7mm on M3

In 2007 I bought an Olympus camera with a 7-14 mm f/4, because that was the widest wide angle at that moment. The camera is old and has done many clicks (about 200000 I guess). It could die every day? To keep the lens in use I've now put it on an M3 via an adapter. The adapter has no electric contacts, focus and aperture do not work. Zooming does, it's mechanical.

But there is a trick: put the lens on the original camera, change aperture and/or focus distance, remove the lens whilst pressing the check DOF button. Put it on the Canon. I used this to test different apertures and different focus distances to find out at which settings the widest usable field is attained.

But first I've tested if it was a real 7mm or not. Roger Cicala said that often zooms do not reach their extremes as noted on the lens. A test with Olympus led to 7.0 mm and a test with Canon gave 6.95. So it will be 7 mm indeed. (I'ld have to know the nodal points to be able to test it more precise.)

With buildings with many straight lines and bricks and so on the best combination turned out to be f/10 and not focused at infinity. Closing the diaphragm worked good, but focusing at shorter than infinity had very much effect on image quality near the borders. Unexpected. Could it be that the system isn't optimised for infinity, but for a distance usable as hyperfocal distance? f/10 by the way is really small.

The sensor of a Canon is bigger than fourthirds sensors. I did not expect the full width of a Canon picture would have good quality, but it has. Horizontal angle is 116 degrees, which is huge. Barrel distortion of the lens is considerable: 5%. Edges of the picture are black caused by the mount of the front lens, this cannot be avoided. Correcting for 5% distortion leads to less black edges. Edges are bad anyway. That's picture 1 (jpeg out of camera, no corrections, sharpening or whatever). Due to correction of distortion the angle of view now is 115 degrees. The sensor size of fourthirds is indicated.

Cropping to 114 degrees and 1.8 : 1 dimensions gives picture 2. Only the edge is clearly worse than the rest of the picture, Picture 3 is the edge. The useful area is much bigger than the image circle needed for Olympus.

nnowak wrote:

Weia wrote:

In 2007 I bought an Olympus camera with a 7-14 mm f/4, because that was the widest wide angle at that moment. The camera is old and has done many clicks (about 200000 I guess). It could die every day? To keep the lens in use I've now put it on an M3 via an adapter. The adapter has no electric contacts, focus and aperture do not work. Zooming does, it's mechanical.

But there is a trick: put the lens on the original camera, change aperture and/or focus distance, remove the lens whilst pressing the check DOF button. Put it on the Canon. I used this to test different apertures and different focus distances to find out at which settings the widest usable field is attained.

But first I've tested if it was a real 7mm or not. Roger Cicala said that often zooms do not reach their extremes as noted on the lens. A test with Olympus led to 7.0 mm and a test with Canon gave 6.95. So it will be 7 mm indeed. (I'ld have to know the nodal points to be able to test it more precise.)

With buildings with many straight lines and bricks and so on the best combination turned out to be f/10 and not focused at infinity. Closing the diaphragm worked good, but focusing at shorter than infinity had very much effect on image quality near the borders. Unexpected. Could it be that the system isn't optimised for infinity, but for a distance usable as hyperfocal distance? f/10 by the way is really small.

The sensor of a Canon is bigger than fourthirds sensors. I did not expect the full width of a Canon picture would have good quality, but it has. Horizontal angle is 116 degrees, which is huge. Barrel distortion of the lens is considerable: 5%. Edges of the picture are black caused by the mount of the front lens, this cannot be avoided. Correcting for 5% distortion leads to less black edges. Edges are bad anyway. That's picture 1 (jpeg out of camera, no corrections, sharpening or whatever). Due to correction of distortion the angle of view now is 115 degrees. The sensor size of fourthirds is indicated.

Cropping to 114 degrees and 1.8 : 1 dimensions gives picture 2. Only the edge is clearly worse than the rest of the picture, Picture 3 is the edge. The useful area is much bigger than the image circle needed for Olympus.

Can you post a picture of your setup? The issues you are having with infinity focus may be due to the thickness of the adapter. The adapter should be no thicker than 1.25mm.

I get the impression he's talking of the FourThirds version rather than M43 ??

https://www.four-thirds.org/en/fourthirds/wide.html

Those had a somewhat longer flange distance!

Peter

Can you post a picture of your setup? The issues you are having with infinity focus may be due to the thickness of the adapter. The adapter should be no thicker than 1.25mm.

I get the impression he's talking of the FourThirds version rather than M43 ??

https://www.four-thirds.org/en/fourthirds/wide.html

Those had a somewhat longer flange distance!

Peter

Yes, she (=me) spoke of four-thirds not micro-four-thirds. And I don't have an issue with infinity, but found it remarkable that focusing (internal) had so much effect on image quality for objects being not close at all.

Weia wrote:

Can you post a picture of your setup? The issues you are having with infinity focus may be due to the thickness of the adapter. The adapter should be no thicker than 1.25mm.

I get the impression he's talking of the FourThirds version rather than M43 ??

https://www.four-thirds.org/en/fourthirds/wide.html

Those had a somewhat longer flange distance!

Peter

Yes, she (=me) spoke of four-thirds not micro-four-thirds. And I don't have an issue with infinity, but found it remarkable that focusing (internal) had so much effect on image quality for objects being not close at all.

Oops - my apologies!

Peter

Add a little (thin) extension tube and see if that improves the lens’ utility.  The corners will improve but you might lose infinity if it’s too thick.  Otherwise you’ll just have to keep on cropping.  

Best of luck!

R2

That's an interesing topic, I'd like to see a picture where the line of sight is as horizontal as possible or, if you prefer, the sensor is as vertical as possible to have an idea of the distortion if any

Oops - my apologies!

Peter

Accepted. Have a nice day!

Weia wrote:

Oops - my apologies!

Peter

Accepted. Have a nice day!

Thank you - and no, the irony of today's special status has not passed me by!

Peter