Should the M9 successor have a Hybrid Viewfinder

[mjh]

Advertisement (gone after registration)

As I understand it, the German transcript, approved by Leica, is being translated into English by a human being.

I’m a human being too. And, as it happens, a native speaker of German. Why should I wait?

[Advertisement]

The Leica Q3 is available for pre-order at B&H Photo Video and Adorama

[bill]

Michael, I was trying to dissuade those with no knowledge of German from impatiently relying upon automated translation services and jumping to conclusions therefrom, as illustrated by your response.

 

Those who are bilingual of course already have an insight. However I do believe that we should work from a single "official" source of information in each language; the debate as to the accuracy of the translation should be kept separate.

 

Regards,

 

Bill

[diogenis]

So, possible weather sealing in a redesigned case (obviously), focus confirmation/assist not possible unless cmos sensor is used.. thats about it.

edit: Actually not bad at all: They still try to improve the model.

[Alberti]

I find the split-image focusing to be a vastly better experience with any lens in the Leica range and don't see an EVF or "focus confirmation" as anything but a step (or 12) downwards and backwards.

 

[---]

Also, until a few minutes ago, I didn't see how one could simultaneously project LCD framelines into the Leica-M viewfinder without interrupting the light path for the (absolutely essential manual) rangefinder. However, by adding another beam splitter to the system, I think I see a way. See diagram.

http://www.l-camera-forum.com/leica-forum/attachments/leica-m9-forum/222201d1284997441-should-m9-successor-have-hybrid-viewfinder-lcdframes-jpg

 

It would take precision to make sure the beam-splitter did not offset the RF image, since precise alignment is how we know things are in focus.

 

[..] .

 

Andy, I greatly appreciate this diagram you put together. It leads me to think out of the box a little bit more.

Hypothesizing:

1. the the beam splitter (double prism) in the viewfinder also will project an image to the left in the diagram. This image has passed through the front optics of the eye piece. If a (slim) optic is added on the left, then it can be made to focus on a specific plane.
   
2. On the left side a focusstrip could be added, say 20 pixels wide (conforming what I understand is the focus mechanism of the S2 and the old patents of Leica)
   
3. The lens focus can be variable to a minute degree, from an mechanical (oops, old fashioned and probably impossible to make) or electronic (truely innovative) mechanism. I am think of a system like Philips used to have in the famous and coveted System200 recorder: a piezo element that moves the optics a few hundred microns.
   
4. there is a capture of the lens distance tab, electronically, that passed it off for this movement, for instance coupled to the moving RF patch.
   
5. The now electronically measured distance, once matched in the camera processor to the lens length, can be tranferred and made available for an electronic frameline.
   
6. an electronic frameline can then move along the axis diagonally on the "matrix board" that is used for creating framelines. (I know this is a different concept from LED-lit framelines a la M9Ti)
   
7. Such a frameline can be projected by the double prism device that Andy points out, behind the second prism.
    
   

 

I am not giving any valuation on the viability, or adamantly saying this is something all should like.

Positive: it keeps the mechanical focus and RF patch, adds just a special indication somewhere in the screen on the focus. And the framelines are electronically projected. So in fact, it is hybrid. It helps those with failing eyes from the times of the Brownie:).

 

It is a way of keeping strong points, and extending the camera concept to fit production techniques to the 21st century.

 

For affectionado's, Leica can easily make a blackbox-MP: a sensor in an MP, just stores images, no fuss. That is the opposite of the scale, IMHO. There is production capacity over now I understand, in the near future.

 

Great conversation.

albert

[giordano]

 

Nice try, Andy, but you're making it look much easier than it is. The problem is that your diagrams omit all the lenses needed to bring the various elements - view through the finder, view through the rangefinder window, and frameline mask or LCD screen - all into a common focus. They're clearly visible in this diagram from Mr Puts's detailed article on Rangefinder Issues:

 

 

Something similar would be needed in and around your double beamsplitter or the information on the LCD simply wouldn't be in focus for the user.

