Interpreting Hasselblad MTF Charts

[larry_menzin]

I was just looking at data sheets for Hasselblad lenses on their web

site and was wondering how to interpret their MTF charts. The image

size varies from 0 to 40mm on the X axis. Should I assume that this

is magnification and the 40mm value is approaching 1:1 since the 6x6

frame is 56x56mm? Most of the worst drop-off in MTF is from 30mm

image size to 40mm, meaning magnification from about 1:2 to 2:3. Is

my interpretation correct?

[Ed_Ingold]

The X-axis represents the distance measured from the center of the image. The Y-axis is the contrast of the image of a standard test target. The higher the contrast, the better the resolution. Each curve represents the results of a test pattern of a certain resolution: 10, 20, 30 and 40 lp/mm.

 

The semi-diagonal distance of a true 6x6 cm image is nearly 4.4 3 cm. The image quality generally falls as you go off-axis of the lens, which is demonstrated by the falling MTF curves.

 

Extraordinary lenses, like the 100/3.5, have MTF curves that are nearly flat, which means the lens is sharp from corner-to-corner. With this lens, the curves are fairly closely spaced down to 40 lp/mm. If you plot the Y-value at a given distance from the center against the lp frequency, you get a curve which is used to interpolate the ultimate resolution, usually in the center of the image.

[nick_wilson2]

Larry

 

I am not sure that you do have it quite right about magnification, as the X axis is the distance from the centre to the corners. The distance on the X axis shows, as stated above, the performance from the centre (0 on the x axis) to the corners. Typically a lens, especially a wide angle, falls away from the centre in optical performance. There are 3 sets of pairs of lines on Hasselblad/Zeiss's charts. These are the 3 sets of test images using different frequencies of repeating black and white lines. It is confusing they are not marked on the graph, but the top one is the 10 cycles/mm test down to the 40 cycles/mm.

The '10 cycles/mm', for example, refers to a test image of black and white lines repeating 10 times per millimetre. The test using it and the other two frquencies are repeated from the centre to the corners. The two lines in the pair refer to the fact that the tests are done twice at each point, with the test black and white lines running from the centre to the corners (sagittal) and across it (tangential).

 

The coarse (10 cycles/mm) test is an indicator of the contrast of the lens. The fine test (40 cyles/mm) is an indicator of the fine resolving power of the lens (ie ability to render fine detail). Together they make the subjective thing we call sharpness.

 

Basically, you want the lines to be as high as possible and as flat as possible across the chart (meaning from the centre of the lens to the corners). The Y axis is the T (or transfer) function that is the percentage 'success' of the lens in accurately transmitting the light information it receives. The ideal is 1.0 (or 100%).

 

A comparison of the 80 and the 100 might thus show contrast as very similarly high and good in both at f8, but falling away a more to the corners in the 80. Perhaps the 100 can resolve fine detail a fraction better in the centre, but more so in the corners.

 

This accords with users' opinions on which is the better lens. The flat field performance of the 100 is exceptional, based on the charts.

 

Most Zeiss Hasselblad lenses score very well! Also, some charts clearly show the influence of stopping down on image quality.

 

There are a number of explanations on the Web, eg the article by Klaus Schroiff on photozone.de which gives an opinion on what is a good result. There are several on http://www.photodo.com/nav/artindex.html.