[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Rollei] Lens charts
This is a multi-part message in MIME format.
- ------=_NextPart_000_004A_01C3158E.64616670
Content-Type: text/plain;
charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
- ----- Original Message -----=20
From: "Dan Kalish" <kaliushkin >
To: <rollei-digest ; =
<rollei
Cc: "Dan Kalish" <kaliushkin >
Sent: Thursday, May 08, 2003 2:30 PM
Subject: [Rollei] Lens charts
>=20
>=20
>=20
> Where would I find a guide, preferably on-line, to understanding lens
> charts? Alternatively, if
> it could be put in a few words, please explain here?
>=20
> I've come across two types of charts:
>=20
> MTF: http://www.butzi.net/rodenstock/apo-sironar-n/180mm.htm
>=20
> Modulation:
> =
http://www.schneideroptics.com/photography/large_format_lenses/apo-symmar=
- -L/pdf/ApoSymmarL_56_180.pdf
>=20
> Thanks,
>=20
> Dan the K.
>=20
>=20
I don't know of an on-line source other than they may be something on =
these web sites.=20
Lets look at the Rodenstock data sheet first because its simpler. There =
are five graphs on this page. Two are modulation transfer curves at two =
different f/stops. One is for geometrical distortion, one for light fall =
off with image angle, and one showing longitudinal chromatic aberration. =
I will take the last one first. Longitudinal chrmomatic aberration =
shows the difference in the focus for different colors. This chart is =
plotted from blue at the left to red at the right. You will note that =
the curve goes through the zero line twice. Actually, the left most zero =
crossing is off the graph but can be extrapolated. Two crossings means =
that the lens is corrected to bring two wavelengths, or colors, of light =
to a common focus. The correction is necessary because glass does not =
bend all colors equally. All glass bends blue light more than red light. =
This results in an aberration called longitudinal chromatic. If not =
corrected the lens will project a rainbow of colors of white objects, =
the size of the fringing will get larger as you move away from the =
center. While this lens is called an APO lens the graph shows it is not =
apochromatic but rather achromatic. An apochromatic lens would cross the =
center line three times, i.e., be in focus for three colors =
simuntaneously.=20
The depth of the curve shows the amount of deviation from focus. This =
is more important than the number of colors brought to common focus. It =
is desirable that this graph be as close to a straight line coincident =
with and parallel to the zero line as possible.=20
The graph showing light fall off shows as a reference a line following =
the cos^4 Theta value. This is the theorectical fall off of light with =
image angle for a _rectilinear_ lens. That is, a lens which reproduces =
an image with correct perspective. Another way of saying this is that it =
is a lens without either barrel or pincussion distortion. It is possible =
to reduce this fall off by one factor by a special design technique but =
its used mainly on very wide angle lenses.=20
The graph shows that the lens has more fall off than the theoretical =
minimum. This is due mainly to partial obscuring of the aperture by the =
lens mount, called vignetting, at larger stops. Note that at the =
smallest stops the fall off follows the theoretical line pretty closely. =
The distortion chart shows geometrical distortion. This lens has =
slight pincussion distortion. Note that the distortion varies with the =
distance of the object. The curves are shown for several values of M or =
magnification of the object.=20
The two MTF curves show the same information but for two values of =
f/stops. Some aberrations vary with the f/stop.
The curves are not as clearly labled as they should be. The curves are =
for the resolution numbers at the top of the chart. They are not =
calibrations of the chart. The graph shows the percentage of fall off of =
resolution for the values of image angle, or coverage size shown on the =
bottom. The curves are for 2.5 line pairs per millimeter, and for 5, 10, =
and 20 lp/mm
The lowest resolution curve is at the top, the greatest at the bottom. =
The graph really shows the contrast of the lines vs: the image angle.
Now, as the stop is made smaller the corrections get better but also =
as the stop is made smaller the loss of resolution due to diffraction =
gets larger. So, we find that the top graph, at f/11, has higher values =
of contrast in the mid section, than the lower graph made at f/22. =
However, the lower graph shows the contrast at greater image angles is =
greater. This is a demonstration of the well known fact that the =
"optimum" stop of a lens becomes smaller with wider image angle. As an =
example, the optimum for a Dagor for use as a "normal" lens, i.e. about =
55 degrees coverage, is around f/22. The optimum stop, or rather, =
necessary stop, for using the lens at its widest coverage, around 87 =
degrees, is f/45. Getting the corners sharp may result in compromising =
the sharpness near the center. These graphs show that.=20
The Schneider PDF data sheet has more complete information but is =
essentially the same. Page one shows the same distortion curve for three =
values of magnification. It also shows the transmittance fo the lens, =
which Rodenstock does not. This graph shows the transmission of the lens =
falling off quite rapidly above visible blue, a chacteristic of many =
high index glasses. It is also rolling off slowly below visible red. The =
curve is pretty flat otherwise. This indicates that the lens probably =
has very little color cast.=20
The MTF curves are for 5, 10, and 20 lp/mm in all cases. They are =
shown for three values of f/stop and for three values of magnification =
image distance.
