Loading ...
Sorry, an error occurred while loading the content.
 

Re: Objective NA and "Modulation-Transfer-Function" Performance?

Expand Messages
  • John
    Hi Merv. In the telescope world mtf seems to be generally accepted as gospel but when I see a generalised plot like this one I m inclined to think stinking
    Message 1 of 14 , Jun 1, 2012
      Hi Merv. In the telescope world mtf seems to be generally accepted as gospel but when I see a generalised plot like this one I'm inclined to think stinking fish

      http://groups.yahoo.com/group/Microscope/photos/album/266762016/pic/1392650841/view

      You can see negative contrast and also that wave front error is introduced by defocusing. In my view it is fairly accurate given moderate wave front errors but gets worse as they increase. There is a lot on the subject on various web sites concerned with digital photography. :-) There is also a tiny bit of info on the subject of too many pixels.

      In my view they are of no use to photographers other than to show general trends and talking about 1/4 wave camera optics is a bit daft. What aperture, focus distance and worse still zoom? I have seen comments that Carl Zeiss himself had an interest in this general subject - mtf50 is usually mentioned. What this probably means is that 50% contrast was regarded as a sensible resolution limit for camera optics which corresponds to around 0.4 of what would be expected from it's na complicated by changes in performance as they are stopped down and distances are varied etc. No mention on the amount of fuzz etc at that resolution only contrast. It seems that this is acceptable due to the nature of the human eye. I would throw in size of print and viewing distance etc.

      As far as microscopy is concerned I also feel that mtf curves do have a use. They indicate what can be expected but having used 1/10 wave, optics, the old and in my view correct metric for precision visual optics, I have serious doubts as to all microscope objectives from any manufacturer even meeting a 1/4 wave specification. These days 1/4 wave optics are often said to be diffraction limited but in real terms they aren't. This is why hubble for instance went for 1/30 wave - it's well past the 1/15 wave that early observers reported as being detectable.

      In these areas I think people need to realise that light behaviour is being modelled rather than actually measured. Same applies to electronic filters too really. There are even oddities in that area. Oblique illumination is odd. I have seen an equation where 2 terms are cancelled which effectively doubles resolution. I have also seen a graph posted on a group by Rene (i think) showing a distorted "mtf curve" where contrast is boosted well down the resolution scale. Which is correct? Probably neither but oblique does improve the view one way or another on many subjects and the noticeable effect is a boost in contrast on many of them even a section. Contrast directly relates to resolution so anything that boosts it helps. That is also why things are stained. MTF curves model the contrast and resolution relationship rather well - that really is their main purpose in life. Strehl ratios are a better metric for resolution but even those have a hole in them when viewed against the classic reasons the diffraction patterns.

      John

      --- In Microscope@yahoogroups.com, "merv" <merv.hobden@...> wrote:
      >
      >
      >
      >
      >
      >
      > Hi John and Rene,
      >
      > MTF is a measure that originated in the early days of video systems as electronic engineers tried to correlate the idea of electronic bandwidth with spatial bandwidth in a lens system. They knew from Nyquist, that the minimum bandwidth required to fully represent the information in the signal was twice the maximum frequency component in that signal, and it seemed reasonable to assume that the same applied to spatial frequencies resolved by a lens. However, the idea is simplistic in that it assumes that the system is linear, with a square edged characteristic rather like a perfect low-pass filter. This is never the case with a real lens system, or for that matter, practical low pass filters! In reality, the aberrations in a lens system produce zonal phase aberration which distorts the MTF, giving zones of apparent better resolution for particular spatial frequencies. In the the summer edition of the 2012 Quekett Journal is an excellent article by John Fletcher on that very problem as found in the immersion objectives made by Robert Tolles. He shows that some of Tolles claims of better resolution depended on the existance of an annular area at the periphery of the lens with particular sensitivity to higher spatial frequencies under oblique illumination. This confirms something I have often observed with oil immersions - that an objective which is nothing special with central illumination, will give sometimes startling performance with oblique or annular illumination, and the only reason I could find is that the corrections were markedly different from that of an objective that worked well with central illumination. John contends that F.H. Wenham has been badly treated, as Wenham insisted that his tests showed that the aberrations in the Tolles objectives were large at the periphery of the aperture, and John's tests and simulations show that is indeed the case. So simplistic notions of image forming, based on numerical aperture, or MTF, can only give us a very limited perception of what is really occuring - what we call an image is a result of the convolution of all of the image components with the complex zonal transfer function of the lens. Zonal interference of periodic elements in the image structure may give higher contrast in some parts of the image, this, as John has shown, is purely fortuitous. It is fascinating that a battle which raged in the 1870s, is still, to a lesser extent,being waged today - for all the progress of the last 140 years, we are all still suckers for the simplistic mathematical explanation.
      >
      > Cheers,
      >
      > Merv
      >
      > --- In Microscope@yahoogroups.com, "John" <a.johnw@> wrote:
      > >
      > > Hello Rene. Maybe I should have added a little on the axis of an mtf graph. Vertical is contrast. The horizontal axis can be viewed as being resolution of target / resolution of optic. Contrast out of the optic for 100% contrast in reaches zero at 2NA / the wavelength of light being used. This is why the charts are the same for all optics as the horizontal axis normalises the curve.
      > >
      > > The resolution limit usually used in respect to microscope objectives is by Rayleigh and represents a contrast of some where around 7 1/2 % for a perfect optic and more or less the same for an optic that reaches his criterion which is less than 1/4 wave wave front error. That is the reason for this figure of merit. It's sensibly perfect for stars and many other purposes. In practice were contrast is low and variations in it are also low it's generally reckoned that wave front error of 1/15 wave are detectable by eye. Some have claimed 1/30 wave.
      > >
      > > John
      > >
      > > --- In Microscope@yahoogroups.com, "rene" <renevanwezel@> wrote:
      > > >
      > > > this is an old discussion. Reversing the question akes it easier to understand: if you put a reducing lens in the system, making the 100x planapo a 63x one, would it give the same quality image as the 63x?
      > > >
      > > > Pretty much so, in the centre. Not plano though, to correct the 100x for that huge increased FOV would need considerable correction. That will give, in practice, reduced contrast. How much depends largely on the date of manufacturing.
      > > >
      > > > Best wishes, René
      > > >
      > > >
      > > > --- In Microscope@yahoogroups.com, "John" <a.johnw@> wrote:
      > > > >
      > > > > The MTF is a ratio in respect to the contrast an optic produces against the resolution it's trying to image so all optics have the same shape of mtf curve. Only what this means in terms of actual resolution varies. It's diffraction spot intensity/energy orientated.
      > > > >
      > > > > There's no point in getting too carried away with it. As most optics are not generally often used to image variably spaced black and white lines it more of an indication of the limitations of imaging optics in general. eg MTF50 is often mention is general photography - the point where black and white contrast is reduced to 50% but even the usefulness of that is dubious as pictures generally do not contain a lot of 100% contrast detail in them. What ever contrast is there will be reduced to 50% of it's real value. It seems that MTF50 has been used to rate quality photographic optics. The mtf curve up to that point is fairly good but goes to pot afterwards. Pass on how that is done because in some ways it makes no sense at all. :-) It's a case of well it's stated some where on the web so it must be true mustn't it?
      > > > >
      > > > > John
      > > > >
      > > > > --- In Microscope@yahoogroups.com, "rich_guy_looker" <rich_guy_looker@> wrote:
      > > > > >
      > > > > > "MTF" stands for "Modulation-Transfer-Function"
      > > > > >
      > > > > > If you have a 100X PlanApo,
      > > > > > with a 1.3 NA, with a 10X eyepiece
      > > > > > for 1000X magnification,
      > > > > > and....
      > > > > >
      > > > > > A 63X PlanApo, with a 1.4 NA,
      > > > > > with a 15X eyepiece for a
      > > > > > magnification of 945X,..
      > > > > >
      > > > > > Do both have approximately the same
      > > > > > performance for the "Modulation-Transfer-Function"?
      > > > > >
      > > > > > thanks.
      > > > > >
      > > > > > Rich
      > > > > >
      > > > >
      > > >
      > >
      >
    • Randall Buck
      Hi Merv, Thank you for the excellent insights. Even so, it would be nice to be able to sift through a set of objectives, using some sort of experimentally
      Message 2 of 14 , Jun 1, 2012
        Hi Merv,

        Thank you for the excellent insights.
        Even so, it would be nice to be able to sift through a set of objectives,
        using some sort of experimentally repeatable test that was a bit less
        subjective than my current method of mentally comparing each one via a
        favorite slide.

        If not the pinhole or MTF tests, is there anything you might recommend?

        Randall

        -----Original Message-----
        From: Microscope@yahoogroups.com [mailto:Microscope@yahoogroups.com]On
        Behalf Of merv
        Sent: Thursday, May 31, 2012 12:22 PM
        To: Microscope@yahoogroups.com
        Subject: [Microscope] Re: Objective NA and
        "Modulation-Transfer-Function" Performance?







