Using Leica M lenses on the Panasonic G1 - some pictures

[sean_reid]

Advertisement (gone after registration)

My own idea for the electronic rangefinder is a bit different. There would be one simple rectangle (LED projected maybe) in the center of the frame lines and another that moves side to side but exactly overlaps the first when focus is correct. So there would be no EVF projected - simply two little box outlines, one of which would move in response to the camera evaluating the actual focus relayed by the sensor. In other words, the projected boxes would simply be a system for displaying progressive focus accuracy - simple LEDs or the like.

 

This would be a camera that could use MFT or 4/3 lenses for autofocus and, with adapters, a wide range of lenses for MF - with RF focus patch-esque focusing. The same AF system would inform both AF and the electronic RF.

 

I'm interested in a fairly simple system that is robust and reasonably inexpensive to develop and produce. It would likely rely on AF technologies that already exist but would use a different way of displaying MF focus accuracy. This would be an MFT camera with a window finder that could use a variety of lenses - not only RF lenses.

 

The challenge that occurs to me is having accurate reading of peak focus (at the sensor) when a (non-electronic) lens was stopped down.

 

The camera would not necessarily have an EVF but it would provide live view via the LCD (when needed).

 

Cheers,

 

Sean

[Bob Ross]

My own idea for the electronic rangefinder is a bit different. There would be one simple rectangle (LED projected maybe) in the center of the frame lines and another that moves side to side but exactly overlaps the first when focus is correct. So there would be no EVF projected - simply two little box outlines, one of which would move in response to the camera evaluating the actual focus relayed by the sensor. In other words, the projected boxes would simply be a system for displaying progressive focus accuracy - simple LEDs or the like.

 

This would be a camera that could use MFT or 4/3 lenses for autofocus and, with adapters, a wide range of lenses for MF - with RF focus patch-esque focusing. The same AF system would inform both AF and the electronic RF.

 

I'm interested in a fairly simple system that is robust and reasonably inexpensive to develop and produce. It would likely rely on AF technologies that already exist but would use a different way of displaying MF focus accuracy. This would be an MFT camera with a window finder that could use a variety of lenses - not only RF lenses.

 

The challenge that occurs to me is having accurate reading of peak focus (at the sensor) when a (non-electronic) lens was stopped down.

 

The camera would not necessarily have an EVF but it would provide live view via the LCD (when needed).

 

Cheers,

 

Sean

Hi Sean,

I like that idea and if you tied the live view sensor to the focus indicator, you'd get electronic focus confirmation as well as visual.

In reading about contrast detection AF used in the G1, a comment was made that CDAF can't tell if you are front focused or back focused, where phase detection AF can. I'm not sure if I have this right, but it would be a problem in using the sensor for focus confirmation that would tell you if you are front or back focused. Contrast detection AF lenses have to be faster computationally to do all the trial and error testing to arrive at the correct focus. This is the reason the older FT lenses don't all do AF with the mFT system or are slower focusing than mFT lenses.

This has been a fun thread, with lots of good ideas.

Bob

[marknorton]

Sean, I see, so you're not suggesting injecting a rangefinder patch image sample from the live view sensor, simply having two rectangles which align when the AF thinks it's in focus and when it's out of focus, the lateral misalignment shows the direction and extent of the mis-focus.

 

That simplifies things by not needing to scale the sensor image and, artefacts aside, an EVF showing just the two rectangles (+ framelines) injected into the viewfinder would do the job. LEDs could also do it but to display what's needed, you either need an array of LEDs or a swivelling LED + mask which would take you back to the mechanical system we have now.

 

Only concern is that the entire focus evaluation is being put in the hands of the AF system without any user override.

[marknorton]

Bob, interesting point, the camera would have to be able to tell the difference between front and back focus to determine which side to display the "moving" rectangle.

