A true digital rangefinder?

[Roger Hicks]

Still, with the increase of EVF quality it should be no biggie to create a rangefinder-like focus system that is mostly digital-based.

First of all a bit of understanding the RF experience: when focusing a Leica, it's in fact two different methods you are using at the same time. One is the superimposed images, but the other comes from the sharp-edged RF patch. Line up the objects inside and outside the patch and you're in focus. This aspect is what is needed for a digital RF mechanism.

The X100 projects digital data in the optical viewfinder, right? So, why not take a square bit of the sensor image and project it in the middle of the optical viewfinder? It does not have to be superimposed, just the sensor image! When the camera is in focus, that little square will be in focus, and it will line up with the optical image.

Along with that it'll be quite easy to have framelines in the viewfinder, like any rangefinder camera.

I really cannot understand why no manufacturer is getting into this. Like I said, I presume that it is a much cheaper to produce RF system that will sufficiently mimic the Leica RF.

But at that point you need to 'throw away' a good deal of the potential resolution of the imaging chip, using it only for the viewfinder and not for the imaging area. Looking through the finder if my M9 (the nearest Leica to hand) I reckon that no more than 2/3 of the viewfinder area is occupied by the 50mm frame. In other words, to equal the 18 Mp resolution of the MP you'd need a 27 Mp sensor. Who is going to go for that?

Unlike you, I find it very easy indeed to see why no manufacturer would bother to make a poor, electronic imitation of an optical rangefnder.

Cheers,

R.

 

[Roger Hicks]

Just wait and see, the "M" will be quite nice to use... No autofocus but red lines on the image where the pinpoint focus will be, no room for focusing errors...

Hmmmm.... Have you tried it? I have. Why do you think they left the 'real' finder in?

Cheers,

R.

 

[Roger Hicks]

very few photographers,not counting view camera shooters, are using the depth of field scale propely. My father, an amateur photographer shooting with leica already in the 30`s taught me the virtues of it when I got my first M2 and 35mm summicron. The depth of field scale was an invention by a finn VILHO SETÄLÄ, a friend if my father. It was adopted by leitz when his leica was sent to factory for repair and they copied it from his lenses. That was the time when rangefinder was the extra gizmo on top of the Leica I. It still has the same function, for those that are smart enough to use it!

Fascinating! And an Esperantist too! Thanks.

Cheers,

R.

 

[bigeye]

So maybe a stupid question but given the advancement in digital focusing ie. focus peaking, auto focus etc, why can't there be a digital rangefinder type of focus?

Effectively you could have an EV with a square patch digitally super imposed to the image, and as you focus on an object, the square aligns

or even better this could be applied to a hybrid optical/EV (ie Xp1). So the image is via the optics but the rangefinder operates digitally

This would remove the cost of the optical rangefinder in a camera like the M9 but still keep the same user experience and couple with existing lenses

thoughts?

I assume that you're trying to replace the mechanical rangefinder in Leicas with a focus-peaking patch, keeping manual lenses? The focus patch could light up when whatever it covers is in focus.

The technology is already available; when you hang a Leica lens on one of the CSCs with focus peaking you get a geological survey chart of red 'horizontal elevations' (well, what it looks like to me) that are in focus, although not the targeted, conventional RF patch you describe. Looked at another way, it would be decoupling the center focusing point in a Nikon from the AF system.

This would be slower than optical RF, though. With a RF patch you can see a super-imposed image describing which way you have to turn the focus ring and by how much, it gives you a solid, quick gauge of "how much" that a binary indicator does not. Also, you can also choose smaller elements within the patch (e.g. an eye) that you can concentrate on, which is more accurate. (The AF boxes on my DSLR are pretty big and I often wonder what inside the box it is actually measuring.)

The trick to your idea is to come up with a way that is as natural and simple to focus as an optical RF.

Perhaps, different shades of color for near/far approaching peak. Also, different sized patches for finer target distance measurement.

-Charlie

The choice of distance measuring technology (phase detect, contrast, radar, string, laser, IR, microwave, loran, sonar, DME, etc.) is a separate question...

.

