A superfast normal on the G1

[petermcwerner]

Thus the 26mm f/1.1 Macro-Switar, being an RX-mount lens, is a very poor choice of "superfast normal" for the G1 camera (unless you plan to mount a 9.5mm thick pane of glass behind the lens to compensate its RX-mount aberrations)

It is only necessary to collimate the 26mm/1.1 for the G1. After that, it works very well on the G1. My experience. More details in the following thread:

http://forum.getdpi.com/forum/showthread.php?t=9593

Cheers
Peter

 

[Ranger 9]

Thanking Ranger9 for getting the matter of the Bolex RX-mount lenses mostly correct.

You're mostly welcome It's not easy to be exact when you have to Google up references from all over the map!

In general, this strand's fascination with using C-mount lenses on the G1 is optically perverse.

Some of it may be perversity and some of it may just be frustration at the current lack of compact, wide-aperture alternatives in these focal-length ranges. When the Micro Four Thirds-native Panasonic 20/1.7 finally appears, I'll probably retire my C-mounts except for special purposes.

Still, since there ARE so many C-mount optical perverts here, though, and since you seem to be the person who has done the deepest thinking on the RX issue... maybe you'd like to write up a summary that could stay in the forum, so people don't have to keep finding my post and then chasing down all the links...?

 

[Vivek]

Still, since there ARE so many C-mount optical perverts here, though, and since you seem to be the person who has done the deepest thinking on the RX issue... maybe you'd like to write up a summary that could stay in the forum, so people don't have to keep finding my post and then chasing down all the links...?

That suggestion is great but would only be helpful if dcouzin actually has any user experience of these lenses, especially for photography.

 

[Vivek]

A typical example of an "optical perversion".:ROTFL::ROTFL:

Panasonic G1, Computar-TV 25/1.3.

 

[woodmancy]

Thanking Ranger9 for getting the matter of the Bolex RX-mount lenses mostly correct.

In general, this strand's fascination with using C-mount lenses on the G1 is optically perverse.

For better or for worse I have a Ph.D in Physics - I never made my living from it, but that's my problem.
I schooled at Reading University and the Physics department was headed up by R W Ditchburn who was an expert on light and optics, and wrote a number of seminal works. I couldn't understand a word he talked about, even though I was heavy into photography - such a pity. I've picked up much more about the practical issues of lenses in these forums.

Onwards and upwards - Vivek has taught me more than Ditchburn. But I'm not going to buy Vivek's recommended camera case :deadhorse:

Keith

 

[dcouzin]

Optically perverse. Love that.:ROTFL::ROTFL:

Do you know of any lens that does not drop sharpness (and other corrections) from the center, that is made for still cameras?

Sure, this is often the case for narrow angle lenses, like 135mm lenses for 35mm still format (43mm diagonal).
The matter has nothing to do with still versus cine. Anyone designing a "normal" angle lens can see that field coverage cannot come without compromising the center. Different manufacturers have different "philosophies" concerning this tradeoff. The Zeiss Distagon I cited is designed to "cover" a 12.6mm circle. Wide open, sharpness is actually greatest at about 5mm off center. Sharpness declines rapidly past about 5.5mm off center. So only the very corners of the intended image are noticeably less sharp than the center. Zeiss knows what they're doing. My point was that by 8mm off center, which is as far as they published data for this lens, the sharpness is gone.

 

[dcouzin]

Some of it may be perversity and some of it may just be frustration at the current lack of compact, wide-aperture alternatives in these focal-length ranges.

You have a 21.6mm diagonal sensor and you want a certain angle of view corresponding to about 25mm focal length. This is a wide-angle lens and you won't find solutions among lenses designed as normal-angle lenses for 16mm cinema (image diagonal 12.6mm) or super 16mm cinema (image diagonal 14.5mm).

Your image diagonal is considerably smaller than that of 35mm cinema (about 27.2mm) and smaller yet than 35mm "half-frame" still (about 30mm). You would find these lenses sharp over your whole sensor. Not as sharp as lenses designed for your format, but sharp enough. But they're expensive.

It is a tough practical problem to have an image sensor with no suitable photographic lenses. Narrow angle lenses can be home-made but not wide-angle lenses. A possible solution is to mount a cheap 50mm lens with adequate coverage and then to stick a cheap 0.5× wide angle adapter in front of that.

 

[dcouzin]

It is only necessary to collimate the 26mm/1.1 for the G1. After that, it works very well on the G1. My experience.

Post #91 (Photomorgana) in this strand shows two pictures made with the 26mm f/1.1 Macro Switar RX lens. The one shot at f/1.1 is grossly unsharp. Look at the lower eyelid hairs. The one shot at f/1.8 looks pretty sharp (in the low resolution JPEG posted). Are your results better?