        Hi John and Rene,

        MTF is a measure that originated in the early days of video systems as
        electronic engineers tried to correlate the idea of electronic bandwidth
        with spatial bandwidth in a lens system. They knew from Nyquist, that the
        minimum bandwidth required to fully represent the information in the signal
        was twice the maximum frequency component in that signal, and it seemed
        reasonable to assume that the same applied to spatial frequencies resolved
        by a lens. However, the idea is simplistic in that it assumes that the
        system is linear, with a square edged characteristic rather like a perfect
        low-pass filter. This is never the case with a real lens system, or for that
        matter, practical low pass filters! In reality, the aberrations in a lens
        system produce zonal phase aberration which distorts the MTF, giving zones
        of apparent better resolution for particular spatial frequencies. In the the
        summer edition of the 2012 Quekett Journal is an excellent article by John
        Fletcher on that very problem as found in the immersion objectives made by
        Robert Tolles. He shows that some of Tolles claims of better resolution
        depended on the existance of an annular area at the periphery of the lens
        with particular sensitivity to higher spatial frequencies under oblique
        illumination. This confirms something I have often observed with oil
        immersions - that an objective which is nothing special with central
        illumination, will give sometimes startling performance with oblique or
        annular illumination, and the only reason I could find is that the
        corrections were markedly different from that of an objective that worked
        well with central illumination. John contends that F.H. Wenham has been
        badly treated, as Wenham insisted that his tests showed that the aberrations
        in the Tolles objectives were large at the periphery of the aperture, and
        John's tests and simulations show that is indeed the case. So simplistic
        notions of image forming, based on numerical aperture, or MTF, can only give
        us a very limited perception of what is really occuring - what we call an
        image is a result of the convolution of all of the image components with the
        complex zonal transfer function of the lens. Zonal interference of periodic
        elements in the image structure may give higher contrast in some parts of
        the image, this, as John has shown, is purely fortuitous. It is fascinating
        that a battle which raged in the 1870s, is still, to a lesser extent,being
        waged today - for all the progress of the last 140 years, we are all still
        suckers for the simplistic mathematical explanation.

        Cheers,

        Merv

        --- In Microscope@yahoogroups.com, "John" <a.johnw@...> wrote:
        >
        > Hello Rene. Maybe I should have added a little on the axis of an mtf
        graph. Vertical is contrast. The horizontal axis can be viewed as being
        resolution of target / resolution of optic. Contrast out of the optic for
        100% contrast in reaches zero at 2NA / the wavelength of light being used.
        This is why the charts are the same for all optics as the horizontal axis
        normalises the curve.
        >
        > The resolution limit usually used in respect to microscope objectives is
        by Rayleigh and represents a contrast of some where around 7 1/2 % for a
        perfect optic and more or less the same for an optic that reaches his
        criterion which is less than 1/4 wave wave front error. That is the reason
        for this figure of merit. It's sensibly perfect for stars and many other
        purposes. In practice were contrast is low and variations in it are also low
        it's generally reckoned that wave front error of 1/15 wave are detectable by
        eye. Some have claimed 1/30 wave.
        >
        > John
        >
        > --- In Microscope@yahoogroups.com, "rene" <renevanwezel@> wrote:
        > >
        > > this is an old discussion. Reversing the question akes it easier to
        understand: if you put a reducing lens in the system, making the 100x
        planapo a 63x one, would it give the same quality image as the 63x?
        > >
        > > Pretty much so, in the centre. Not plano though, to correct the 100x for
        that huge increased FOV would need considerable correction. That will give,
        in practice, reduced contrast. How much depends largely on the date of
        manufacturing.
        > >
        > > Best wishes, René
        > >
        > >
        > > --- In Microscope@yahoogroups.com, "John" <a.johnw@> wrote:
        > > >
        > > > The MTF is a ratio in respect to the contrast an optic produces
        against the resolution it's trying to image so all optics have the same
        shape of mtf curve. Only what this means in terms of actual resolution
        varies. It's diffraction spot intensity/energy orientated.
        > > >
        > > > There's no point in getting too carried away with it. As most optics
        are not generally often used to image variably spaced black and white lines
        it more of an indication of the limitations of imaging optics in general. eg
        MTF50 is often mention is general photography - the point where black and
        white contrast is reduced to 50% but even the usefulness of that is dubious
        as pictures generally do not contain a lot of 100% contrast detail in them.
        What ever contrast is there will be reduced to 50% of it's real value. It
        seems that MTF50 has been used to rate quality photographic optics. The mtf
        curve up to that point is fairly good but goes to pot afterwards. Pass on
        how that is done because in some ways it makes no sense at all. :-) It's a
        case of well it's stated some where on the web so it must be true mustn't
        it?
        > > >
        > > > John
        > > >
        > > > --- In Microscope@yahoogroups.com, "rich_guy_looker"
        <rich_guy_looker@> wrote:
        > > > >
        > > > > "MTF" stands for "Modulation-Transfer-Function"
        > > > >
        > > > > If you have a 100X PlanApo,
        > > > > with a 1.3 NA, with a 10X eyepiece
        > > > > for 1000X magnification,
        > > > > and....
        > > > >
        > > > > A 63X PlanApo, with a 1.4 NA,
        > > > > with a 15X eyepiece for a
        > > > > magnification of 945X,..
        > > > >
        > > > > Do both have approximately the same
        > > > > performance for the "Modulation-Transfer-Function"?
        > > > >
        > > > > thanks.
        > > > >
        > > > > Rich
        > > > >
        > > >
        > >
        >




        ------------------------------------

        Yahoo! Groups Links
      • merv
        Hi Randall, Apart from a Twyman-Green interferometer (Ha ha!) there is no singular test strategy which gives all the relevant information.I came to the
        Message 3 of 14 , Jun 3, 2012
          Hi Randall,

          Apart from a Twyman-Green interferometer (Ha ha!) there is no singular test strategy which gives all the relevant information.I came to the conclusion some years ago that the major differences in objective performance were down to aberrations, irrespective of the NA. Neither the pinhole test (aka Francon) or the Abbe test platte can give detailed information on the extent of zonal aberrations.

          Only use of an interferometer, with a spread of wavelengths can uncover the true performance, and it is a tedious process. If you know the element curvatures, refractive indices and dispersions, you can do what Dr Fletcher did - put it through a ray trace programme, but even then it is difficult to state with complete accuracy. So I take my hat off to the 19th century pioneers - they were able to assemble very well corrected objectives with limited tests. Wenham was very well aware of the importance of zonal corrections, and he had some fairly simple, but sophisticated techniques for sorting out what was going on. One thing that seems to be essential, is the ability to think in 3D, with six degrees of freedom!I think both J.J. Lister and F.H Wenham had that ability, which no amount of algebra can replace easily!