[sean_reid]

Sean, I see, so you're not suggesting injecting a rangefinder patch image sample from the live view sensor, simply having two rectangles which align when the AF thinks it's in focus and when it's out of focus, the lateral misalignment shows the direction and extent of the mis-focus.

 

That simplifies things by not needing to scale the sensor image and, artefacts aside, an EVF showing just the two rectangles (+ framelines) injected into the viewfinder would do the job. LEDs could also do it but to display what's needed, you either need an array of LEDs or a swivelling LED + mask which would take you back to the mechanical system we have now.

 

Only concern is that the entire focus evaluation is being put in the hands of the AF system without any user override.

 

Hi Mark,

 

Yes, that's exactly what I was thinking of. The over-ride, however, could come from using the LCD for a live view (magnified if needed) when and if one did not trust the focus of the AF system.

 

I'm specifically thinking about this for an affordable window finder MFT camera that could use a variety of lenses, including RF lenses. Actual AF could be used with MFT and 4/3 lenses but MF (via the rectangles and live view) could be used with various lenses.

 

Cheers,

 

Sean

[sean_reid]

Bob, interesting point, the camera would have to be able to tell the difference between front and back focus to determine which side to display the "moving" rectangle.

 

Yes it would, do you think that's possible?

 

Cheers,

 

Sean

[Riley]

Advertisement (gone after registration)

just a note in support

with the more sophisticated EXIF readers, E series cameras record the computed AF distance the resulting DoF and the hyperfocal distance

 

it ought be possible to transpose those points on an EVF display, but there is a question as to how. I would suggest an oblique line across the active AF point, with target limits and distances displayed.

[carstenw]

A tiny little tweak in one direction would tell it that. I think that it would need two data points anyway, to tell how far to move the square.

[sean_reid]

I'm not a technology person so I don't know the specifics on how the implementation would work. But this system, for an affordable viewfinder camera, wouldn't use the RF cams at all. It could use MFT and 4/3 AF lenses (with AF) as well as MF lenses (including RF lenses) with MF.

 

It would preserve the bright window view (with frame lines) that distinguish an RF camera and, hopefully, use a focus method that is as quick and useful as we have in the M8.

 

Cheers,

 

Sean

[monza]

The concept sounds good, not knowing the technical challenges behind the engineering.

 

How would such a system work with say a 90 or 135mm lens? How big (or small) would the framelines be in the optical view?

[Bob Ross]

Bob, interesting point, the camera would have to be able to tell the difference between front and back focus to determine which side to display the "moving" rectangle.

Hi Mark,

One simple, but limited, way to do it would be to use an external phase detection unit (PDAF) like the one in the DigiLux2 or even the hybrid unit, if live view was enabled. The external PDAF unit would be limited by focal length range. Let the PDAF unit set the AF range and the lens' cam would be manually focused to match it and signal a match in the window finder. The PDAF/lens cam link unit would know if the lens match was in front or behind its pick. The RF patch in the finder would be the PDAF unit's FOV and the whole frame, of the patch, could light up with a match or only one side to indicate which way to turn the lens focus ring to get a match.

Phase detection technology has come a way since the DigiLux2 (and Nikon Coolpix 8400), so there probably could be greater accuracy. One thing for sure, is that this would be a low cost solution with only one mechanical link between the lens cam and the PDAF unit. Add to this the M-type window finder withe a transparent TFT to show frame lines that would correct for parallax and field size and you would have something that would mimic a classic. Enabling live view would allow macro, tele & even perspective control lenses. We would even be haunted by focus shift and lens calibrations to make us feel right at home

Bob

[peterb]

Greetings Mark,

 

As Howard points out lugging around a D3 or D700 with two or three zoom lenses is just plain cumbersome.

 

 

Hi Terry (and Howard),

 

I couldn't agree more. My only digital purchases have been the Digilux 2 (which is an amazing SILENT camera with a phenomenal lens for a small sensor and an LX-1. (I had the Ricoh GR-D but it got fried on a rather moist trip to the Canadian Rockies.)