It is conceivable that Kern made some 26mm f/1.1 Macro Switars which were not RX-type. These would be true C-mount lenses and not require the 9.5mm glass pane between them and the image. If they exist, and if you luckily found one, then your experience won't match Photomorgana's or the other less lucky f/1.1 users.

And how sharp is your image 8mm, 9mm, 10mm off the center?

 

[dcouzin]

... maybe you'd like to write up a summary that could stay in the forum, so people don't have to keep finding my post and then chasing down all the links...?

My last article on RX and C was in 1987.
The short answer today is: avoid RX-mount lenses. They are lenses designed for a 9.5mm pane of BK7 glass between them and the image. Such a pane introduces spherical aberration, astigmatism, and chromatic aberrations. The RX lens is designed with a load of aberrations to cancel those of the pane, so the RX lens without the pane is an incomplete optic. Why pay big money for an incomplete optic unless you happen to own a camera, namely the Bolex H16 REX, which has the completing 9.5mm thick pane in front of the image plane?
Similarly, for large NA (like small f-number) miicroscope objectives you generally need a different lens when there will be a cover glass over the specimen.
Similarly lenses for 3-sensor video cameras are designed with aberrations to cancel those of the camera's prism systems. You would not be happy using one of these lenses on a 1-sensor camera.

I lived near the American Science Center in Chicago in its heyday when it got the surplus of Bell & Howell. The risk in buying mystery lenses cheap is that you don't know whether they are optically complete. (You also don't know what wavelengths, what image surface, what magnification, etc. they're intended for.) The enthusiasts in this room have enough uncertainty about their C-mount lenses. They don't need the addition of the RX problem.

 

[Ranger 9]

All too true. Fortunately, a lot of these lenses are cheap enough that we can just buy them and try them out to see if we like the pictorial effects they produce.

I don't always care if my pictures are sharp, as long as they're pretty -- although I realize that as a professional cinematographer you can't afford to take such a cavalier attitude!

 

[Vivek]

My last article on RX and C was in 1987.

The short answer today is: avoid RX-mount lenses. They are lenses designed for a 9.5mm pane of BK7 glass between them and the image. Such a pane introduces spherical aberration, astigmatism, and chromatic aberrations.

Correct. However, you are still in the film era and are forgetting the digital situation.

The construction of the sensor (based on a 4/3rds Olympus E-410) is like this:

1. ~1mm glass permanently fixed on the sensor which is on a flex board.

2. ~7-8mm thick glass sandwich of AA/UV/IR cut filters (about 4 or 5 layers of various materials).

3. Dust shake "cover" glass of ~ 1mm thickness.

Any amount of glass material between the sensor and the lens is going to induce aberrations. See Dr. Brian Caldwell's post here: http://photo.net/leica-rangefinders-forum/00M5ag

(Brian is a lens designer by profession)

So, I am afraid your reasoning about RX lenses (and other optical perversions) are grossly misplaced. The amount of glass that Olympus and Panasonic load on their sensors and the issues that ensue can not be beaten by any other glass.

 

[petermcwerner]

It is conceivable that Kern made some 26mm f/1.1 Macro Switars which were not RX-type. These would be true C-mount lenses and not require the 9.5mm glass pane between them and the image.

Another example of an "optical perversion", shot through the glass pane of the shop window. Uncropped, 4 to 3 format

Yverdon - Macro Switar 26mm/1.1 RX on G1 - 1/60 @ f/5.6 - Raw Therapee
​

 

[petermcwerner]

..

 

[petermcwerner]

More examples of "optical perversion", again 4 to 3 uncropped
​

Yverdon Bike Shop - Macro Switar 26mm/1.1 RX on G1 - ISO=100 - 1/125 @ f/5.6 - Raw Therapee

Macro Switar 26mm/1.1 - ISO=100 - 1/160 @ f/2.8 - Raw Therapee

​

 

[Vivek]

"Optical perverts" are everywhere.:ROTFL:

Josep, in Barcelona (Spain) photographs with classic Nikkors mounted on an Olympus E-410.

Panasonic G1, Olympus 17/2.8 plus a Ricoh GW-1 (0.75X converter).

He was pleased to see my Olympus (real) Pen F 40/1.4 and gave me his email to correspond.

 

[woodmancy]

Someone needs to start an optical perversion thread. Looks like we're getting a lot of perverted images.
May want to use my G1 Holga thread as a start
I think we also need an in depth analysis of the Holga lens. Shouldn't take long - it only has one element, or does it? (have to read up on that) How can anything so perverted have such a following?
Has anyone made an adapter for a Holga lens to a G1 or EP-1 (maybe I'm onto something here, I've already butchered my Pentax 110, why not my Holga?)
Sorry
Keith

 

[m3photo]

Re: Optical Perverts Anonymous

"Optical perverts" are everywhere.:ROTFL:

Josep, in Barcelona (Spain) photographs with classic Nikkors mounted on an Olympus E-410.