 

[Fabio Ruffet]

announcement from Fuji's new x100s - well it seems this wasn't a crazy or impossible idea after all...

Also new to the FUJIFILM X100S is the world’s first (*5) “Digital Split Image” feature, which displays dual images on the left and right to be lined up for manual focusing. This enables accurate focusing especially when working with an open aperture or macro shooting.

 

[Roger Hicks]

announcement from Fuji's new x100s - well it seems this wasn't a crazy or impossible idea after all...

Also new to the FUJIFILM X100S is the world’s first (*5) “Digital Split Image” feature, which displays dual images on the left and right to be lined up for manual focusing. This enables accurate focusing especially when working with an open aperture or macro shooting.

Did anyone say it was impossible? I don't think so.

As for crazy... Well, we'll have to wait and see (a) what it's like and (b) what the market says.

Cheers,

R.

 

[johannielscom]

announcement from Fuji's new x100s - well it seems this wasn't a crazy or impossible idea after all...

Also new to the FUJIFILM X100S is the world’s first (*5) “Digital Split Image” feature, which displays dual images on the left and right to be lined up for manual focusing. This enables accurate focusing especially when working with an open aperture or macro shooting.

Nice.

Let us hope they have implemented this idea sufficiently accurate to show all the nay-sayers that it can be done and actually is a good idea. Lining two image halves up in the finder would be technically easier to implement.

I've been pleading for a focussing like this from the moment they announced the X100. I've several times written Fujifilm through the X100 feedback forms from the moment the X100 announcing website was online, pointing out this was an easily achievable option that might bring them more users.

If there's any feel of mechanics to the lens of that new X100S, I might now actually get me one. Sticky blades on the aperture will likely have been cured with this new carnation of the camera. The lack of Rf-like focusing, sticky blades and the lack of any mechanical feel to the focusing were the reasons that kept me away from the original X100.

 

[JoeV]

Many Thanks for the kind explanation but I have a couple of Leicas and fuji and I understand this principle. My question is, why does this have to be done optically. Why can't the same principle be replicated digitally?

Take a look around in your own household right now and you will see things that were unthinkable 5 years ago are just built into simple everyday products. So the answer can't just be technical difficulty. Lack of demand versus investment would be a more plausible explanation though

But again, this could be done mostly or entirely via software as the camera already 'knows' what's in focus and what isn't (that was my original badly phrased point)

You say in another response that you want to retain the manual focus experience, but want the viewfinder to display the rangefinder patch electronically. This means that you are suggesting replacing a currently all mechanical system with a combination of mechanical and electronic components. If you want to manually focus, the user will be operating a focus ring by hand, which is a mechanical component. Now, do you suggest this new system dispense with legacy M-series optics and instead go to a new lens format with digital encoders for focus position? Or are you suggesting a mechanical to digital position encoder be employed in the new camera body, to convert the lens focus position to digital information?

Either way, the new proposed hybrid system will be at least as complex, and perhaps more costly, than the existing all-mechanical method.

Of course it's not necessary for me to state that the real world of lenses and optics is physical, especially if you are intending on employing manual focus. The only advantage such a hybrid mechanical-electronic focus indication would provide would be the possibility for electronically calibrating individual lens focus cams to compensate for physical wear. But Leica would have no incentive to undermine their current repair business with such a technology.

You haven't addressed the advantage of current RF VFs as enabling the user to see outside the sensor area, to anticipate subject motion.

Lastly, you state that the camera would already "know what's in focus", but it does not know what should be in focus, that's the job of the photographer.

~Joe

 

[larmarv916]

Things are going to get real interesting next week. Fuji is having a major international release of many products in Las Vegas at CES. As well as all the other major camera producers...to put your seat belt on !

announcement from Fuji's new x100s - well it seems this wasn't a crazy or impossible idea after all...

Also new to the FUJIFILM X100S is the world’s first (*5) “Digital Split Image” feature, which displays dual images on the left and right to be lined up for manual focusing. This enables accurate focusing especially when working with an open aperture or macro shooting.