          Cheers,

          Merv
          --- In Microscope@yahoogroups.com, "Randall Buck" <rbuck@...> wrote:
          >
          > Hi Merv,
          >
          > Thank you for the excellent insights.
          > Even so, it would be nice to be able to sift through a set of objectives,
          > using some sort of experimentally repeatable test that was a bit less
          > subjective than my current method of mentally comparing each one via a
          > favorite slide.
          >
          > If not the pinhole or MTF tests, is there anything you might recommend?
          >
          > Randall
          >
          > -----Original Message-----
          > From: Microscope@yahoogroups.com [mailto:Microscope@yahoogroups.com]On
          > Behalf Of merv
          > Sent: Thursday, May 31, 2012 12:22 PM
          > To: Microscope@yahoogroups.com
          > Subject: [Microscope] Re: Objective NA and
          > "Modulation-Transfer-Function" Performance?
          >
          >
          >
          >
          >
          >
          >
          > Hi John and Rene,
          >
          > MTF is a measure that originated in the early days of video systems as
          > electronic engineers tried to correlate the idea of electronic bandwidth
          > with spatial bandwidth in a lens system. They knew from Nyquist, that the
          > minimum bandwidth required to fully represent the information in the signal
          > was twice the maximum frequency component in that signal, and it seemed
          > reasonable to assume that the same applied to spatial frequencies resolved
          > by a lens. However, the idea is simplistic in that it assumes that the
          > system is linear, with a square edged characteristic rather like a perfect
          > low-pass filter. This is never the case with a real lens system, or for that
          > matter, practical low pass filters! In reality, the aberrations in a lens
          > system produce zonal phase aberration which distorts the MTF, giving zones
          > of apparent better resolution for particular spatial frequencies. In the the
          > summer edition of the 2012 Quekett Journal is an excellent article by John
          > Fletcher on that very problem as found in the immersion objectives made by
          > Robert Tolles. He shows that some of Tolles claims of better resolution
          > depended on the existance of an annular area at the periphery of the lens
          > with particular sensitivity to higher spatial frequencies under oblique
          > illumination. This confirms something I have often observed with oil
          > immersions - that an objective which is nothing special with central
          > illumination, will give sometimes startling performance with oblique or
          > annular illumination, and the only reason I could find is that the
          > corrections were markedly different from that of an objective that worked
          > well with central illumination. John contends that F.H. Wenham has been
          > badly treated, as Wenham insisted that his tests showed that the aberrations
          > in the Tolles objectives were large at the periphery of the aperture, and
          > John's tests and simulations show that is indeed the case. So simplistic
          > notions of image forming, based on numerical aperture, or MTF, can only give
          > us a very limited perception of what is really occuring - what we call an
          > image is a result of the convolution of all of the image components with the
          > complex zonal transfer function of the lens. Zonal interference of periodic
          > elements in the image structure may give higher contrast in some parts of
          > the image, this, as John has shown, is purely fortuitous. It is fascinating
          > that a battle which raged in the 1870s, is still, to a lesser extent,being
          > waged today - for all the progress of the last 140 years, we are all still
          > suckers for the simplistic mathematical explanation.
          >
          > Cheers,
          >
          > Merv
          >
          > --- In Microscope@yahoogroups.com, "John" <a.johnw@> wrote:
          > >
          > > Hello Rene. Maybe I should have added a little on the axis of an mtf
          > graph. Vertical is contrast. The horizontal axis can be viewed as being
          > resolution of target / resolution of optic. Contrast out of the optic for
          > 100% contrast in reaches zero at 2NA / the wavelength of light being used.
          > This is why the charts are the same for all optics as the horizontal axis
          > normalises the curve.
          > >
          > > The resolution limit usually used in respect to microscope objectives is
          > by Rayleigh and represents a contrast of some where around 7 1/2 % for a
          > perfect optic and more or less the same for an optic that reaches his
          > criterion which is less than 1/4 wave wave front error. That is the reason
          > for this figure of merit. It's sensibly perfect for stars and many other
          > purposes. In practice were contrast is low and variations in it are also low
          > it's generally reckoned that wave front error of 1/15 wave are detectable by
          > eye. Some have claimed 1/30 wave.
          > >
          > > John
          > >
          > > --- In Microscope@yahoogroups.com, "rene" <renevanwezel@> wrote:
          > > >
          > > > this is an old discussion. Reversing the question akes it easier to
          > understand: if you put a reducing lens in the system, making the 100x
          > planapo a 63x one, would it give the same quality image as the 63x?
          > > >
          > > > Pretty much so, in the centre. Not plano though, to correct the 100x for
          > that huge increased FOV would need considerable correction. That will give,
          > in practice, reduced contrast. How much depends largely on the date of
          > manufacturing.
          > > >
          > > > Best wishes, René
          > > >
          > > >
          > > > --- In Microscope@yahoogroups.com, "John" <a.johnw@> wrote:
          > > > >
          > > > > The MTF is a ratio in respect to the contrast an optic produces
          > against the resolution it's trying to image so all optics have the same
          > shape of mtf curve. Only what this means in terms of actual resolution
          > varies. It's diffraction spot intensity/energy orientated.
          > > > >
          > > > > There's no point in getting too carried away with it. As most optics
          > are not generally often used to image variably spaced black and white lines
          > it more of an indication of the limitations of imaging optics in general. eg
          > MTF50 is often mention is general photography - the point where black and
          > white contrast is reduced to 50% but even the usefulness of that is dubious
          > as pictures generally do not contain a lot of 100% contrast detail in them.
          > What ever contrast is there will be reduced to 50% of it's real value. It
          > seems that MTF50 has been used to rate quality photographic optics. The mtf
          > curve up to that point is fairly good but goes to pot afterwards. Pass on
          > how that is done because in some ways it makes no sense at all. :-) It's a
          > case of well it's stated some where on the web so it must be true mustn't
          > it?
          > > > >
          > > > > John
          > > > >
          > > > > --- In Microscope@yahoogroups.com, "rich_guy_looker"
          > <rich_guy_looker@> wrote:
          > > > > >
          > > > > > "MTF" stands for "Modulation-Transfer-Function"
          > > > > >
          > > > > > If you have a 100X PlanApo,
          > > > > > with a 1.3 NA, with a 10X eyepiece
          > > > > > for 1000X magnification,
          > > > > > and....
          > > > > >
          > > > > > A 63X PlanApo, with a 1.4 NA,
          > > > > > with a 15X eyepiece for a
          > > > > > magnification of 945X,..
          > > > > >
          > > > > > Do both have approximately the same
          > > > > > performance for the "Modulation-Transfer-Function"?
          > > > > >
          > > > > > thanks.
          > > > > >
          > > > > > Rich
          > > > > >
          > > > >
          > > >
          > >
          >
          >
          >
          >
          > ------------------------------------
          >
          > Yahoo! Groups Links
          >
        • John
          You could always try these Randall - with mirth or maybe not. I haven t tried it yet. http://www.microscopy-uk.org.uk/mag/artfeb11/dw-insect-test2.html
          Message 4 of 14 , Jun 3, 2012
            You could always try these Randall - with mirth or maybe not. I haven't tried it yet.

            http://www.microscopy-uk.org.uk/mag/artfeb11/dw-insect-test2.html

            http://en.wikisource.org/wiki/Page:Popular_Science_Monthly_Volume_7.djvu/194

            They are interesting. The ends disappear. Some say they are 125nm wide at the end. You could scale how far down one you can see.

            The micscape article refers to a test slide. As the advert mentions this is better than another slide I have. The mounting is excellent making it easier to focus on them.