The f/stops are (from left to right) f/5.6, f/11, f/22
From top to bottom the conjugates are for infinity focus, for a =
magnification of -10 and for -5. The minus sign meaning the images are =
smaller than the object, i.e., reduction even though the general term is =
magnification ratio.=20
The image angle is shown at the bottom as in the Rodenstock graphs.=20
Both sets of MTF charts show both meridonal or radial, and tangential =
values. This is because of the eliptical shape of the aperture away from =
the exact center. The loss of resolution due to diffraction is greatest =
in the direction where the aperture is smallest. So, if we image a =
spider's web we will find that the sharpness of the round parts =
(tangential) are sharper than the spokes (radial or meridonal lines). =
This is true of all lenses although modified in those mentioned earlier =
where distortion of the aperture is introduced to decrease fall off.=20
The hills and valleys of the MTF curves are due to the interaction of =
the various aberrations, some of which vary with the stop and some which =
do not, plus the effect of diffraction. While diffraction alone suggests =
that the resolution is maximum at the center and falls off with angle, =
the complex interaction of aberrations in an actual lens may result in =
the resolution actually going up with angle and then falling off again. =
This is one reason for the peculiar performance of some lenses.=20
I will add that neither manufacturer has any data on lateral chromatic =
aberration. I explained what longidutinal chromatic is above. However, =
correcting it does not insure good images. Its possible for a lens to =
bring the light of all colors of interest to close to the same focus but =
still have fringing. This is because the magnification may vary with =
color. This defect is called lateral color. While the iamge of each =
color is sharp they are of different size. The result is blurry images =
for black and white and unpleasant fringing for color pictures.=20
Symmetrical lenses are automatically corrected for lateral color by =
the symmetry. Even though the cancellation is complete only whent the =
entire system is symmetrical, i.e., image and object distance the same, =
or unity magnification, it is substantially reduced even for infinity =
focus.=20
Symmetrical _type_ lenses meant for use at infinity are often made =
somewhat asymmetrical to improve the cancellation of lateral color and =
also coma and distortion, at infinity.=20
Unsymmetrical lenses, like Tessars, are more difficult to correct, but =
can be. An example is the Kodak Ektar series, which have virtually no =
lateral color.=20
I hope this helps your understanding. It is very far from a "few =
words" but its rather more complicated than it may seem at first glance. =
I also think that you could tell a lot more from a few minutes =
comparing the lenses than these graphs will tell you.=20
=20
- ---
Richard Knoppow
Los Angeles, CA, USA
dickburk
- ------=_NextPart_000_004A_01C3158E.64616670
Content-Type: text/html;
charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META http-equiv=3DContent-Type content=3D"text/html; =
charset=3Diso-8859-1">
<META content=3D"MSHTML 6.00.2800.1170" name=3DGENERATOR>
<STYLE></STYLE>
</HEAD>
<BODY>
<DIV> </DIV>
<DIV>----- Original Message -----=20
<DIV>From: "Dan Kalish" <<A=20
href=3D"mailto:kaliushkin </A>></DIV>
<DIV>To: <<A=20
href=3D"mailto:rollei-digest >rollei-digest@mejac.p=
alo-alto.ca.us</A>>;=20
<<A=20
href=3D"mailto:rollei
</A>></DIV>
<DIV>Cc: "Dan Kalish" <<A=20
href=3D"mailto:kaliushkin </A>>; <<A=20
href=3D"mailto:butzi </A>></DIV>
<DIV>Sent: Thursday, May 08, 2003 2:30 PM</DIV>
<DIV>Subject: [Rollei] Lens charts</DIV></DIV>
<DIV><BR></DIV>
<DIV>> <BR>> <BR>> <BR>> Where would I find a guide,=20
preferably on-line, to understanding lens<BR>> charts? =
Alternatively,=20
if<BR>> it could be put in a few words, please explain =
here?<BR>>=20
<BR>> I've come across two types of charts:<BR>> <BR>> =
MTF:=20
<A=20
href=3D"http://www.butzi.net/rodenstock/apo-sironar-n/180mm.htm";>http://w=
ww.butzi.net/rodenstock/apo-sironar-n/180mm.htm</A><BR>>=20
<BR>> Modulation:<BR>> <A=20
href=3D"http://www.schneideroptics.com/photography/large_format_lenses/ap=
o-symmar-L/pdf/ApoSymmarL_56_180.pdf">http://www.schneideroptics.com/phot=
