Lens evaluation: AF-S Nikkor 58/1.4G

[Bjørn Rørslett]

It's essential to keep the detail of magnification of the original capture identical in order to compare 'depth of field'. I don't think this is the case here from what Sten writes. So the comparison can only tell something about the transitions not the actual depth of field.

[Airy]

Just counting the dots... If the camera was at the same position for both shots, I'd anticipate shallower DOF with the 58, I mean : less focussed dots here. Am I wrong ?

[Bjørn Rørslett]

Stens says one of the images had been rescaled. And in the near range, having the camera in the same position is an incorrect procedure.

Only by adjusting the capture and camera standpoint so as to make the details in the plane of focus exactly the same size, valid comparisons can be made. Put a ruler in the frame and make it equal in size and the comparison can be carried out.

[Thomas G]

Well, both pictures count 25 full black dots on the sharpest row.
I've used the 4th in the sharpest zone counted from the bottom up.
I think sharpness on this row does not perfectly match in the left and right image but I assume it's close enough.

The right image is half of a gap line wider, thus a little (little equals 1/50th row) less magnification.
The difference in mm is 158 (left) to 155 (right), which is as litte as 1.94%.
At 50cm object distance for the 50mm this is 1:9 magnification for the left image and 59cm object distance for the 58mm at 1:9.17 magnification for the right one.
Assuming an acceptable blurr circle of 0.025mm we get DOF for left: 8.994mm and right:9.324mm at f=2.0.
At 58cm object distance for the 58mm the magnificaton would be both 1:9 and the DOF would only vary insignificantly on the 3rd digit.
I used the calulator app at http://www.erik-krause.de/schaerfe.htm.

Despite this calculation I recognize the same phenomenon as  Airy does: the left image printed does show more DOF.

edit: please note reply #17 /edit

[Bjørn Rørslett]

It bears repetition: it is the detail magnification of the capture that is important. NOT the scaled image.

The 'DOF calculators' use simplified equations and aren't really of any help. They cannot prove anything as they are models. Only a visual comparison of captures done with the conditions correct can be of assistance.

[Thomas G]

Sorry Bjørn,
The images shown are misleading to a certain degree, which can be helped.
I think the calculation helps a lot to identify the issue with the printed images.
It also backs your argument of matching equal magnification for theses tests.

The calculator result allows us to conclude that the pictures were shot at about  the same object distance, thus the rescaling.

For optimal presentation the 58mm pic should have been shot at greater distance, ie 59cm if 50cm was used for the 50mm lens,
which I think is as close as it gets for practical use to receive the same magnification.

The images shown are misleading to a certain degree, which can be helped.

That and only that is what I wanted to figure out and show.

[stenrasmussen]

50/1.8G to the left.

[Bjørn Rørslett]

Do note that the relevant conjugate distances (u, v) are measured from the nodal planes. The simplified equations further assume the internodal distance is zero, which is acceptable for captures at distance, but *not* in the near range.

The only acceptable way of getting a meaningful comparison is making the image scale in the focused plane identical. You cannot calculate the required distances as the required parameters for the lens in general are not known.

I urge Sten to make a new set of comparison images with a valid test configuration. Pay also attention to the angle of incidence so as to keep the front- and rear perspective similar as well.

[stenrasmussen]

The two images are shot from the same distance and angle (tripoded). The 58 image has been scaled to match the magnification of the 50 one. It may not be scientifically correct but it does actually show the difference in rendering between the two lenses.

[richardHaw]

i was choosing between the 58 and the 50 ART more than a year ago and decided to get the sigma for the AF since i do intend to use the lens in dimly lit stages and the AF of the 58 just isn't fast enough for what I need it for.  cost is also another thing.

this (58) is a sharp lens wide open to be honest, and the bokeh is superior to the 50 ART. this is saying a lot since the 50 ART is a pretty good 50mm lens.