 

Obviously - now one thinks about it - the Fuji X100 must have something more than a beamsplitter and an LCD or it would have the same problem. The way things work out, the lower the magnification of the finder, the higher the power of the positive lens at the eyepiece and the lower the power required in the field lens(es) nearer the LCD. This may explain Fuji's choice of 0.5x magnification: anything bigger would have meant making the camera bigger.

[giordano]

Hypothesizing:

1. the the beam splitter (double prism) in the viewfinder also will project an image to the left in the diagram. This image has passed through the front optics of the eye piece. If a (slim) optic is added on the left, then it can be made to focus on a specific plane.
   
2. On the left side a focusstrip could be added, say 20 pixels wide (conforming what I understand is the focus mechanism of the S2 and the old patents of Leica)
   
3. The lens focus can be variable to a minute degree, from an mechanical (oops, old fashioned and probably impossible to make) or electronic (truely innovative) mechanism. I am think of a system like Philips used to have in the famous and coveted System200 recorder: a piezo element that moves the optics a few hundred microns.
   
4. there is a capture of the lens distance tab, electronically, that passed it off for this movement, for instance coupled to the moving RF patch.
   
5. The now electronically measured distance, once matched in the camera processor to the lens length, can be tranferred and made available for an electronic frameline.
   
6. an electronic frameline can then move along the axis diagonally on the "matrix board" that is used for creating framelines. (I know this is a different concept from LED-lit framelines a la M9Ti)
   
7. Such a frameline can be projected by the double prism device that Andy points out, behind the second prism.

Thanks for the contribution. It's great that we've got some serious thinking going.

 

But I'm not sure that (1) above is practical. In the existing M viewfinder, the only place that a real image is actually focused is on your retina. The "image" passing through the beamsplitter is a virtual image from the (concave) front lense of the viewfinder; It would take much more than "a slim optic" to focus it on a little sensor.

[Alberti]

Advertisement (gone after registration)

Thanks for the contribution. It's great that we've got some serious thinking going.

 

But I'm not sure that (1) above is practical. In the existing M viewfinder, the only place that a real image is actually focused is on your retina. The "image" passing through the beamsplitter is a virtual image from the (concave) front lense of the viewfinder; It would take much more than "a slim optic" to focus it on a little sensor.

 

Hi Giordano,

indeed you've got a good point.

 

Well, for a focus confirmation strip we don't need a whole wide angle and full frame (diagonally full viewable area in viewfinder some 82-85 degrees, I guess).

 

• Just a few degrees of view in the center will work. In fact, roughly the size of the RF patch might be OK. Then there are very small lenses possible, like on a CD.
  
• Such lenses are positioned on the CD player extremely fast and reliable too (to keep along with small warps in a CD). And the positioning of such devices might be possible energy efficient.
  

 

regards

albert

[giordano]

Hi Giordano,

indeed you've got a good point.

 

Well, for a focus confirmation strip we don't need a whole wide angle and full frame (diagonally full viewable area in viewfinder some 82-85 degrees, I guess).

 

• Just a few degrees of view in the center will work. In fact, roughly the size of the RF patch might be OK. Then there are very small lenses possible, like on a CD.
  
• Such lenses are positioned on the CD player extremely fast and reliable too (to keep along with small warps in a CD). And the positioning of such devices might be possible energy efficient.
  

 

regards

albert

 

I'm not certain I understand what you expect to achieve by moving this little sensor anyway. Don't forget that the smaller the lens (i.e. the shorter the focal length) the greater the depth of field. Maybe you've done the calculations that will prove me wrong, but I doubt whether one could in practice use a (say) 5mm focal length lens and a tiny sensor to control the critical focusing of a 50mm f/0.95 lens in dim light.

[Alberti]

I'm not certain I understand what you expect to achieve by moving this little sensor anyway. Don't forget that the smaller the lens (i.e. the shorter the focal length) the greater the depth of field. Maybe you've done the calculations that will prove me wrong, but I doubt whether one could in practice use a (say) 5mm focal length lens and a tiny sensor to control the critical focusing of a 50mm f/0.95 lens in dim light.