ography/large_format_lenses/apo-symmar-L/pdf/ApoSymmarL_56_180.pdf</A><BR=
>>=20
<BR>> Thanks,<BR>> <BR>> Dan the K.<BR>> =
<BR>>=20
<BR> I don't know of an on-line source other than they may be =
something on=20
these web sites. </DIV>
<DIV> </DIV>
<DIV>Lets look at the Rodenstock data sheet first because its simpler. =
There are=20
five graphs on this page. Two are modulation transfer curves at two =
different=20
f/stops. One is for geometrical distortion, one for light fall off with =
image=20
angle, and one showing longitudinal chromatic aberration. </DIV>
<DIV> I will take the last one first. Longitudinal chrmomatic =
aberration=20
shows the difference in the focus for different colors. This chart is =
plotted=20
from blue at the left to red at the right. You will note that the curve =
goes=20
through the zero line twice. Actually, the left most zero crossing is =
off the=20
graph but can be extrapolated. Two crossings means that the lens =
is=20
corrected to bring two wavelengths, or colors, of light to a common =
focus. The=20
correction is necessary because glass does not bend all colors equally. =
All=20
glass bends blue light more than red light. This results in an =
aberration called=20
longitudinal chromatic. If not corrected the lens will project a rainbow =
of=20
colors of white objects, the size of the fringing will get larger as you =
move=20
away from the center. While this lens is called an APO lens the graph =
shows it=20
is not apochromatic but rather achromatic. An apochromatic lens would =
cross the=20
center line three times, i.e., be in focus for three colors =
simuntaneously.=20
</DIV>
<DIV> The depth of the curve shows the amount of deviation from =
focus.=20
This is more important than the number of colors brought to common =
focus. It is=20
desirable that this graph be as close to a straight line coincident with =
and=20
parallel to the zero line as possible. </DIV>
<DIV> The graph showing light fall off shows as a reference a line =
following the cos^4 Theta value. This is the theorectical fall off of =
light with=20
image angle for a _rectilinear_ lens. That is, a lens which reproduces =
an image=20
with correct perspective. Another way of saying this is that it is a =
lens=20
without either barrel or pincussion distortion. It is possible to reduce =
this=20
fall off by one factor by a special design technique but its used mainly =
on very=20
wide angle lenses. </DIV>
<DIV> The graph shows that the lens has more fall off than the =
theoretical=20
minimum. This is due mainly to partial obscuring of the aperture by the =
lens=20
mount, called vignetting, at larger stops. Note that at the smallest =
stops the=20
fall off follows the theoretical line pretty closely. </DIV>
<DIV> The distortion chart shows geometrical distortion. This lens =
has=20
slight pincussion distortion. Note that the distortion varies with the =
distance=20
of the object. The curves are shown for several values of M or =
magnification of=20
the object. </DIV>
<DIV> The two MTF curves show the same information but for two =
values of=20
f/stops. Some aberrations vary with the f/stop.</DIV>
<DIV> The curves are not as clearly labled as they should be. The =
curves=20
are for the resolution numbers at the top of the chart. They are not=20
calibrations of the chart. The graph shows the percentage of fall off of =
resolution for the values of image angle, or coverage size shown on the =
bottom.=20
The curves are for 2.5 line pairs per millimeter, and for 5, 10, and 20=20
lp/mm</DIV>
<DIV> The lowest resolution curve is at the top, the greatest at =
the=20
bottom. <BR>The graph really shows the contrast of the lines vs: the =
image=20
angle.</DIV>
<DIV> Now, as the stop is made smaller the corrections get better =
but also=20
as the stop is made smaller the loss of resolution due to diffraction =
gets=20
larger. So, we find that the top graph, at f/11, has higher =
values of=20
contrast in the mid section, than the lower graph made at f/22. However, =
the=20
lower graph shows the contrast at greater image angles is greater. This =
is a=20
demonstration of the well known fact that the "optimum" stop of a lens =
becomes=20
smaller with wider image angle. As an example, the optimum for a Dagor =
for use=20
as a "normal" lens, i.e. about 55 degrees coverage, is around f/22. The =
optimum=20