            John

            --- In Microscope@yahoogroups.com, "merv" <merv.hobden@...> wrote:
            >
            >
            >
            >
            > Hi Randall,
            >
            > Apart from a Twyman-Green interferometer (Ha ha!) there is no singular test strategy which gives all the relevant information.I came to the conclusion some years ago that the major differences in objective performance were down to aberrations, irrespective of the NA. Neither the pinhole test (aka Francon) or the Abbe test platte can give detailed information on the extent of zonal aberrations.
            >
            > Only use of an interferometer, with a spread of wavelengths can uncover the true performance, and it is a tedious process. If you know the element curvatures, refractive indices and dispersions, you can do what Dr Fletcher did - put it through a ray trace programme, but even then it is difficult to state with complete accuracy. So I take my hat off to the 19th century pioneers - they were able to assemble very well corrected objectives with limited tests. Wenham was very well aware of the importance of zonal corrections, and he had some fairly simple, but sophisticated techniques for sorting out what was going on. One thing that seems to be essential, is the ability to think in 3D, with six degrees of freedom!I think both J.J. Lister and F.H Wenham had that ability, which no amount of algebra can replace easily!
            >
            > Cheers,
            >
            > Merv
            > --- In Microscope@yahoogroups.com, "Randall Buck" <rbuck@> wrote:
            > >
            > > Hi Merv,
            > >
            > > Thank you for the excellent insights.
            > > Even so, it would be nice to be able to sift through a set of objectives,
            > > using some sort of experimentally repeatable test that was a bit less
            > > subjective than my current method of mentally comparing each one via a
            > > favorite slide.
            > >
            > > If not the pinhole or MTF tests, is there anything you might recommend?
            > >
            > > Randall
            > >
            > > -----Original Message-----
            > > From: Microscope@yahoogroups.com [mailto:Microscope@yahoogroups.com]On
            > > Behalf Of merv
            > > Sent: Thursday, May 31, 2012 12:22 PM
            > > To: Microscope@yahoogroups.com
            > > Subject: [Microscope] Re: Objective NA and
            > > "Modulation-Transfer-Function" Performance?
            > >
            > >
            > >
            > >
            > >
            > >
            > >
            > > Hi John and Rene,
            > >
            > > MTF is a measure that originated in the early days of video systems as
            > > electronic engineers tried to correlate the idea of electronic bandwidth
            > > with spatial bandwidth in a lens system. They knew from Nyquist, that the
            > > minimum bandwidth required to fully represent the information in the signal
            > > was twice the maximum frequency component in that signal, and it seemed
            > > reasonable to assume that the same applied to spatial frequencies resolved
            > > by a lens. However, the idea is simplistic in that it assumes that the
            > > system is linear, with a square edged characteristic rather like a perfect
            > > low-pass filter. This is never the case with a real lens system, or for that
            > > matter, practical low pass filters! In reality, the aberrations in a lens
            > > system produce zonal phase aberration which distorts the MTF, giving zones
            > > of apparent better resolution for particular spatial frequencies. In the the
            > > summer edition of the 2012 Quekett Journal is an excellent article by John
            > > Fletcher on that very problem as found in the immersion objectives made by
            > > Robert Tolles. He shows that some of Tolles claims of better resolution
            > > depended on the existance of an annular area at the periphery of the lens
            > > with particular sensitivity to higher spatial frequencies under oblique
            > > illumination. This confirms something I have often observed with oil
            > > immersions - that an objective which is nothing special with central
            > > illumination, will give sometimes startling performance with oblique or
            > > annular illumination, and the only reason I could find is that the
            > > corrections were markedly different from that of an objective that worked
            > > well with central illumination. John contends that F.H. Wenham has been
            > > badly treated, as Wenham insisted that his tests showed that the aberrations
            > > in the Tolles objectives were large at the periphery of the aperture, and
            > > John's tests and simulations show that is indeed the case. So simplistic
            > > notions of image forming, based on numerical aperture, or MTF, can only give
            > > us a very limited perception of what is really occuring - what we call an
            > > image is a result of the convolution of all of the image components with the
            > > complex zonal transfer function of the lens. Zonal interference of periodic
            > > elements in the image structure may give higher contrast in some parts of
            > > the image, this, as John has shown, is purely fortuitous. It is fascinating
            > > that a battle which raged in the 1870s, is still, to a lesser extent,being
            > > waged today - for all the progress of the last 140 years, we are all still
            > > suckers for the simplistic mathematical explanation.
            > >
            > > Cheers,
            > >
            > > Merv
            > >
            > > --- In Microscope@yahoogroups.com, "John" <a.johnw@> wrote:
            > > >
            > > > Hello Rene. Maybe I should have added a little on the axis of an mtf
            > > graph. Vertical is contrast. The horizontal axis can be viewed as being
            > > resolution of target / resolution of optic. Contrast out of the optic for
            > > 100% contrast in reaches zero at 2NA / the wavelength of light being used.
            > > This is why the charts are the same for all optics as the horizontal axis
            > > normalises the curve.
            > > >
            > > > The resolution limit usually used in respect to microscope objectives is
            > > by Rayleigh and represents a contrast of some where around 7 1/2 % for a
            > > perfect optic and more or less the same for an optic that reaches his
            > > criterion which is less than 1/4 wave wave front error. That is the reason
            > > for this figure of merit. It's sensibly perfect for stars and many other
            > > purposes. In practice were contrast is low and variations in it are also low
            > > it's generally reckoned that wave front error of 1/15 wave are detectable by
            > > eye. Some have claimed 1/30 wave.
            > > >
            > > > John
            > > >
            > > > --- In Microscope@yahoogroups.com, "rene" <renevanwezel@> wrote:
            > > > >
            > > > > this is an old discussion. Reversing the question akes it easier to
            > > understand: if you put a reducing lens in the system, making the 100x
            > > planapo a 63x one, would it give the same quality image as the 63x?
            > > > >
            > > > > Pretty much so, in the centre. Not plano though, to correct the 100x for
            > > that huge increased FOV would need considerable correction. That will give,
            > > in practice, reduced contrast. How much depends largely on the date of
            > > manufacturing.
            > > > >
            > > > > Best wishes, René
            > > > >
            > > > >
            > > > > --- In Microscope@yahoogroups.com, "John" <a.johnw@> wrote:
            > > > > >
            > > > > > The MTF is a ratio in respect to the contrast an optic produces
            > > against the resolution it's trying to image so all optics have the same
            > > shape of mtf curve. Only what this means in terms of actual resolution
            > > varies. It's diffraction spot intensity/energy orientated.
            > > > > >
            > > > > > There's no point in getting too carried away with it. As most optics
            > > are not generally often used to image variably spaced black and white lines
            > > it more of an indication of the limitations of imaging optics in general. eg
            > > MTF50 is often mention is general photography - the point where black and
            > > white contrast is reduced to 50% but even the usefulness of that is dubious
            > > as pictures generally do not contain a lot of 100% contrast detail in them.
            > > What ever contrast is there will be reduced to 50% of it's real value. It
            > > seems that MTF50 has been used to rate quality photographic optics. The mtf
            > > curve up to that point is fairly good but goes to pot afterwards. Pass on
            > > how that is done because in some ways it makes no sense at all. :-) It's a
            > > case of well it's stated some where on the web so it must be true mustn't
            > > it?
            > > > > >
            > > > > > John
            > > > > >
            > > > > > --- In Microscope@yahoogroups.com, "rich_guy_looker"
            > > <rich_guy_looker@> wrote:
            > > > > > >
            > > > > > > "MTF" stands for "Modulation-Transfer-Function"
            > > > > > >
            > > > > > > If you have a 100X PlanApo,
            > > > > > > with a 1.3 NA, with a 10X eyepiece
            > > > > > > for 1000X magnification,
            > > > > > > and....
            > > > > > >
            > > > > > > A 63X PlanApo, with a 1.4 NA,
            > > > > > > with a 15X eyepiece for a
            > > > > > > magnification of 945X,..
            > > > > > >
            > > > > > > Do both have approximately the same
            > > > > > > performance for the "Modulation-Transfer-Function"?
            > > > > > >
            > > > > > > thanks.
            > > > > > >
            > > > > > > Rich
            > > > > > >
            > > > > >
            > > > >
            > > >
            > >
            > >
            > >
            > >
            > > ------------------------------------
            > >
            > > Yahoo! Groups Links
            > >
            >
          • Randall Buck
            Thanks Merv, I assume your Ha ha modifier is due to the unlikely probability of obtaining the necessary spherical mirror of appropriate focal length and
            Message 5 of 14 , Jun 3, 2012
              Thanks Merv,

              I assume your "Ha ha" modifier is due to the unlikely probability of
              obtaining the necessary spherical mirror of appropriate focal length and
              figure, which for high magnification objectives, is accurate over surface
              distances that are themselves near the wavelength of light?

              What if the Twyman-Green spherical mirror was larger and concave instead?
              I.e., placed at a distance from the front lens of the objective under test
              that is much greater than its working distance.

              Or perhaps I should just start working on those six degrees of freedom ?

              Randall



              -----Original Message-----
              From: Microscope@yahoogroups.com [mailto:Microscope@yahoogroups.com]On
              Behalf Of merv
              Sent: Sunday, June 03, 2012 5:33 AM
              To: Microscope@yahoogroups.com
              Subject: [Microscope] Re: Objective NA and
              "Modulation-Transfer-Function" Performance?





              Hi Randall,

              Apart from a Twyman-Green interferometer (Ha ha!) there is no singular test
              strategy which gives all the relevant information.I came to the conclusion
              some years ago that the major differences in objective performance were down
              to aberrations, irrespective of the NA. Neither the pinhole test (aka
              Francon) or the Abbe test platte can give detailed information on the extent
              of zonal aberrations.

              Only use of an interferometer, with a spread of wavelengths can uncover the
              true performance, and it is a tedious process. If you know the element
              curvatures, refractive indices and dispersions, you can do what Dr Fletcher
              did - put it through a ray trace programme, but even then it is difficult to
              state with complete accuracy. So I take my hat off to the 19th century
              pioneers - they were able to assemble very well corrected objectives with
              limited tests. Wenham was very well aware of the importance of zonal
              corrections, and he had some fairly simple, but sophisticated techniques for
              sorting out what was going on. One thing that seems to be essential, is the
              ability to think in 3D, with six degrees of freedom!I think both J.J. Lister
              and F.H Wenham had that ability, which no amount of algebra can replace
              easily!

