike focusing with an RF. It is also a lot less fun.

With the the brilliant engineers at Sony, Fuji and other companies, why hasn't anyone come up with an EVF that shows overlapping images like an RF for manual focusing (including legacy lenses using adapters)?

I'm no engineer but it seems to me that EVF is just a presentation layer, so engineers should be able to develop an interface that displays overlapping images like an RF.

Is it a matter of financial feasibility?

I for one would be interested. I would think that many buyers of such cameras would be interested as well.

 

[kxl]

As others have suggested, I'm asking about an electronic presentation that shows overlapping images, but obviously it has to be functional.

IMO, that kind of feature perpetuates the RF-shooting experience.

 

[rluka]

Use a very cheap and low res variation of Lytro's camera to work the rangefinder and then set the focus automatically like SLR camera stopping down the lens automatically when shooting ?

A question. Isn't DSLR's phase detect AF works like rangefinder in certain way ?
It uses beam splitter and then try to merge the two resulting images to get a focus confirmation. There has also been patents for phase detection for CSC.

I'm not sure if this can be done. The EVF view is through the lens, so it will not always be sharp unlike an RF VF. How can this image be used for co-incident type RF focusing?

The image would also be of different size/magnification, unless the rangefinder sensor has complex magnification system to adjust with the attached lens.

Other thing. To make it simpler, it should be in the same horizontal axis as the lens right ?
Left side is blocked by the hand turning the lens, while the right side is limited by the space available between the lens mount and the user's gripping hand. Most likely will have low baseline accuracy

 

[rogue_designer]

To make it simpler, it should be in the same horizontal axis as the lens right ?

Not strictly necessary if you're only dealing with a single point/focus area. Makes it easier mechanically, but would work just fine either way from an accuracy standpoint.