stop, or rather, necessary stop, for using the lens at its widest =
coverage,=20
around 87 degrees, is f/45. Getting the corners sharp may result in =
compromising=20
the sharpness near the center. These graphs show that. </DIV>
<DIV> </DIV>
<DIV> The Schneider PDF data sheet has more complete information =
but is=20
essentially the same. Page one shows the same distortion curve for three =
values=20
of magnification. It also shows the transmittance fo the lens, which =
Rodenstock=20
does not. This graph shows the transmission of the lens falling off =
quite=20
rapidly above visible blue, a chacteristic of many high index glasses. =
It is=20
also rolling off slowly below visible red. The curve is pretty flat =
otherwise.=20
This indicates that the lens probably has very little color cast. </DIV>
<DIV> The MTF curves are for 5, 10, and 20 lp/mm in all cases. =
They are=20
shown for three values of f/stop and for three values of=20
magnification image distance.</DIV>
<DIV> The f/stops are (from left to right) f/5.6, f/11, f/22</DIV>
<DIV> From top to bottom the conjugates are for infinity focus, =
for a=20
magnification of -10 and for -5. The minus sign meaning the images are =
smaller=20
than the object, i.e., reduction even though the general term is =
magnification=20
ratio. </DIV>
<DIV> The image angle is shown at the bottom as in the Rodenstock =
graphs.=20
</DIV>
<DIV> </DIV>
<DIV> Both sets of MTF charts show both meridonal or =
radial, and=20
tangential values. This is because of the eliptical shape of the =
aperture away=20
from the exact center. The loss of resolution due to diffraction is =
greatest in=20
the direction where the aperture is smallest. So, if we image a spider's =
web we=20
will find that the sharpness of the round parts (tangential) are =
sharper=20
than the spokes (radial or meridonal lines). This is true of all lenses =
although=20
modified in those mentioned earlier where distortion of the aperture is=20
introduced to decrease fall off. </DIV>
<DIV> The hills and valleys of the MTF curves are due to the =
interaction=20
of the various aberrations, some of which vary with the stop and some =
which do=20
not, plus the effect of diffraction. While diffraction alone suggests =
that the=20
resolution is maximum at the center and falls off with angle, the =
complex=20
interaction of aberrations in an actual lens may result in the =
resolution=20
actually going up with angle and then falling off again. This is one =
reason for=20
the peculiar performance of some lenses. </DIV>
<DIV> </DIV>
<DIV> I will add that neither manufacturer has any data on lateral =
chromatic aberration. I explained what longidutinal chromatic is above. =
However,=20
correcting it does not insure good images. Its possible for a lens to =
bring the=20
light of all colors of interest to close to the same focus but still =
have=20
fringing. This is because the magnification may vary with color. This =
defect is=20
called lateral color. While the iamge of each color is sharp they are of =
different size. The result is blurry images for black and white and =
unpleasant=20
fringing for color pictures. </DIV>
<DIV> Symmetrical lenses are automatically corrected for lateral =
color by=20
the symmetry. Even though the cancellation is complete only whent the =
entire=20
system is symmetrical, i.e., image and object distance the same, or =
unity=20
magnification, it is substantially reduced even for infinity focus. =
</DIV>
<DIV> Symmetrical _type_ lenses meant for use at infinity are =
often made=20
somewhat asymmetrical to improve the cancellation of lateral color and =
also coma=20
and distortion, at infinity. </DIV>
<DIV> Unsymmetrical lenses, like Tessars, are more difficult to =
correct,=20
but can be. An example is the Kodak Ektar series, which have virtually =
no=20
lateral color. </DIV>
<DIV> </DIV>
<DIV> I hope this helps your understanding. It is very far from a =
"few=20
words" but its rather more complicated than it may seem at first glance. =
</DIV>
<DIV> I also think that you could tell a lot more from a few =
minutes=20
comparing the lenses than these graphs will tell you. </DIV>
<DIV> <BR>---<BR>Richard Knoppow<BR>Los Angeles, CA, USA<BR><A=20
href=3D"mailto:dickburk </A></DIV></=
BODY></HTML>
- ------=_NextPart_000_004A_01C3158E.64616670--
------------------------------