Hi, my name is Michael ... and I'm an Optical Pervert. Hello Michael

I can't help myself, it's been two days since I took these and, well, I'm starting to get the craves and I might slip back and put another Nikkor or Pentax 110 lens on my G1 ...
The last two were with a 105mm f/2.5 wide open and the first with a 70mm Pentax 110 lens also wide open, this latter one doesn't even have a diaphragm and always shoots at f/2.8 ...

 

[dcouzin]

I don't always care if my pictures are sharp, as long as they're pretty -- although I realize that as a professional cinematographer you can't afford to take such a cavalier attitude!

Hey, I'm no professional cinematographer. I'm fine with pinhole lenses, cokebottle lenses, etc. I love the images from my Gemini 300 kilopixel camera. I tape old (positive) eyeglass lenses on for focusing. I find misunderstood $1000 lenses in well-machined adapters on a 12 megapixel camera to lack a certain "sportiveness".

For photography's first 100 years or so lenses had much character (from aberrations and ghosts) and lent their character to the pictures. For about the next 40 years many photographers and cinematographer scrambled to simulate lost lens character with lens front paraphernalia. Then for a while straight-and-clear was in full dominance. Then came digital photography. Among else, image quality modification in Photoshop (and like programs) became easy. This of course is after the original photography and it's often not the photographer doing it.

It will take more than retro dabbling to shake the new digital dominion. If you believe that lens and light are crucial to photography, then you have to prove this with lenses and light. The picture given to Photoshop is 2-dimensional although the scene given to the camera is 3-dimensional (really 4-dimensional but we're not discussing motion and shutters here). The lens alone transforms the three dimensions into two. The limited depth of field and the lens's "defects" of astigmatism and field curvature can't be simulated after the picture is taken (unlike most other aberrations, distortion, vignetting, etc.). Also the picture given to Photoshop has its dynamic range much limited compared to the scene. The bright sun can send fabulous ghosts through an old uncoated lens which the merely white disk in the image cannot be coaxed to produce. These are some aspects of optical character left to lens-and-light-lovers.

 

[woodmancy]

Re: Optical Perverts Anonymous

Hi, my name is Michael ... and I'm an Optical Pervert. Hello Michael

I can't help myself, it's been two days since I took these and, well, I'm starting to get the craves and I might slip back and put another Nikkor or Pentax 110 lens on my G1 ...
The last two were with a 105mm f/2.5 wide open and the first with a 70mm Pentax 110 lens also wide open, this latter one doesn't even have a diaphragm and always shoots at f/2.8 ...

Hey m3, do you want to tell us how you created the 110 adapter - I'm half way through mine.

Keith

 

[dcouzin]

Correct. However, you are still in the film era and are forgetting the digital situation.

The construction of the sensor (based on a 4/3rds Olympus E-410) is like this:
1. ~1mm glass permanently fixed on the sensor which is on a flex board.
2. ~7-8mm thick glass sandwich of AA/UV/IR cut filters (about 4 or 5 layers of various materials).
3. Dust shake "cover" glass of ~ 1mm thickness.
...
So, I am afraid your reasoning about RX lenses (and other optical perversions) are grossly misplaced. The amount of glass that Olympus and Panasonic load on their sensors and the issues that ensue can not be beaten by any other glass.

Your information that there is approximately 9 to 10 mm of flat glass between the lens and the G1 sensor is completely new to me. That would make the optical situation practically the same as in the Bolex H16 RX camera. Then of course a RX type lens is needed for the G1 camera.
It's surprising information, but I'll assume you or someone else has measured it. The old RX/C rule is still very applicable to the interests of this strand. Just the conclusions will be different based on the new information. 1987 RX/C Rule:

A C-mount lens works well on a RX camera or a RX-mount lens works well on a C camera, if and only if:

(1) the lens is slower than about f/2 or f/2.8, or stopped down this far;​

AND (2) the lens has a deep set exit pupil, about 1½ inches or farther into its screw mount.​

[Note the second clause in the rule is based on the 12.6mm image diagonal of 16mm. The exit pupil distance would need to be increased for a 21.6mm diagonal sensor.]

This is not really about C and RX mount lenses and cameras. It's about lenses designed to image directly onto the sensor versus those designed to image onto sensors covered by 9.5mm of BK7 glass, and about cameras with uncovered sensors and those with sensors covered by 9.5mm of BK7 glass. It is not terribly important that the thickness be exactly 9.5mm or the glass be exactly BK7. (The glasses you describe are generally like BK7.) The location of the 9.5mm pane, nearer or farther from the sensor, is immaterial.

The point remains that you can't have it both ways. If at full aperture the 26mm f/1.1 Macro-Switar RX lens works well on a camera (because the camera is effectively RX type) then at full aperture the popular and common 25mm f/0.95 Angenieux does not work well on that camera. Likewise the Zeiss, Canon, Kowa, Nikon and most all superfast normals don't work well at full aperture on that camera.