              Cheers,

              Merv
              --- In Microscope@yahoogroups.com, "Randall Buck" <rbuck@...> wrote:
              >
              > Hi Merv,
              >
              > Thank you for the excellent insights.
              > Even so, it would be nice to be able to sift through a set of objectives,
              > using some sort of experimentally repeatable test that was a bit less
              > subjective than my current method of mentally comparing each one via a
              > favorite slide.
              >
              > If not the pinhole or MTF tests, is there anything you might recommend?
              >
              > Randall
              >
              > -----Original Message-----
              > From: Microscope@yahoogroups.com [mailto:Microscope@yahoogroups.com]On
              > Behalf Of merv
              > Sent: Thursday, May 31, 2012 12:22 PM
              > To: Microscope@yahoogroups.com
              > Subject: [Microscope] Re: Objective NA and
              > "Modulation-Transfer-Function" Performance?
              >
              >
              >
              >
              >
              >
              >
              > Hi John and Rene,
              >
              > MTF is a measure that originated in the early days of video systems as
              > electronic engineers tried to correlate the idea of electronic bandwidth
              > with spatial bandwidth in a lens system. They knew from Nyquist, that the
              > minimum bandwidth required to fully represent the information in the
              signal
              > was twice the maximum frequency component in that signal, and it seemed
              > reasonable to assume that the same applied to spatial frequencies resolved
              > by a lens. However, the idea is simplistic in that it assumes that the
              > system is linear, with a square edged characteristic rather like a perfect
              > low-pass filter. This is never the case with a real lens system, or for
              that
              > matter, practical low pass filters! In reality, the aberrations in a lens
              > system produce zonal phase aberration which distorts the MTF, giving zones
              > of apparent better resolution for particular spatial frequencies. In the
              the
              > summer edition of the 2012 Quekett Journal is an excellent article by John
              > Fletcher on that very problem as found in the immersion objectives made by
              > Robert Tolles. He shows that some of Tolles claims of better resolution
              > depended on the existance of an annular area at the periphery of the lens
              > with particular sensitivity to higher spatial frequencies under oblique
              > illumination. This confirms something I have often observed with oil
              > immersions - that an objective which is nothing special with central
              > illumination, will give sometimes startling performance with oblique or
              > annular illumination, and the only reason I could find is that the
              > corrections were markedly different from that of an objective that worked
              > well with central illumination. John contends that F.H. Wenham has been
              > badly treated, as Wenham insisted that his tests showed that the
              aberrations
              > in the Tolles objectives were large at the periphery of the aperture, and
              > John's tests and simulations show that is indeed the case. So simplistic
              > notions of image forming, based on numerical aperture, or MTF, can only
              give
              > us a very limited perception of what is really occuring - what we call an
              > image is a result of the convolution of all of the image components with
              the
              > complex zonal transfer function of the lens. Zonal interference of
              periodic
              > elements in the image structure may give higher contrast in some parts of
              > the image, this, as John has shown, is purely fortuitous. It is
              fascinating
              > that a battle which raged in the 1870s, is still, to a lesser extent,being
              > waged today - for all the progress of the last 140 years, we are all still
              > suckers for the simplistic mathematical explanation.
              >
              > Cheers,
              >
              > Merv
              >
              > --- In Microscope@yahoogroups.com, "John" <a.johnw@> wrote:
              > >
              > > Hello Rene. Maybe I should have added a little on the axis of an mtf
              > graph. Vertical is contrast. The horizontal axis can be viewed as being
              > resolution of target / resolution of optic. Contrast out of the optic for
              > 100% contrast in reaches zero at 2NA / the wavelength of light being used.
              > This is why the charts are the same for all optics as the horizontal axis
              > normalises the curve.
              > >
              > > The resolution limit usually used in respect to microscope objectives is
              > by Rayleigh and represents a contrast of some where around 7 1/2 % for a
              > perfect optic and more or less the same for an optic that reaches his
              > criterion which is less than 1/4 wave wave front error. That is the reason
              > for this figure of merit. It's sensibly perfect for stars and many other
              > purposes. In practice were contrast is low and variations in it are also
              low
              > it's generally reckoned that wave front error of 1/15 wave are detectable
              by
              > eye. Some have claimed 1/30 wave.
              > >
              > > John
              > >
              > > --- In Microscope@yahoogroups.com, "rene" <renevanwezel@> wrote:
              > > >
              > > > this is an old discussion. Reversing the question akes it easier to
              > understand: if you put a reducing lens in the system, making the 100x
              > planapo a 63x one, would it give the same quality image as the 63x?
              > > >
              > > > Pretty much so, in the centre. Not plano though, to correct the 100x
              for
              > that huge increased FOV would need considerable correction. That will
              give,
              > in practice, reduced contrast. How much depends largely on the date of
              > manufacturing.
              > > >
              > > > Best wishes, René
              > > >
              > > >
              > > > --- In Microscope@yahoogroups.com, "John" <a.johnw@> wrote:
              > > > >
              > > > > The MTF is a ratio in respect to the contrast an optic produces
              > against the resolution it's trying to image so all optics have the same
              > shape of mtf curve. Only what this means in terms of actual resolution
              > varies. It's diffraction spot intensity/energy orientated.
              > > > >
              > > > > There's no point in getting too carried away with it. As most optics
              > are not generally often used to image variably spaced black and white
              lines
              > it more of an indication of the limitations of imaging optics in general.
              eg
              > MTF50 is often mention is general photography - the point where black and
              > white contrast is reduced to 50% but even the usefulness of that is
              dubious
              > as pictures generally do not contain a lot of 100% contrast detail in
              them.
              > What ever contrast is there will be reduced to 50% of it's real value. It
              > seems that MTF50 has been used to rate quality photographic optics. The
              mtf
              > curve up to that point is fairly good but goes to pot afterwards. Pass on
              > how that is done because in some ways it makes no sense at all. :-) It's a
              > case of well it's stated some where on the web so it must be true mustn't
              > it?
              > > > >
              > > > > John
              > > > >
              > > > > --- In Microscope@yahoogroups.com, "rich_guy_looker"
              > <rich_guy_looker@> wrote:
              > > > > >
              > > > > > "MTF" stands for "Modulation-Transfer-Function"
              > > > > >
              > > > > > If you have a 100X PlanApo,
              > > > > > with a 1.3 NA, with a 10X eyepiece
              > > > > > for 1000X magnification,
              > > > > > and....
              > > > > >
              > > > > > A 63X PlanApo, with a 1.4 NA,
              > > > > > with a 15X eyepiece for a
              > > > > > magnification of 945X,..
              > > > > >
              > > > > > Do both have approximately the same
              > > > > > performance for the "Modulation-Transfer-Function"?
              > > > > >
              > > > > > thanks.
              > > > > >
              > > > > > Rich
              > > > > >
              > > > >
              > > >
              > >
              >
              >
              >
              >
              > ------------------------------------
              >
              > Yahoo! Groups Links
              >




              ------------------------------------

              Yahoo! Groups Links
            • Randall Buck
              Excellent, thanks John. Randall ... From: Microscope@yahoogroups.com [mailto:Microscope@yahoogroups.com]On Behalf Of John Sent: Sunday, June 03, 2012 6:21 AM
              Message 6 of 14 , Jun 3, 2012
                Excellent, thanks John.

                Randall

                -----Original Message-----
                From: Microscope@yahoogroups.com [mailto:Microscope@yahoogroups.com]On
                Behalf Of John
                Sent: Sunday, June 03, 2012 6:21 AM
                To: Microscope@yahoogroups.com
                Subject: [Microscope] Re: Objective NA and
                "Modulation-Transfer-Function" Performance?


                You could always try these Randall - with mirth or maybe not. I haven't
                tried it yet.

                http://www.microscopy-uk.org.uk/mag/artfeb11/dw-insect-test2.html

                http://en.wikisource.org/wiki/Page:Popular_Science_Monthly_Volume_7.djvu/194

                They are interesting. The ends disappear. Some say they are 125nm wide at
                the end. You could scale how far down one you can see.

                The micscape article refers to a test slide. As the advert mentions this is
                better than another slide I have. The mounting is excellent making it easier
                to focus on them.

