Jump to content

M (240) calibration


Recommended Posts

  • Replies 47
  • Created
  • Last Reply
Am I the only one who can't link to the movie? (" Requested URL ... not available on this server")

 

Terribly sorry for the inconvenience. I have placed a second copy on another server. Let me know if this works for you also. I was not aware of limitations in access to my site from the US.

 

http://129.125.165.58/~egbertotten/LeicaForum/L1009413four.MP4

Link to post
Share on other sites

Terribly sorry for the inconvenience. I have placed a second copy on another server. Let me know if this works for you also. I was not aware of limitations in access to my site from the US.

 

http://129.125.165.58/~egbertotten/LeicaForum/L1009413four.MP4

 

Thank you for popping the cover plate. Nice sequel to the M4 movie; with the higher viewpoint you can now see the two parts...

 

thanks much

Link to post
Share on other sites

...

Both the offset and gain errors give an increased error in focus with growing object distance, also when expressed as percentage of that distance. When infinity is set correctly (even when there is a gain error), the percentage drops when going towards infinity...

 

...last sentence is correct?

 

thanks

Link to post
Share on other sites

Amazing. I'm falling in love with the RF all over again. I just can't get enough of this opto-mechanical stuff. Really.

 

Love it too, Rick. There is something organic about the combination of lenses, moving parts and precision that comes from curved shapes rather than brute numerical processes.

 

Organic, since the human body is also using those principles. Look at the shape of joint surfaces, the changing shapes of human eye-lenses when focussing and gradually changing lever arms of muscles creating movements in your fingertips.

 

Oops, I think I'm getting carried away...

Link to post
Share on other sites

Yes: rangefindererrorpercentagecompensatedinfinity.png

 

|% error| is decreasing I guess ... what is the (uncompensated) function you are using? Looks like bgain1 = bcenter + (b - bcenter) X bgain + boffset. What is boffset after you "compensate infinity"?

 

thanks

Link to post
Share on other sites

|% error| is decreasing I guess ... what is the (uncompensated) function you are using? Looks like bgain1 = bcenter + (b - bcenter) X bgain + boffset. What is boffset after you "compensate infinity"?

 

thanks

 

boffset is chosen such that at infinity (in this case 1000 meter) the error in gain is compensated. This is what you do when you only adjust the roller for a target at infinity to get correct overlap while you have a small error in rangefinder arm length.

Link to post
Share on other sites

This graph is not linear, but comes directly from the simulations of the rangefinder mechanism for which I already offered the equations and the software on this forum.

http://www.l-camera-forum.com/leica-forum/leica-m9-forum/147259-focussing-issues-3.html#post1509238

 

I was asking about your notebook, not the graph; that's your notation.

 

boffset is chosen such that at infinity (in this case 1000 meter) the error in gain is compensated...

 

lol, now I see what you did. Use mathematical infinity, not a finite distance like 1000, to get an increasing but bounded |% error|.

Link to post
Share on other sites

Yes: rangefindererrorpercentagecompensatedinfinity.png

 

eh...I still think the graph is problematic.

 

I don't think it's physically correct. "1000m" is just picked out of a hat. How do you know that 100 or 50 is not the appropriate number?

 

I don't think it's mathematically correct. If it wasn't conveniently (!) cut off at 1000m, you could see the |% error| growing larger and larger after that point.

 

Not sure that my comment was any improvement.

Link to post
Share on other sites

eh...I still think the graph is problematic.

 

I don't think it's physically correct. "1000m" is just picked out of a hat. How do you know that 100 or 50 is not the appropriate number?

 

I don't think it's mathematically correct. If it wasn't conveniently (!) cut off at 1000m, you could see the |% error| growing larger and larger after that point.

 

Not sure that my comment was any improvement.

 

1000 meter is chosen to be the practical infinity for a 50 mm lens. If you want to redefine practical infinity to for instance 10000 meter, I can insert that and redraw the graph. The result is similar.