 

Giordano,

Absolutely right, the RF is king of focussing, and any artifice for focus confirmation faces an uphill struggle.

By the way, my idea is to move the 'slim lens', not the little sensor strip. Indeed, one might face a small DOF of that lens, but it can be juggled and jousted around with in a mathematical model to find a workable solution in combination with a sensor, maybe without the microprisms... A little stepper motor could also work wonders, there are micro versions of that available.

No, I didn't calculate it, I can't competently do that.

And yes, agree, dimmed light, without an auxiliary lighting function, with all what that might imply in loss of resolution (noise) of such a strip spoils the function. But hey, then we still master the scene with our superior RF capability.

The whole idea I have of hybrid is adding, not replacing. And then it might have a lower quality in some circumstances. It definitely will never be equal or better.

regards

albert

[Bob Andersson]

Michael, I was trying to dissuade those with no knowledge of German from impatiently relying upon automated translation services and jumping to conclusions therefrom, as illustrated by your response...

Goodness, I felt rather like a motorist who's been pulled over for doing 31 mph in a 30 mh zone when I read that. But I dutifully held off from further comment until the official translation was available as I was always brought up to respect the Old Bill. The reason I then waited some more was that, in contrast to the German version which was featured on the Blog the English version was placed here in the Customer section so I was looking in the wrong place. The relevant part to this thread reads:

 

AF Help / Focus confirmation

The question was raised whether it is possible to integrate an AF Help into M9 or successors. The answer was no: The M9 is based on a CCD sensor, who cannot produce a live view. The sensor can only be read out 2 times in a second. In addition the sensor would warm up which would increase noise.

 

Leica prefers optical focusing over the rangefinder. AF functionality would be possible with a future CMOS sensor, which could be used for upcoming generations.

 

No statements were given to further developments of the M-System.

The first paragraph effectively says no live view so no hybrid viewfinder. The second paragraph goes on to say that Leica prefers rangefinder focussing (no surprise there) and then just restates the correlation of CMOS with Live View and offers that Leica could use CMOS in future.

 

So what's the likelihood of Leica using a CMOS sensor in a future M-body? They've shown no particular inclination to move away from CCD that I've seen but maybe I've missed something? Of course CMOS may be imposed at some stage if a suitable CCD can't be sourced or a suitable CMOS sensor is available at a cheaper price and at that point maybe hybrid viewfinder technology may be possible within the constraints we've discussed. Whether this will be a priority for Leica is something they are keeping close to their chests.

 

With luck the M10 should be available by the time I'm 60 (not too long now!) and I hope I can be tempted. I want that camera to be traditional "M" but, as the attentive reader has no doubt already deduced, I'd like the extra functionality of a hybrid viewfinder to be available on demand. Wish me luck.

 

Bob.

[henning]

So what's the likelihood of Leica using a CMOS sensor in a future M-body? They've shown no particular inclination to move away from CCD that I've seen but maybe I've missed something? Of course CMOS may be imposed at some stage if a suitable CCD can't be sourced or a suitable CMOS sensor is available at a cheaper price and at that point maybe hybrid viewfinder technology may be possible within the constraints we've discussed. Whether this will be a priority for Leica is something they are keeping close to their chests.

 

With luck the M10 should be available by the time I'm 60 (not too long now!) and I hope I can be tempted. I want that camera to be traditional "M" but, as the attentive reader has no doubt already deduced, I'd like the extra functionality of a hybrid viewfinder to be available on demand. Wish me luck.

 

Bob.

 

I should probably first take the time to read all 8 pages, but here goes anyway...

 

At the LHSA meeting in Wetzlar Stefan Daniel said that they were working on a live view-CMOS camera when asked about the 'solution' for R lenses. When asked whether that would be for R lenses only, he said that would make no sense, as the installed base is too small. It would be an M mount camera with an adapter for R lenses. When I asked whether it would have auto diaphragm and speculated that that would be required for R lens owners to prefer Leica's solution to Nikon or Canon adaptation, he said that they would look into it.