 

[lcpr]

I'm bored so I've decided to come up with this. Let's say we're replacing the Leica M's mechanical RF system with an all electronic one, you'd need:

- A way to project the secondary image into the OVF: maybe a smaller, lower res camera module built into where the RF window is now. Let's call that the RF camera. The image from the RF camera is then projected Fuji-hybrid-VF style into the center of the OVF

- The lens is mechanically linked to the focusing roller in the body - the user will set start and end points for infinity and close focus as well at the distances indicated on the lens barrel when a lens is attached. Like 6-bit coding this will be part of the lens's profile. This is so the camera knows what distance the lens is focused at

- The roller is also mechanically linked to the camera in the RF window, so turning the focus ring will displace the image projected from the RF camera to the OVF along the X-axis

- The user will then align the projection from the RF camera with the desired subject/point of focus they see in the OVF

- Framelines and info are projected much like the Fuji way

However, I imagine that such a system would be just as complicated as the standard RF. What will need to be done is an abstraction of the way you focus an RF - you align two points together. That is always the case. So you can have a module in the RF window which will measure the distance to the subject (like a IR beam or some sort of phase detect). The lens will be linked to the body in the same way as above. To focus, half press the shutter to activate the beam (or it'll always be on, so no need for the shutter press) - this will bring up a graphic representation (a marker of sorts) of the point of focus in the OVF. Since the camera will know what distance the lens is focused at (mentioned above) a second marker will be projected in the OVF - that one represents the distance the lens is focused at. Furthermore, tertiary markers could be placed either side of the RF distance and Lens distance markers to signify depth of field at certain apertures (much like the patches in some M3s/M2s). The user will then align the two markers to achieve focus. See the diagram below:

The only mechanical part will be the linkage between the roller in the body and the lens. The rest would be electronic. It's fundamentally the separation of the distance acquisition and lens focusing that's inside a DSLR. In fact, Nikon DSLRs do this with manual lenses, but the way they represent which way to focus via LEDs isn't graphic enough. I imagine that Fuji have gone down a similar route with the X100s (not in terms of implementation, but the separation of the two stages).

 

[lcpr]

double post, delete

 

[JoeV]

A hybrid form of camera user interface is an intelligent approach, and I at the very least appreciate Fuji's first attempts in this area. But we should not wish for a certain part of a camera system to be electronic merely through some philosophical preference uninformed by actual knowledge of the technology employed. There are some areas of a system that would be optimally engineered to be mechanical, while other areas would more optimally be electronic in function.

Specific to viewfinder technology, any form of EVF will have an intrinsic delay between a scene's activity and the user's response to what's seen in the EVF, it's just a fact of reality that electronic signals take time to process. A relevant question is whether future advances will make such a delay insignificant in actual operation. In my personal experience, the Lumix G5 EVF is pretty decent, but there is still a delay that can affect timely response to quick-moving action. In this regard I suspect an optical VF will always be superior.

Regarding focus determination, a non-TTL manual focus system will always use some form of mechanical interface coupled to the lens, regardless of whether the overlay patch is optical or electronic, or whether the lens' focus information is relayed mechanically or electronically.

But in terms of speed of response for a manual focus system, an optical VF and mechanical coupling of lens to camera will always be faster than a hybrid mechanical - electronic system. This is because, in the real world of electronics, signals are processed in real time, not instantaneously like in the movies, and employing servo or stepper motors to move a lens element or focus prism, or employing digital position sensors over direct mechanical lens control, will always present some additional delay that could be prevented by using an all-mechanical, directly hand-operated system. The trade-off is that accuracy of composition and DOF preview are compromised with the present all-mechanical M-system, as compared to EVF-equipped cameras.

There might never be an optimally perfect camera system that can do all things ideally suited for all people, so perhaps having choices available is a good thing after all, with the present all-mechanical M-system being one of those choices that we can be glad to have available.

~Joe

 

[lcpr]

I think that the problem most people have with the current batch of electronic focusing aids is that they aren't as objective feeling as lining up two points together - magnification and peaking, while a huge help, still feels a little subjective. It's difference between judging whether the image in a ground glass screen looks sharp or knowing that two points have been aligned together. Personally I think Leica should stick with an all mechanical system, but other companies who don't have a reliance on legacy lenses should look into ways of representing the point of (manual) focus more objectively.