                John

                --- In Microscope@yahoogroups.com, "merv" <merv.hobden@...> wrote:
                >
                >
                >
                >
                > Hi Randall,
                >
                > Apart from a Twyman-Green interferometer (Ha ha!) there is no singular
                test strategy which gives all the relevant information.I came to the
                conclusion some years ago that the major differences in objective
                performance were down to aberrations, irrespective of the NA. Neither the
                pinhole test (aka Francon) or the Abbe test platte can give detailed
                information on the extent of zonal aberrations.
                >
                > Only use of an interferometer, with a spread of wavelengths can uncover
                the true performance, and it is a tedious process. If you know the element
                curvatures, refractive indices and dispersions, you can do what Dr Fletcher
                did - put it through a ray trace programme, but even then it is difficult to
                state with complete accuracy. So I take my hat off to the 19th century
                pioneers - they were able to assemble very well corrected objectives with
                limited tests. Wenham was very well aware of the importance of zonal
                corrections, and he had some fairly simple, but sophisticated techniques for
                sorting out what was going on. One thing that seems to be essential, is the
                ability to think in 3D, with six degrees of freedom!I think both J.J. Lister
                and F.H Wenham had that ability, which no amount of algebra can replace
                easily!
                >
                > Cheers,
                >
                > Merv
                > --- In Microscope@yahoogroups.com, "Randall Buck" <rbuck@> wrote:
                > >
                > > Hi Merv,
                > >
                > > Thank you for the excellent insights.
                > > Even so, it would be nice to be able to sift through a set of
                objectives,
                > > using some sort of experimentally repeatable test that was a bit less
                > > subjective than my current method of mentally comparing each one via a
                > > favorite slide.
                > >
                > > If not the pinhole or MTF tests, is there anything you might recommend?
                > >
                > > Randall
                > >
                > > -----Original Message-----
                > > From: Microscope@yahoogroups.com [mailto:Microscope@yahoogroups.com]On
                > > Behalf Of merv
                > > Sent: Thursday, May 31, 2012 12:22 PM
                > > To: Microscope@yahoogroups.com
                > > Subject: [Microscope] Re: Objective NA and
                > > "Modulation-Transfer-Function" Performance?
                > >
                > >
                > >
                > >
                > >
                > >
                > >
                > > Hi John and Rene,
                > >
                > > MTF is a measure that originated in the early days of video systems as
                > > electronic engineers tried to correlate the idea of electronic bandwidth
                > > with spatial bandwidth in a lens system. They knew from Nyquist, that
                the
                > > minimum bandwidth required to fully represent the information in the
                signal
                > > was twice the maximum frequency component in that signal, and it seemed
                > > reasonable to assume that the same applied to spatial frequencies
                resolved
                > > by a lens. However, the idea is simplistic in that it assumes that the
                > > system is linear, with a square edged characteristic rather like a
                perfect
                > > low-pass filter. This is never the case with a real lens system, or for
                that
                > > matter, practical low pass filters! In reality, the aberrations in a
                lens
                > > system produce zonal phase aberration which distorts the MTF, giving
                zones
                > > of apparent better resolution for particular spatial frequencies. In the
                the
                > > summer edition of the 2012 Quekett Journal is an excellent article by
                John
                > > Fletcher on that very problem as found in the immersion objectives made
                by
                > > Robert Tolles. He shows that some of Tolles claims of better resolution
                > > depended on the existance of an annular area at the periphery of the
                lens
                > > with particular sensitivity to higher spatial frequencies under oblique
                > > illumination. This confirms something I have often observed with oil
                > > immersions - that an objective which is nothing special with central
                > > illumination, will give sometimes startling performance with oblique or
                > > annular illumination, and the only reason I could find is that the
                > > corrections were markedly different from that of an objective that
                worked
                > > well with central illumination. John contends that F.H. Wenham has been
                > > badly treated, as Wenham insisted that his tests showed that the
                aberrations
                > > in the Tolles objectives were large at the periphery of the aperture,
                and
                > > John's tests and simulations show that is indeed the case. So simplistic
                > > notions of image forming, based on numerical aperture, or MTF, can only
                give
                > > us a very limited perception of what is really occuring - what we call
                an
                > > image is a result of the convolution of all of the image components with
                the
                > > complex zonal transfer function of the lens. Zonal interference of
                periodic
                > > elements in the image structure may give higher contrast in some parts
                of
                > > the image, this, as John has shown, is purely fortuitous. It is
                fascinating
                > > that a battle which raged in the 1870s, is still, to a lesser
                extent,being
                > > waged today - for all the progress of the last 140 years, we are all
                still
                > > suckers for the simplistic mathematical explanation.
                > >
                > > Cheers,
                > >
                > > Merv
                > >
                > > --- In Microscope@yahoogroups.com, "John" <a.johnw@> wrote:
                > > >
                > > > Hello Rene. Maybe I should have added a little on the axis of an mtf
                > > graph. Vertical is contrast. The horizontal axis can be viewed as being
                > > resolution of target / resolution of optic. Contrast out of the optic
                for
                > > 100% contrast in reaches zero at 2NA / the wavelength of light being
                used.
                > > This is why the charts are the same for all optics as the horizontal
                axis
                > > normalises the curve.
                > > >
                > > > The resolution limit usually used in respect to microscope objectives
                is
                > > by Rayleigh and represents a contrast of some where around 7 1/2 % for
                a
                > > perfect optic and more or less the same for an optic that reaches his
                > > criterion which is less than 1/4 wave wave front error. That is the
                reason
                > > for this figure of merit. It's sensibly perfect for stars and many other
                > > purposes. In practice were contrast is low and variations in it are also
                low
                > > it's generally reckoned that wave front error of 1/15 wave are
                detectable by
                > > eye. Some have claimed 1/30 wave.
                > > >
                > > > John
                > > >
                > > > --- In Microscope@yahoogroups.com, "rene" <renevanwezel@> wrote:
                > > > >
                > > > > this is an old discussion. Reversing the question akes it easier to
                > > understand: if you put a reducing lens in the system, making the 100x
                > > planapo a 63x one, would it give the same quality image as the 63x?
                > > > >
                > > > > Pretty much so, in the centre. Not plano though, to correct the 100x
                for
                > > that huge increased FOV would need considerable correction. That will
                give,
                > > in practice, reduced contrast. How much depends largely on the date of
                > > manufacturing.
                > > > >
                > > > > Best wishes, René
                > > > >
                > > > >
                > > > > --- In Microscope@yahoogroups.com, "John" <a.johnw@> wrote:
                > > > > >
                > > > > > The MTF is a ratio in respect to the contrast an optic produces
                > > against the resolution it's trying to image so all optics have the same
                > > shape of mtf curve. Only what this means in terms of actual resolution
                > > varies. It's diffraction spot intensity/energy orientated.
                > > > > >
                > > > > > There's no point in getting too carried away with it. As most
                optics
                > > are not generally often used to image variably spaced black and white
                lines
                > > it more of an indication of the limitations of imaging optics in
                general. eg
                > > MTF50 is often mention is general photography - the point where black
                and
                > > white contrast is reduced to 50% but even the usefulness of that is
                dubious
                > > as pictures generally do not contain a lot of 100% contrast detail in
                them.
                > > What ever contrast is there will be reduced to 50% of it's real value.
                It
                > > seems that MTF50 has been used to rate quality photographic optics. The
                mtf
                > > curve up to that point is fairly good but goes to pot afterwards. Pass
                on
                > > how that is done because in some ways it makes no sense at all. :-) It's
                a
                > > case of well it's stated some where on the web so it must be true
                mustn't
                > > it?
                > > > > >
                > > > > > John
                > > > > >
                > > > > > --- In Microscope@yahoogroups.com, "rich_guy_looker"
                > > <rich_guy_looker@> wrote:
                > > > > > >
                > > > > > > "MTF" stands for "Modulation-Transfer-Function"
                > > > > > >
                > > > > > > If you have a 100X PlanApo,
                > > > > > > with a 1.3 NA, with a 10X eyepiece
                > > > > > > for 1000X magnification,
                > > > > > > and....
                > > > > > >
                > > > > > > A 63X PlanApo, with a 1.4 NA,
                > > > > > > with a 15X eyepiece for a
                > > > > > > magnification of 945X,..
                > > > > > >
                > > > > > > Do both have approximately the same
                > > > > > > performance for the "Modulation-Transfer-Function"?
                > > > > > >
                > > > > > > thanks.
                > > > > > >
                > > > > > > Rich
                > > > > > >
                > > > > >
                > > > >
                > > >
                > >
                > >
                > >
                > >
                > > ------------------------------------
                > >
                > > Yahoo! Groups Links
                > >
                >




                ------------------------------------

                Yahoo! Groups Links
              • merv
                Hi Randall, The ha ha refered to the fact that the memsahib would be looking up the number of her lawyer on her mobile phone if you brought one through the
                Message 7 of 14 , Jun 3, 2012
                  Hi Randall,

                  The 'ha ha' refered to the fact that the memsahib would be looking up the number of her lawyer on her mobile phone if you brought one through the door, and as I was reliably informed by an ex-employee of Barr and Stroud, continual use of such a device in dimly lit surroundings, can (as with other activities), send you blind....
                  The original Twyman-Green interferometer was only capable of measuring on axis aberrations.

                  A very clever modification by R. J. Bracey of the Scientific Instrument Research Association allowed the quantification of off-axis results. The final instrument was built by Adam Hilger. Details of this instrument can be found in 'Lens and Prism Making', F. Twyman, Hilger and Watts Ltd. Re your suggestion of the use of a concave lens, on page 438 of that book are drawings of a number of arrangements of lens interferometer, some of which use a concave lens outside the normal working distance, as well as the use of a reference objective and plane mirrors.

                  The Sira-Hilger interferometer was used with a concave mirror for short focus objectives. This consisted of glass hemisphere, silvered on the back, and with a piece of glass the thickness of a cover glass on its front surface. You could use this principle to make a simple interferometer to measure on-axis aberrations only - the basic instrument is a Michelson division of wavefront interferometer, and this could be relatively small if all you were interested in were microscope objectives. You could even brew something up from an old vertical illumination unit as the beamsplitter and use an old monocular microscope as the basic support structure..... F. H. Wenham describes a very neat way of making small glass hemispheres...
                  ('Your honour, he was a perfectly normal person until that wicked, malicious person supplied him with the details of how to make a microscope lens interferometer, which is why the accused has entered a plea of justifiable homicide.')

                  Cheers,

                  Merv (who also has too many microscopes...)