 

Further questions resulted in indications that this camera would likely be announced in 2012, and would essentially be the next M. Clearly they are in the early stages of design, and are probably talking with CMOS producers about the chip and their requirements. Some people spoke up about 'no anti-alias filter' and he indicated agreement.

 

Where this will ultimately lead is open, but Leica is definitely thinking about an EVF M. The Titanium M9 has LED projected framelines, and Leica seems to think this is a good idea at present, and from handling the camera I agree. If a traditional RF is maintained, I'm all for these projected framelines. If an EVF camera is to be the future of the M line, I'd like to wait to see what develops. I hope that the concept can improve, as right now I'm not in favour of replacing the M range/viewfinder with the current state of EVF's.

 

Henning

[Bob Andersson]

Hi Henning,

 

That's truly interesting and very much appreciated feedback from a primary source. I'm fascinated that Stefan is indicating that the R-lens solution may be via an M-body plus an adaptor and that he hasn't ruled out incorporating auto-diaphragm functionality for those lenses. My assumption is that such an R-lens solution would have to incorporate live view and it would also benefit from contrast based AF confirmation but I haven't followed the R-system debate very closely I'm afraid.

 

On a personal note I'll echo your wish that the M10 should not be primarily EVF and that's been pretty much the consensus of this thread. But if the EVF part of a hybrid viewfinder can be made truly invisible until switched on then I don't think many have objected strongly. And thanks for that indication of 2012 - I'll not hold my breath but guess when I'm 60!

 

Bob.

[dseelig]

For me no just a focus confirmation chip and better high iso that is all I could want. Actually I just want higher iso.

[dseelig]

One other thing I do not want is a cmos sensor.

[Bob Andersson]

...would also benefit from contrast based AF confirmation...

Whoops, can't edit the original typo but I should, of course, have written "focus confirmation" and not "AF confirmation".

 

Bob.

[henning]

One other thing I do not want is a cmos sensor.

 

I really don't care at a primary level whether it's CMOS, CCD or whatever. What matters is the implementation, and if it is possible to get the same image quality from a CMOS as from the Leica M9 sensor, who cares?

 

From what I have seen to date, I agree that the CMOS results do not equal the best of the CCD results, but surely there is room for development. I don't think that Leica is interested in taking a step back in quality.

 

CMOS has a number of advantages over CCD (as CCD has over CMOS). Let's hope that any CMOS implementation can result in an overall advance.

 

Henning

[Bob Andersson]

.

Hmm. Having finally seen the Fujifilm X100's hybrid viewfinder in action my personal reaction is that the full EVF mode is horrible, giving me a feeling of complete disconnection from the scene being photographed. Using it in optical mode with the hybrid aspect just being used to superimpose frame lines and shooting information worked a lot better but I've definitely come away from the experience believing that a full-on hybrid viewfinder should be kept as far away from a Leica rangefinder as possible unless the technology gets a lot, and by a lot I actually mean hugely, better and maybe not even then!

 

Bob.

 

P.S. Apologies for resurrecting this thread but I wanted to share my experience and effectively vindicate the views of a number of contributions to the thread which were, as it turns out, quite rightly IMHO very suspicious about the appropriateness of a hybrid viewfinder.

[Washington]

Hey: I just don't like the damn electronic viewfinder. OK?

[Advertisement]

Hallo,

Look here: Leica M

[AlanG]

 

At the LHSA meeting in Wetzlar Stefan Daniel said that they were working on a live view-CMOS camera when asked about the 'solution' for R lenses. When asked whether that would be for R lenses only, he said that would make no sense, as the installed base is too small. It would be an M mount camera with an adapter for R lenses. When I asked whether it would have auto diaphragm and speculated that that would be required for R lens owners to prefer Leica's solution to Nikon or Canon adaptation, he said that they would look into it.

 

 

This has always seemed to me to be the logical and obvious direction for Leica to go.