 

I'd been eyeing larger sensor DSLRs but ultimately shyed away because when held in my hands, they just seemed soooooo big. And the Sigma DP1, as conceptually nice as it was, was too flawed to put money down. And then you have to deal with the optics which are not any less bulky.

 

I like the M8 (I used to have the M6) but it's just a bit pricey. And like Sean Reid and others have said, for 4 kilobucks, the thing should (1) NOT require a filter over the prize glass you want to use with it and (2) be weather sealed.

 

So, when the G1 appeared on the horizon with it's far less bulk but with the possibility of using M lenses on what appears to be a decent sensor and image processor, I leaped whole hog. And I haven't enjoyed shooting as much as I have in a long time. Even more than with the D2.

 

Peter

[Guest stnami]

post 13 new leica it or not!!!:D

 

The GetDPI Workshop Forums - View Single Post - Different u4/3-LM adapter designs - what differences?

[peterb]

My own idea for the electronic rangefinder is a bit different. There would be one simple rectangle (LED projected maybe) in the center of the frame lines and another that moves side to side but exactly overlaps the first when focus is correct. So there would be no EVF projected - simply two little box outlines, one of which would move in response to the camera evaluating the actual focus relayed by the sensor. In other words, the projected boxes would simply be a system for displaying progressive focus accuracy - simple LEDs or the like.

 

This would be a camera that could use MFT or 4/3 lenses for autofocus and, with adapters, a wide range of lenses for MF - with RF focus patch-esque focusing. The same AF system would inform both AF and the electronic RF.

 

I'm interested in a fairly simple system that is robust and reasonably inexpensive to develop and produce. It would likely rely on AF technologies that already exist but would use a different way of displaying MF focus accuracy. This would be an MFT camera with a window finder that could use a variety of lenses - not only RF lenses.

 

The challenge that occurs to me is having accurate reading of peak focus (at the sensor) when a (non-electronic) lens was stopped down.

 

The camera would not necessarily have an EVF but it would provide live view via the LCD (when needed).

 

 

Hi Sean,

 

I'd been meaning to comment on your idea here and I have to tell you I just love the simplicity of it. Aligning two coincident frames super-imposed onto a window like current rangefinders do. With the sensor as the ultimate arbiter of sharpness as opposed to relying on cams and properly milled lenses. Sensor focus would, as you and others have pointed out, focus shift resulting from whatever.

 

The issue of achieving focus when non-auto aperture lenses are stopped down is certainly something to consider. Especially when that's when contrast focus is at its weakest.

 

I'm not a fan of those red beacon assist lamps that call everyone's attention to your presence.

 

Bob Ross's suggestion of an auxiliary focusing system (external) would certainly help but it just seems to be getting more complicated again. And I don't think that's what you had intended.

 

It's sure hard to top what's worked so simply and so beautifully for so long.

 

Peter

[marknorton]

I have now received my Novoflex adapter and it's great to be able to use my M lenses on the camera. The only one which cannot be used is the 50/2.8 collapsible Elmar because of the danger of damaging the camera when it's collapsed.

 

The longer lenses (and in this context, long is >= 50mm) work very well without the focus assist function and the live view through the Noctilux is extraordinary. The full-frame image would be more so.

 

Focussing when stopped down is easier that with a straight optical TTL viewfinder because the EVF increases the gain but there's no denying that focussing becomes more difficult in low light conditions, but it does too on an M camera if you can't find edges to align against.

 

I agree Peter, the M rangefinder has worked well and it's only now that it's up against limits of accuracy and with new technology that it's worth exploring alternatives.

 

M stands for "Messesucher", or "measuring finder" so there's a question of whether a camera without a rangefinder as such could still be called an M. Maybe, in the way that Leica appear to be producing a new R camera which supports existing R lenses, there's a case for a different camera which doesn't tread on the toes of the M but happens to be able to use its lenses.