 

What I can not do is calculate to true infinity: that is too far away :D (any small percentage of infinity is infinitely large)

Link to post
Share on other sites

Here are the graphs for practical infinity is 10000 meter for a 50 mm lens.

 

rangefindererror10000me.png

 

rangefindererrorpercentage10000me.png

 

The difference between the focal length and image distance is now only 0.0002 mm, which is less than any Leica rangefinder mechanism can consistently match in accuracy, so quite a meaningless improvement over the 1000 meter I took earlier.

 

Yet the principle is still demonstrated on the importance of adjusting the roller of the mechanism.

Link to post
Share on other sites

1000 meter is chosen to be the practical infinity for a 50 mm lens. If you want to redefine practical infinity to for instance 10000 meter, I can insert that and redraw the graph. The result is similar.

 

...

 

no you have to draw the two graphs running out to 10,000m with the boffset for 1,000m. Every time you draw the graph with the distance chopped off at the distance you compensated for, it looks like the error and percent error are 0 going out to infinity...especially when you tell us that the error is "compensated at infinity" and the "percentage error drops going out to infinity". But as long as you say what you're doing, it's no matter.

 

What I can not do is calculate to true infinity: that is too far away :D (any small percentage of infinity is infinitely large)

 

You and Zeno. Actually, if you can get the percentage error for any arbitrary finite distance, you are all set.

Link to post
Share on other sites

We tend to think in accuracy of focussing on objects in object space (the real world) and object space runs from very close to infinity. However, for the image (the most important result), image space is more interesting and that runs between focal length and about two times focal length (giving 1:1 macro). Any error in image space (defect of focus) of the focal point of an object has an effect on sharpness (modulation as a function of spatial frequency) of the image. So it is more interesting to look at image space.

 

And in image space the problem of infinity does not occur. You can just express the focus error as a percentage of the focus image distance. Again, the same principle holds: if you adjust for "infinity", the error in rangefinder arm length results in a very limited error in focus for objects more nearby. So adjustment of the rangefinder roller for focussing on an object at "infinity" (say 1000 meter for a 50 mm lens) is the most important adjustment.

Link to post
Share on other sites

We tend to think in accuracy of focussing on objects in object space (the real world) and object space runs from very close to infinity. However, for the image (the most important result), image space is more interesting and that runs between focal length and about two times focal length (giving 1:1 macro). Any error in image space (defect of focus) of the focal point of an object has an effect on sharpness (modulation as a function of spatial frequency) of the image. So it is more interesting to look at image space.

 

And in image space the problem of infinity does not occur. You can just express the focus error as a percentage of the focus image distance. Again, the same principle holds: if you adjust for "infinity", the error in rangefinder arm length results in a very limited error in focus for objects more nearby. So adjustment of the rangefinder roller for focussing on an object at "infinity" (say 1000 meter for a 50 mm lens) is the most important adjustment.

 

You would need to go to the image space to get the effect of a change in arm length or offset. You test the focus for a DIY adjustment on the other side, so it's natural to express results there. It is easy enough to go back and forth. The problem of infinity does not go away in image space.

Link to post
Share on other sites

If you calculate what happens between 10000 meter in object space and anything further, approaching an infinite distance, the focus error due to an error gain (rangefinder arm length error), compensated at 10000 meter with the rangefinder roller, expressed as a percentage of the distance to the camera in object space, climbs from 0% upward and approaches asymptotically 100% towards infinity. Now that looks like a real problem for taking sharp images you may say. But in image space the error climbs from 0 mm to 0.0002 mm and so you could never detect the error in the image. All other tolerances of the rangefinder mechanism and the optical tolerances, and resolution of the sensor are all much worse.

 

So if we step into the real world of a rangefinder photographer for a moment :), the problem of infinity does not exist.

Link to post
Share on other sites

Archived

This topic is now archived and is closed to further replies.

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...