                  --- In Microscope@yahoogroups.com, "Randall Buck" <rbuck@...> wrote:
                  >
                  > Thanks Merv,
                  >
                  > I assume your "Ha ha" modifier is due to the unlikely probability of
                  > obtaining the necessary spherical mirror of appropriate focal length and
                  > figure, which for high magnification objectives, is accurate over surface
                  > distances that are themselves near the wavelength of light?
                  >
                  > What if the Twyman-Green spherical mirror was larger and concave instead?
                  > I.e., placed at a distance from the front lens of the objective under test
                  > that is much greater than its working distance.
                  >
                  > Or perhaps I should just start working on those six degrees of freedom ?
                  >
                  > Randall
                  >
                  >
                  >
                  > -----Original Message-----
                  > From: Microscope@yahoogroups.com [mailto:Microscope@yahoogroups.com]On
                  > Behalf Of merv
                  > Sent: Sunday, June 03, 2012 5:33 AM
                  > To: Microscope@yahoogroups.com
                  > Subject: [Microscope] Re: Objective NA and
                  > "Modulation-Transfer-Function" Performance?
                  >
                  >
                  >
                  >
                  >
                  > Hi Randall,
                  >
                  > Apart from a Twyman-Green interferometer (Ha ha!) there is no singular test
                  > strategy which gives all the relevant information.I came to the conclusion
                  > some years ago that the major differences in objective performance were down
                  > to aberrations, irrespective of the NA. Neither the pinhole test (aka
                  > Francon) or the Abbe test platte can give detailed information on the extent
                  > of zonal aberrations.
                  >
                  > Only use of an interferometer, with a spread of wavelengths can uncover the
                  > true performance, and it is a tedious process. If you know the element
                  > curvatures, refractive indices and dispersions, you can do what Dr Fletcher
                  > did - put it through a ray trace programme, but even then it is difficult to
                  > state with complete accuracy. So I take my hat off to the 19th century
                  > pioneers - they were able to assemble very well corrected objectives with
                  > limited tests. Wenham was very well aware of the importance of zonal
                  > corrections, and he had some fairly simple, but sophisticated techniques for
                  > sorting out what was going on. One thing that seems to be essential, is the
                  > ability to think in 3D, with six degrees of freedom!I think both J.J. Lister
                  > and F.H Wenham had that ability, which no amount of algebra can replace
                  > easily!
                  >
                  > Cheers,
                  >
                  > Merv
                  > --- In Microscope@yahoogroups.com, "Randall Buck" <rbuck@> wrote:
                  > >
                  > > Hi Merv,
                  > >
                  > > Thank you for the excellent insights.
                  > > Even so, it would be nice to be able to sift through a set of objectives,
                  > > using some sort of experimentally repeatable test that was a bit less
                  > > subjective than my current method of mentally comparing each one via a
                  > > favorite slide.
                  > >
                  > > If not the pinhole or MTF tests, is there anything you might recommend?
                  > >
                  > > Randall
                  > >
                  > > -----Original Message-----
                  > > From: Microscope@yahoogroups.com [mailto:Microscope@yahoogroups.com]On
                  > > Behalf Of merv
                  > > Sent: Thursday, May 31, 2012 12:22 PM
                  > > To: Microscope@yahoogroups.com
                  > > Subject: [Microscope] Re: Objective NA and
                  > > "Modulation-Transfer-Function" Performance?
                  > >
                  > >
                  > >
                  > >
                  > >
                  > >
                  > >
                  > > Hi John and Rene,
                  > >
                  > > MTF is a measure that originated in the early days of video systems as
                  > > electronic engineers tried to correlate the idea of electronic bandwidth
                  > > with spatial bandwidth in a lens system. They knew from Nyquist, that the
                  > > minimum bandwidth required to fully represent the information in the
                  > signal
                  > > was twice the maximum frequency component in that signal, and it seemed
                  > > reasonable to assume that the same applied to spatial frequencies resolved
                  > > by a lens. However, the idea is simplistic in that it assumes that the
                  > > system is linear, with a square edged characteristic rather like a perfect
                  > > low-pass filter. This is never the case with a real lens system, or for
                  > that
                  > > matter, practical low pass filters! In reality, the aberrations in a lens
                  > > system produce zonal phase aberration which distorts the MTF, giving zones
                  > > of apparent better resolution for particular spatial frequencies. In the
                  > the
                  > > summer edition of the 2012 Quekett Journal is an excellent article by John
                  > > Fletcher on that very problem as found in the immersion objectives made by
                  > > Robert Tolles. He shows that some of Tolles claims of better resolution
                  > > depended on the existance of an annular area at the periphery of the lens
                  > > with particular sensitivity to higher spatial frequencies under oblique
                  > > illumination. This confirms something I have often observed with oil
                  > > immersions - that an objective which is nothing special with central
                  > > illumination, will give sometimes startling performance with oblique or
                  > > annular illumination, and the only reason I could find is that the
                  > > corrections were markedly different from that of an objective that worked
                  > > well with central illumination. John contends that F.H. Wenham has been
                  > > badly treated, as Wenham insisted that his tests showed that the
                  > aberrations
                  > > in the Tolles objectives were large at the periphery of the aperture, and
                  > > John's tests and simulations show that is indeed the case. So simplistic
                  > > notions of image forming, based on numerical aperture, or MTF, can only
                  > give
                  > > us a very limited perception of what is really occuring - what we call an
                  > > image is a result of the convolution of all of the image components with
                  > the
                  > > complex zonal transfer function of the lens. Zonal interference of
                  > periodic
                  > > elements in the image structure may give higher contrast in some parts of
                  > > the image, this, as John has shown, is purely fortuitous. It is
                  > fascinating
                  > > that a battle which raged in the 1870s, is still, to a lesser extent,being
                  > > waged today - for all the progress of the last 140 years, we are all still
                  > > suckers for the simplistic mathematical explanation.
                  > >
                  > > Cheers,
                  > >
                  > > Merv
                  > >
                  > > --- In Microscope@yahoogroups.com, "John" <a.johnw@> wrote:
                  > > >
                  > > > Hello Rene. Maybe I should have added a little on the axis of an mtf
                  > > graph. Vertical is contrast. The horizontal axis can be viewed as being
                  > > resolution of target / resolution of optic. Contrast out of the optic for
                  > > 100% contrast in reaches zero at 2NA / the wavelength of light being used.
                  > > This is why the charts are the same for all optics as the horizontal axis
                  > > normalises the curve.
                  > > >
                  > > > The resolution limit usually used in respect to microscope objectives is
                  > > by Rayleigh and represents a contrast of some where around 7 1/2 % for a
                  > > perfect optic and more or less the same for an optic that reaches his
                  > > criterion which is less than 1/4 wave wave front error. That is the reason
                  > > for this figure of merit. It's sensibly perfect for stars and many other
                  > > purposes. In practice were contrast is low and variations in it are also
                  > low
                  > > it's generally reckoned that wave front error of 1/15 wave are detectable
                  > by
                  > > eye. Some have claimed 1/30 wave.
                  > > >
                  > > > John
                  > > >
                  > > > --- In Microscope@yahoogroups.com, "rene" <renevanwezel@> wrote:
                  > > > >
                  > > > > this is an old discussion. Reversing the question akes it easier to
                  > > understand: if you put a reducing lens in the system, making the 100x
                  > > planapo a 63x one, would it give the same quality image as the 63x?
                  > > > >
                  > > > > Pretty much so, in the centre. Not plano though, to correct the 100x
                  > for
                  > > that huge increased FOV would need considerable correction. That will
                  > give,
                  > > in practice, reduced contrast. How much depends largely on the date of
                  > > manufacturing.
                  > > > >
                  > > > > Best wishes, René
                  > > > >
                  > > > >
                  > > > > --- In Microscope@yahoogroups.com, "John" <a.johnw@> wrote:
                  > > > > >
                  > > > > > The MTF is a ratio in respect to the contrast an optic produces
                  > > against the resolution it's trying to image so all optics have the same
                  > > shape of mtf curve. Only what this means in terms of actual resolution
                  > > varies. It's diffraction spot intensity/energy orientated.
                  > > > > >
                  > > > > > There's no point in getting too carried away with it. As most optics
                  > > are not generally often used to image variably spaced black and white
                  > lines
                  > > it more of an indication of the limitations of imaging optics in general.
                  > eg
                  > > MTF50 is often mention is general photography - the point where black and
                  > > white contrast is reduced to 50% but even the usefulness of that is
                  > dubious
                  > > as pictures generally do not contain a lot of 100% contrast detail in
                  > them.
                  > > What ever contrast is there will be reduced to 50% of it's real value. It
                  > > seems that MTF50 has been used to rate quality photographic optics. The
                  > mtf
                  > > curve up to that point is fairly good but goes to pot afterwards. Pass on
                  > > how that is done because in some ways it makes no sense at all. :-) It's a
                  > > case of well it's stated some where on the web so it must be true mustn't
                  > > it?
                  > > > > >
                  > > > > > John
                  > > > > >
                  > > > > > --- In Microscope@yahoogroups.com, "rich_guy_looker"
                  > > <rich_guy_looker@> wrote:
                  > > > > > >
                  > > > > > > "MTF" stands for "Modulation-Transfer-Function"
                  > > > > > >
                  > > > > > > If you have a 100X PlanApo,
                  > > > > > > with a 1.3 NA, with a 10X eyepiece
                  > > > > > > for 1000X magnification,
                  > > > > > > and....
                  > > > > > >
                  > > > > > > A 63X PlanApo, with a 1.4 NA,
                  > > > > > > with a 15X eyepiece for a
                  > > > > > > magnification of 945X,..
                  > > > > > >
                  > > > > > > Do both have approximately the same
                  > > > > > > performance for the "Modulation-Transfer-Function"?
                  > > > > > >
                  > > > > > > thanks.
                  > > > > > >
                  > > > > > > Rich
                  > > > > > >
                  > > > > >
                  > > > >
                  > > >
                  > >
                  > >
                  > >
                  > >
                  > > ------------------------------------
                  > >
                  > > Yahoo! Groups Links
                  > >
                  >
                  >
                  >
                  >
                  > ------------------------------------
                  >
                  > Yahoo! Groups Links
                  >
                • merv
                  Thanks John, for the link to a truly excellent article by Dave Walker. It is fascinating that he shows all of the recorded aspects of the podura scale image,
                  Message 8 of 14 , Jun 4, 2012
                    Thanks John, for the link to a truly excellent article by Dave Walker. It is fascinating that he shows all of the recorded aspects of the podura scale image, apart from Royston-Pigott's beaded structure. With thin structures with a periodic content, the distance to the coverglass and the relative angle is a critical factor as you start to see multiple beam interference effects - often seen with Amphipleura Pellucida. A first class piece of work!

