Re: [Leica] MTF comparison

["dominique pellissier" <noct@club-internet.fr>]

- ----- Original Message -----
From: Erwin Puts <imxputs@knoware.nl>
To: <leica-users@mejac.palo-alto.ca.us>
Sent: Saturday, February 19, 2000 12:29 PM
Subject: [Leica] MTF comparison


> Dominique wrote in part
> >But, if we compare MTF tests published by Leica, between the 35 asph and
the
> >tri-elmar at a focal length of  35 mm, and both the 2 lenses at 5.6, we
> >observe that the 35 is clearly the winner.
>
> Well I have the Leica graphs in front of me and I wonder how this
> inference can be drawn. Look at this:
>
> 5lp/mm:identical for the Tri-E at 35mm focal length  and the asph 35
> from center to image height 21.6mm
> 10lp/mm: on axis the TriE has slightly better contrast, at the 21.6
> position the asph is slightly better
> 20lp/mm: the Tri-E is better than or equal to the asph from axis till
> image height 15mm, after that the asph is better.
> 40lp/mm: from axis till 6mm image height: tri-E wins, from 6mm to
> 12mm the asph has an advantage, from 12mm to 15 they are equal, after
> that the asph is better. But the area from 6 to 12mm where the asph
> scores a bit better there is much more astigmatism to be seen so it
> is not an unqualified :better".
>
>  From these figures it is impossible to conclude that the "35 is
> clearly the winner". Only if you give 100% of your personal score to
> the unimportant area at image height 20mm then the 35 wins. At more
> relevant image zones from axis till 15 to 18mm, the Tri-E is as good
> as or even better.
> The MTF comparison will tell you that both lenses at 5.6 deliver
> outstanding image quality with of course some character differences.
>
> But if someone just wants to prove the Tri-E to be not as good he
> will interpret the facts with this view in mind. As some scholar
> noted ages ago: if a person wants to distort reality, no amount of
> facts will stop him.
>
>
> On the 2 versus one aspherical surfaces. It were Zeiss engineers who
> told me that, not Leica engineers.
> On the image quality: the differences between the aspherical  and
> ASPH are so small as to be insignificant. There are some marginal
> differences (a shade higher contrast on axis for the first version,
> but better contrast in the field for the second version) but I wonder
> if any Leica user would be able to demonstrate this in even the most
> demanding of test situations.
>
> Erwin
>

#########
We partially agree at 5.6, and for 40lp/mm !
I prefer not to observe the charts for 5, 10 and 20 lp/mm because you wrote
in LF n° 1/2000 (i translate from the french) : "We considered that 40
lp/mm(...) was a significant value of the good quality of an image".
I wrote "partially" because, on the axis the tri- gives nearly 72  % and the
summicron 68 %. The difference is lowest. But if we go far from the axis 0,
we see that the contrast of the tri- in saggital structures is constant up
to 6 mm, then it diminishes continually.At 15 mm we have 58 % (saggital) and
49 %(tangential) only. For the summicron we see a growing chart with some
astigmatism (as you observed) but, at 15 mm from the axis, we have 58 %
(saggital) and 85 % (tangential).
Look : 49 % v.85 % !.
And the MTF-test made by Leica is very favourable to the tri- ! At 5.6 the
tri is at its best. If we could compare at f4 the tri- and the summicron,
the victory would be total for the summicron.

So the reality is not distorted. And the tests made by CDI give same
results.

You wrote :"if a person wants to distort reality, no amount of  facts will
stop him."
The argument can be reversed. If someone wants to prove that the tri- is
better than (or at least equal to) the summicron 35, it's easy as a cornish
pie : take the figures for 5 or 10 lp/mm. (BTW same story for the noctilux).
In fact we are facing an ordinary situation in statistics : how to extract
true facts from raw data ?