[peterb]

Mark,

 

I am so envious of you (and TEBNewYork, Hudson, Vivek and others) who are now free to use the beloved Leitz glass on a G1 (which I agree is seductively quiet and in my opinion let's you roam about pretty much unnoticed as you happily snap away they way Bresson must have done in the last century, only without the need to worry about running out of film so quickly).

 

I look forward to seeing what you see.

 

The question is, will you post your first results here????

 

Or....(dum dum dum dummmmmmm....) that 'other' site.

 

 

 

Peter

[marknorton]

It really helps to have the aperture ring at the front of M lenses so that it's easy to open up the lens for focussing and then stop it down for shooting. You can use Aperture Priority mode which is identical to setting the exposure on an M8 - you see the camera selected shutter speed in the viewfinder as you stop down; alternatively, you can set to Manual mode and do "match needle".

 

The big plus of course is that you get to preview depth of field as you stop down, something that even a DSLR cannot do while retaining finder brightness. As you stop down, the EVF cranks up the gain to compensate for the reduced light coming through the lens; sure, there's noise at f16 but DoF is much easier to evaluate than through a murky optical viewfinder.

 

Look through a Noctilux and marvel as the depth of field changes as you stop down, 75mm Summilux too.

 

Just as a sanity check, I dug out my Digilux 2 and had a look at its EVF. Yes, just as bad as we remember it but the difference between the two shows just how much progress can be made in a few years and, presumably, there can be more to come.

 

Then factor in the cost - G1, adapter, a trio of lenses, say, CV 15, old 28mm Elmarit/35mm Summicron IV and 50mm Summicron/pre-ASPH Summilux. What better way to enjoy photography through Leica glass at an affordable price? Makes the R-D1 look faintly ridiculous now!

[howard_cummer]

Hi Mark,

Welcome to the G1 M adapter users. Another versatile lens for the G1 is the Tri Elmar which is becoming my walk around lens like it is (was) on the M8..

Looking forward to seeing some pictures - especially with the Nocti

Happy New Year

Howard

[Riley]

i had a think about this, and i dont think emulating a RF is doable

 

1/ you cant see more than the sensor sees, you cannot therefore have a view outside of the sensor frame.

 

2/ with CDAF it needs to reach the focus point to compute anything, so it cant give you a superimposed window with any accuracy until its actually focussed, which makes the whole thing kinda superfluous. Since it doesnt have an AF specific sensor as in a dSLR, there are no other options.

[Advertisement]

Hallo,

Look here: Interesting Leica gear

[sean_reid]

It really helps to have the aperture ring at the front of M lenses so that it's easy to open up the lens for focussing and then stop it down for shooting. You can use Aperture Priority mode which is identical to setting the exposure on an M8 - you see the camera selected shutter speed in the viewfinder as you stop down; alternatively, you can set to Manual mode and do "match needle".

 

The big plus of course is that you get to preview depth of field as you stop down, something that even a DSLR cannot do while retaining finder brightness. As you stop down, the EVF cranks up the gain to compensate for the reduced light coming through the lens; sure, there's noise at f16 but DoF is much easier to evaluate than through a murky optical viewfinder.

 

Look through a Noctilux and marvel as the depth of field changes as you stop down, 75mm Summilux too.

 

Just as a sanity check, I dug out my Digilux 2 and had a look at its EVF. Yes, just as bad as we remember it but the difference between the two shows just how much progress can be made in a few years and, presumably, there can be more to come.

 

Then factor in the cost - G1, adapter, a trio of lenses, say, CV 15, old 28mm Elmarit/35mm Summicron IV and 50mm Summicron/pre-ASPH Summilux. What better way to enjoy photography through Leica glass at an affordable price? Makes the R-D1 look faintly ridiculous now!

 

Hi Mark,

 

I agreed with everything you said here until that last line. The R-D1 still offers important things that the G1 does not. I don't think it makes the R-D1 look ridiculous at all but it is a very interesting new camera.

 

Cheers,

 

Sean