                    Cheers,

                    Merv

                    --- In Microscope@yahoogroups.com, "John" <a.johnw@...> wrote:
                    >
                    > You could always try these Randall - with mirth or maybe not. I haven't tried it yet.
                    >
                    > http://www.microscopy-uk.org.uk/mag/artfeb11/dw-insect-test2.html
                    >
                    > http://en.wikisource.org/wiki/Page:Popular_Science_Monthly_Volume_7.djvu/194
                    >
                    > They are interesting. The ends disappear. Some say they are 125nm wide at the end. You could scale how far down one you can see.
                    >
                    > The micscape article refers to a test slide. As the advert mentions this is better than another slide I have. The mounting is excellent making it easier to focus on them.
                    >
                    > John
                    >
                    > --- In Microscope@yahoogroups.com, "merv" <merv.hobden@> wrote:
                    > >
                    > >
                    > >
                    > >
                    > > Hi Randall,
                    > >
                    > > Apart from a Twyman-Green interferometer (Ha ha!) there is no singular test strategy which gives all the relevant information.I came to the conclusion some years ago that the major differences in objective performance were down to aberrations, irrespective of the NA. Neither the pinhole test (aka Francon) or the Abbe test platte can give detailed information on the extent of zonal aberrations.
                    > >
                    > > Only use of an interferometer, with a spread of wavelengths can uncover the true performance, and it is a tedious process. If you know the element curvatures, refractive indices and dispersions, you can do what Dr Fletcher did - put it through a ray trace programme, but even then it is difficult to state with complete accuracy. So I take my hat off to the 19th century pioneers - they were able to assemble very well corrected objectives with limited tests. Wenham was very well aware of the importance of zonal corrections, and he had some fairly simple, but sophisticated techniques for sorting out what was going on. One thing that seems to be essential, is the ability to think in 3D, with six degrees of freedom!I think both J.J. Lister and F.H Wenham had that ability, which no amount of algebra can replace easily!
                    > >
                    > > Cheers,
                    > >
                    > > Merv
                    > > --- In Microscope@yahoogroups.com, "Randall Buck" <rbuck@> wrote:
                    > > >
                    > > > Hi Merv,
                    > > >
                    > > > Thank you for the excellent insights.
                    > > > Even so, it would be nice to be able to sift through a set of objectives,
                    > > > using some sort of experimentally repeatable test that was a bit less
                    > > > subjective than my current method of mentally comparing each one via a
                    > > > favorite slide.
                    > > >
                    > > > If not the pinhole or MTF tests, is there anything you might recommend?
                    > > >
                    > > > Randall
                    > > >
                    > > > -----Original Message-----
                    > > > From: Microscope@yahoogroups.com [mailto:Microscope@yahoogroups.com]On
                    > > > Behalf Of merv
                    > > > Sent: Thursday, May 31, 2012 12:22 PM
                    > > > To: Microscope@yahoogroups.com
                    > > > Subject: [Microscope] Re: Objective NA and
                    > > > "Modulation-Transfer-Function" Performance?
                    > > >
                    > > >
                    > > >
                    > > >
                    > > >
                    > > >
                    > > >
                    > > > Hi John and Rene,
                    > > >
                    > > > MTF is a measure that originated in the early days of video systems as
                    > > > electronic engineers tried to correlate the idea of electronic bandwidth
                    > > > with spatial bandwidth in a lens system. They knew from Nyquist, that the
                    > > > minimum bandwidth required to fully represent the information in the signal
                    > > > was twice the maximum frequency component in that signal, and it seemed
                    > > > reasonable to assume that the same applied to spatial frequencies resolved
                    > > > by a lens. However, the idea is simplistic in that it assumes that the
                    > > > system is linear, with a square edged characteristic rather like a perfect
                    > > > low-pass filter. This is never the case with a real lens system, or for that
                    > > > matter, practical low pass filters! In reality, the aberrations in a lens
                    > > > system produce zonal phase aberration which distorts the MTF, giving zones
                    > > > of apparent better resolution for particular spatial frequencies. In the the
                    > > > summer edition of the 2012 Quekett Journal is an excellent article by John
                    > > > Fletcher on that very problem as found in the immersion objectives made by
                    > > > Robert Tolles. He shows that some of Tolles claims of better resolution
                    > > > depended on the existance of an annular area at the periphery of the lens
                    > > > with particular sensitivity to higher spatial frequencies under oblique
                    > > > illumination. This confirms something I have often observed with oil
                    > > > immersions - that an objective which is nothing special with central
                    > > > illumination, will give sometimes startling performance with oblique or
                    > > > annular illumination, and the only reason I could find is that the
                    > > > corrections were markedly different from that of an objective that worked
                    > > > well with central illumination. John contends that F.H. Wenham has been
                    > > > badly treated, as Wenham insisted that his tests showed that the aberrations
                    > > > in the Tolles objectives were large at the periphery of the aperture, and
                    > > > John's tests and simulations show that is indeed the case. So simplistic
                    > > > notions of image forming, based on numerical aperture, or MTF, can only give
                    > > > us a very limited perception of what is really occuring - what we call an
                    > > > image is a result of the convolution of all of the image components with the
                    > > > complex zonal transfer function of the lens. Zonal interference of periodic
                    > > > elements in the image structure may give higher contrast in some parts of
                    > > > the image, this, as John has shown, is purely fortuitous. It is fascinating
                    > > > that a battle which raged in the 1870s, is still, to a lesser extent,being
                    > > > waged today - for all the progress of the last 140 years, we are all still
                    > > > suckers for the simplistic mathematical explanation.
                    > > >
                    > > > Cheers,
                    > > >
                    > > > Merv
                    > > >
                    > > > --- In Microscope@yahoogroups.com, "John" <a.johnw@> wrote:
                    > > > >
                    > > > > Hello Rene. Maybe I should have added a little on the axis of an mtf
                    > > > graph. Vertical is contrast. The horizontal axis can be viewed as being
                    > > > resolution of target / resolution of optic. Contrast out of the optic for
                    > > > 100% contrast in reaches zero at 2NA / the wavelength of light being used.
                    > > > This is why the charts are the same for all optics as the horizontal axis
                    > > > normalises the curve.
                    > > > >
                    > > > > The resolution limit usually used in respect to microscope objectives is
                    > > > by Rayleigh and represents a contrast of some where around 7 1/2 % for a
                    > > > perfect optic and more or less the same for an optic that reaches his
                    > > > criterion which is less than 1/4 wave wave front error. That is the reason
                    > > > for this figure of merit. It's sensibly perfect for stars and many other
                    > > > purposes. In practice were contrast is low and variations in it are also low
                    > > > it's generally reckoned that wave front error of 1/15 wave are detectable by
                    > > > eye. Some have claimed 1/30 wave.
                    > > > >
                    > > > > John
                    > > > >
                    > > > > --- In Microscope@yahoogroups.com, "rene" <renevanwezel@> wrote:
                    > > > > >
                    > > > > > this is an old discussion. Reversing the question akes it easier to
                    > > > understand: if you put a reducing lens in the system, making the 100x
                    > > > planapo a 63x one, would it give the same quality image as the 63x?
                    > > > > >
                    > > > > > Pretty much so, in the centre. Not plano though, to correct the 100x for
                    > > > that huge increased FOV would need considerable correction. That will give,
                    > > > in practice, reduced contrast. How much depends largely on the date of
                    > > > manufacturing.
                    > > > > >
                    > > > > > Best wishes, René
                    > > > > >
                    > > > > >
                    > > > > > --- In Microscope@yahoogroups.com, "John" <a.johnw@> wrote:
                    > > > > > >
                    > > > > > > The MTF is a ratio in respect to the contrast an optic produces
                    > > > against the resolution it's trying to image so all optics have the same
                    > > > shape of mtf curve. Only what this means in terms of actual resolution
                    > > > varies. It's diffraction spot intensity/energy orientated.
                    > > > > > >
                    > > > > > > There's no point in getting too carried away with it. As most optics
                    > > > are not generally often used to image variably spaced black and white lines
                    > > > it more of an indication of the limitations of imaging optics in general. eg
                    > > > MTF50 is often mention is general photography - the point where black and
                    > > > white contrast is reduced to 50% but even the usefulness of that is dubious
                    > > > as pictures generally do not contain a lot of 100% contrast detail in them.
                    > > > What ever contrast is there will be reduced to 50% of it's real value. It
                    > > > seems that MTF50 has been used to rate quality photographic optics. The mtf
                    > > > curve up to that point is fairly good but goes to pot afterwards. Pass on
                    > > > how that is done because in some ways it makes no sense at all. :-) It's a
                    > > > case of well it's stated some where on the web so it must be true mustn't
                    > > > it?
                    > > > > > >
                    > > > > > > John
                    > > > > > >
                    > > > > > > --- In Microscope@yahoogroups.com, "rich_guy_looker"
                    > > > <rich_guy_looker@> wrote:
                    > > > > > > >
                    > > > > > > > "MTF" stands for "Modulation-Transfer-Function"
                    > > > > > > >
                    > > > > > > > If you have a 100X PlanApo,
                    > > > > > > > with a 1.3 NA, with a 10X eyepiece
                    > > > > > > > for 1000X magnification,
                    > > > > > > > and....
                    > > > > > > >
                    > > > > > > > A 63X PlanApo, with a 1.4 NA,
                    > > > > > > > with a 15X eyepiece for a
                    > > > > > > > magnification of 945X,..
                    > > > > > > >
                    > > > > > > > Do both have approximately the same
                    > > > > > > > performance for the "Modulation-Transfer-Function"?
                    > > > > > > >
                    > > > > > > > thanks.
                    > > > > > > >
                    > > > > > > > Rich
                    > > > > > > >
                    > > > > > >
                    > > > > >
                    > > > >
                    > > >
                    > > >
                    > > >
                    > > >
                    > > > ------------------------------------
                    > > >
                    > > > Yahoo! Groups Links
                    > > >
                    > >
                    >
                  Your message has been successfully submitted and would be delivered to recipients shortly.