Great Old Performers (and M8)

[sergiolov]

It seems that very old lenses are still capable of great performance with the m8. This is merit of the original designers, but also of this little sensor with displaced microlenses.
The images posted were taken with a 1954 Zeiss biogon 21 4,5 for the postwar Contax (and left to me by my father many years ago)
-# 1 is the protagonist, taken with d2x
-# 2 is full image, to evaluate vignetting
-# 3 is 100% crop left lower corner for resolution
-# 4 is 100% crop center for resolution
-# 5 is full image, to have an idea of (very small) distortion and color
Not a scientific test, but a modest tribute by me to a veteran.

I am sure there are many other and possibly older great performers around.
Sergio

[marknorton]

Interesting shots, but isn't that the great thing about the M8, able to being to life old lenses? Try doing that on a dSLR something or other...

[robsteve]

Originally Posted by marknorton

Interesting shots, but isn't that the great thing about the M8, able to being to life old lenses? Try doing that on a dSLR something or other...

I think the M8 is going to bring new life to old photographers too

[sergiolov]

Hi Robert and Mark,
I completely agree, expecially on the OLD PHOTOGRAPHERS part, LOL.
The fact is, I am unable to part from this camera,notwistanding the moirè problem that is practically precluding me to use it for my main job. (but I'll find a solution..)
I am doing something I have not done for many many years, keep a camera in my pocket and go around just to take photos, as I did in the sixties.

That said, I don't know what a new asph etc can do better, substantially better, than this old lens with the m8. Probably only now, due to the camera, it is showing all its possibilities.

Sergio

[lct]

Could you show vignetting at f/4.5 with the M8?

[sergiolov]

Originally Posted by lct

Could you show vignetting at f/4.5 with the M8?

Here it is, full open, probably slightly faster than 4,5 as blades open a little more.
Sergio

[scott kirkpatrick]

Originally Posted by sergiolov

Here it is, full open, probably slightly faster than 4,5 as blades open a little more.
Sergio

How does it look in color (no IR filter needed for this, there is still some red vignetting from the internal IR filter)? Just shoot a nice white wall in indirect daylight.

scott

[scott kirkpatrick]

Originally Posted by sergiolov

Here it is, full open, probably slightly faster than 4,5 as blades open a little more.
Sergio

here's what Imatest ( Imatest - image quality evaluation software )'s Light Falloff action shows for the luminance channel vignetting in the frame above. It's impressive -- less than one stop at the worst corner, and some of that is due to the red vignetting. If you can send the original without turning it into black and white, I can separate the two effects.



The "lens focal length" reported is an estimate of exit pupil height, used by Picture Window Pro in its vignette correction tool.

scott

[sergiolov]

Hi Scott,
Norman is a good friend of mine, we occasionally exchange info and I appreciate his experienced and qualified opinion.
That was a BW jpg, the shot was made with mixed tungsten and fluorescent lighting.
It was underexposed, so the only change I did was to increase luminance- this should not have changed the ratios.
To day I'll perform the color test in suitable light.
Sergio

[scott kirkpatrick]

Originally Posted by sergiolov

Hi Scott,
Norman is a good friend of mine, we occasionally exchange info and I appreciate his experienced and qualified opinion.
That was a BW jpg, the shot was made with mixed tungsten and fluorescent lighting.
It was underexposed, so the only change I did was to increase luminance- this should not have changed the ratios.
To day I'll perform the color test in suitable light.
Sergio

great, I'll dig it out and run it, even if it does sink below the fold. I also enjoy talking to Norman. He gave us a frontispiece for a book recently, a shot of peaks and valleys which he took in Monument Valley.

Incidentally, I looked at the noise background (one of the steps available in Light Falloff) and I think I can see the vertical line at .58 of the width in your smooth shot.

scott

[sergiolov]

Conversion with CS2, contrast in neutral position.Shadows at 0
First at 4,5 second at 8

[lct]

Originally Posted by sergiolov

Here it is, full open..

Thank you Sergio.
Great lens indeed.
What lens adapter do you use on the M8?

[sergiolov]

Originally Posted by lct

Thank you Sergio.
Great lens indeed.
What lens adapter do you use on the M8?

