08-15-2010, 01:03 PM
[quote name='Yakim' timestamp='1281854898' post='1782']
That's what everybody in PN said. That's why I re-tested it. Look at my reply on page #3: "Update: I re-tested and took pictures. You wouldn't be surprised if I told you shutter speeds are the same, right?".
[/quote]
Some remarks on light loss at magnification ratio 1:1.
With the old macro lens design types close focusing was accomplished by increasing the extension through the built-in helical up to 1:2 and through additional extension tubes between 1:2 and 1:1.
Modern macro lenses which provide continuous focusing up to a magnification ratio of 1:1 without the need of an additional extension tube are in fact zoom lenses because their focal length decreases from infinity towards minimum focus distance. In the older (non-VR) AF-Micro-Nikkor 105/2.8 the lens group behind the fixed back element moves backwards(!) when focusing from infinity to 1:2. But from 1:2 on to 1:1 it reverses its movement.
At 1:1 the focus distance (distance between a focused object and its image) equals four times the nominal focal length. Thus, we can calculate an estimated nominal focal length at 1:1 by dividing the minimal focus distance by four:
minimal focus distance @1:1 - estimated nominal focal length @1:1 - example lenses
350 mm - 87.5 - Sony 100/2.8
310 mm - 77.5 - Canon 100/2.8, Nikon 105/2.8, Nikon 105/2.8 VR, Pentax 100/2.8
300 mm - 75.0 - Canon 100/2.8 IS
286 mm - 71.5 - Nikon 85/3.5
So the loss of 1.5 f-stops that you observed with your Canon 100/2.8 IS corresponds well with a decrease of the focal length to 75% of the focal length @infinity and therefore corresponds with a gain in relative aperture compared to an expected loss of 2 f-stops @1:1 with a traditionally designed macro lens.
My AF Micro-Nikkor 105/2.8 has a calculated aperture diameter of 37.5 mm (= 105 mm / 2.8). Given an effective focal length @1:1 of 155 mm (= twice the estimated nominal focal length @1:1), this lens has a calculated aperture of f/4.1 (= 155 mm / 37.5 mm). But the camera indicates and measures an effective value of f/5.
So this shows that this calculation of the aperture is an over-simplification and that something else is at play. Most likely this is caused by a change of the effective exit pupil while focusing towards 1:1. The role of the pupil magnification is mentioned at http://www.cambridgeincolour.com/tutoria...lenses.htm
Considering the rather extreme light loss of 3.5 f-stops with the Mamiya 120/4: Did you observe this also with an MF body? Maybe the adapter and/or the relative small hight and width of the mirror chamber of the 7D hinder(s) peripheral light from the back lens element to hit the sensor.
Back in the days of slide projectors with mediocre projector lenses I "stepped-down" the projector lens by fitting an aperture ring cut from cardboard directly to the back lens element. This induced vignetting but it also improved the sharpness of the projected image.
HTH, Ralf.
That's what everybody in PN said. That's why I re-tested it. Look at my reply on page #3: "Update: I re-tested and took pictures. You wouldn't be surprised if I told you shutter speeds are the same, right?".
[/quote]
Some remarks on light loss at magnification ratio 1:1.
With the old macro lens design types close focusing was accomplished by increasing the extension through the built-in helical up to 1:2 and through additional extension tubes between 1:2 and 1:1.
Modern macro lenses which provide continuous focusing up to a magnification ratio of 1:1 without the need of an additional extension tube are in fact zoom lenses because their focal length decreases from infinity towards minimum focus distance. In the older (non-VR) AF-Micro-Nikkor 105/2.8 the lens group behind the fixed back element moves backwards(!) when focusing from infinity to 1:2. But from 1:2 on to 1:1 it reverses its movement.
At 1:1 the focus distance (distance between a focused object and its image) equals four times the nominal focal length. Thus, we can calculate an estimated nominal focal length at 1:1 by dividing the minimal focus distance by four:
minimal focus distance @1:1 - estimated nominal focal length @1:1 - example lenses
350 mm - 87.5 - Sony 100/2.8
310 mm - 77.5 - Canon 100/2.8, Nikon 105/2.8, Nikon 105/2.8 VR, Pentax 100/2.8
300 mm - 75.0 - Canon 100/2.8 IS
286 mm - 71.5 - Nikon 85/3.5
So the loss of 1.5 f-stops that you observed with your Canon 100/2.8 IS corresponds well with a decrease of the focal length to 75% of the focal length @infinity and therefore corresponds with a gain in relative aperture compared to an expected loss of 2 f-stops @1:1 with a traditionally designed macro lens.
My AF Micro-Nikkor 105/2.8 has a calculated aperture diameter of 37.5 mm (= 105 mm / 2.8). Given an effective focal length @1:1 of 155 mm (= twice the estimated nominal focal length @1:1), this lens has a calculated aperture of f/4.1 (= 155 mm / 37.5 mm). But the camera indicates and measures an effective value of f/5.
So this shows that this calculation of the aperture is an over-simplification and that something else is at play. Most likely this is caused by a change of the effective exit pupil while focusing towards 1:1. The role of the pupil magnification is mentioned at http://www.cambridgeincolour.com/tutoria...lenses.htm
Considering the rather extreme light loss of 3.5 f-stops with the Mamiya 120/4: Did you observe this also with an MF body? Maybe the adapter and/or the relative small hight and width of the mirror chamber of the 7D hinder(s) peripheral light from the back lens element to hit the sensor.
Back in the days of slide projectors with mediocre projector lenses I "stepped-down" the projector lens by fitting an aperture ring cut from cardboard directly to the back lens element. This induced vignetting but it also improved the sharpness of the projected image.
HTH, Ralf.