You wellcome.
The adapter was made by a repair shop in the sixties to use it on m2, but if you take a look on ebay you'll find that it is actually produced and offered.
eBay Italia: Contax to Leica Adapter for Zeiss & Nikon RF Lens M ZI (oggetto 190065691226 scade il 01-Gen-07 02:56:37 CET)
Sergio

[scott kirkpatrick]

Originally Posted by sergiolov

Conversion with CS2, contrast in neutral position.Shadows at 0
First at 4,5 second at 8

Thanks for making these tests available. I think this is easily correctible in the M8 firmware, once we see how the Leica 21/2.8 behaves, and mark the mount of your lens to impersonate the coded Leica 21. (Reid Reviews should have that information soon, but perhaps not until the next firmware is released.)

Here are analyses of these two vignetting tests. First the vignetting of the Blue and Green channels, translated into light intensities by assuming a gamma of 2.2. It shows a modest aperture dependence, and about one stop of vignetting.



and then the additional effect of the red vignetting, obtained by plotting Red intensity over the average of the Blue and Green intensities. This is smaller, and does not depend much on aperture, hence should be not too hard to correct in firmware.



Imatest doesn't let us separate the two types of vignetting, so this is my own Matlab program. I average the pixels of your 900x600 plots in 3x2 groups, leaving a square array of averaged pixels, and then plot the diagonal values, averaging the diagonal starting from upper left to lower right with the diagonal starting from upper right to lower left. The left side of each figure is the upper two corners of the test shot, and the right side is the lower two corners. The numbers I get are pretty consistent with the contour plot that Imatest produces for both effects together.

Speaking of great old performers, i have a Canon 19/3.5 rangefinder lens from the 60's which I will try out once my M8 gets to me in another week or two. It sits very deep in the camera body, so i am not sure how well it will do, but it was fun to use with film.

scott

[rvaubel]

Scott
What I find interesting about all these old lenses is how well they perform on the M8 despite their theoritical design disadvantage. It seems that many of these traditional symmetrical designs with rear elements colse to the focal plane, are performing just as well, if not better, than the newer "digital" retrofocus designs.
Anybody else notice this?

Rex

[lct]

Will be interesting to compare it to the new Biogon 21/4.5 as well.

[Woody Campbell]

Originally Posted by sergiolov

It seems that very old lenses are still capable of great performance with the m8. This is merit of the original designers, but also of this little sensor with displaced microlenses.
. . . .
I am sure there are many other and possibly older great performers around.
Sergio

It would be great to start a sticky thread for samples with vintage lenses. It would be nice to see the results from some exotics like Angenieux, Ross, etc., as well as vintage Leica and CZ glass.

[rvaubel]

Originally Posted by lct

Will be interesting to compare it to the new Biogon 21/4.5 as well.

True. But the new Biogon at least makes some consessions to the older design. I would also like to see a comparision with the Zeiss 21mm F2.8, a more "modern" retrofocus telecentric design.
Is "modern"=Telecentric=Better on the M8 or has the offset microlensing changed the equation?

Rex

[sergiolov]

Originally Posted by rvaubel

True. But the new Biogon at least makes some consessions to the older design. I would also like to see a comparision with the Zeiss 21mm F2.8, a more "modern" retrofocus telecentric design.
Is "modern"=Telecentric=Better on the M8 or has the offset microlensing changed the equation?

Rex

The rear lens of the old biogon is only 9mm from the focal plane, so I was especting to see a terrible amount of vignetting. Offset microlensing for sure changed the equation, if there is'nt something else not disclosed that helps.
As a side note, I also expected light metering blockage, but, on the contrary, auto exposure is correct.

Sergio

[rvaubel]

Originally Posted by sergiolov

The rear lens of the old biogon is only 9mm from the focal plane, so I was especting to see a terrible amount of vignetting. Offset microlensing for sure changed the equation, if there is'nt something else not disclosed that helps.
As a side note, I also expected light metering blockage, but, on the contrary, auto exposure is correct.

Sergio

Yes, I was expecting problems from all short back focus lenses. Although I love my RD1, she really had a problem with them. Not so the M8. All those VC super wides seem to work great. Now it appears that many of the old school wides will also work, even better than they did on film cameras. This opens up a whole new horizon for us lens nuts

Rex