Contrast methods and resolution

Begonnen von Rene, Dezember 12, 2008, 16:32:33 NACHMITTAGS

Vorheriges Thema - Nächstes Thema

Rene

Dear all,

We're working on a handbook for hydrobiological methods, in order to standarize protocols in use in the Netherlands. One chapter is on diatom analyses, and we have noticed that people use brightfield, phasecontrast, DIC and even someone adviced us to use Olbricht's Ergonom500 (...). I'd like to present an oversight (poster) on these different contrasting methods, but my math is lacking on some points.

Can someone explain to me why most (biological) phase contrast sets are not able resolve F saxonica, let alone A pellucida? My Leitz Heine can because it has a far larger annulus, aren't they made anymore? I've once heard the term 'strenges Zernike Verfahren', is this what it is called? What would the effect be with a ring at the edge of the aperture?

Can I call Anoptral a phase contrast method? I do not think the sooth ring in the objective is a retarder, but part of the method might be based on 'pseudo' phase effects like with annular lighting (COL). This would mean objectives from other manufacturers might give different results.

About Nomarski DIC I've once showed a MTF from MicroscopyU http://www.microscopyu.com/articles/optics/mtfintro.html (fig 5) that showed DIC worse then brightfield at the extreme limit. There are differences in setups nowadays (contrast/resolution?), has anyone numbers?

Thanks for any answers,

Rene.

Detlef Kramer

Lieber René,

Anoptral-Kontrast ist eine sehr kontrastreiche Variante des Phasenkontrastes. Zu den anderen Fragen möchte ich mich mangels Kompetenz nicht äußern. Eine kleine Anmerkung zum DIC vielleicht doch: Zeiss bietet aktuell zwei Varianten an: Kontrast-optimiert und Auflösung-optimiert. Ich denke, das hat seine Gründe!

Herzliche Grüße
Detlef
Dr. Detlef Kramer, gerne per DU

Vorstellung: Hier klicken

hinrich husemann

#2
Hallo Herr van Wezel,
Das Auflösungsvermögen im Phasenkontrast ist wegen der merklich geringeren Beleuchtungsapertur praktisch immer geringer als bei Hellfeld mit weitgehend ausgeleuchteter Objektivapertur (vorausgesetzt, das Objekt hat ausreichenden Kontrast und ist kein reines Phasenobjekt). Ein Beispiel: Mein Neofluar 40/0,75 löst die hexagonalen Strukturen von Pleurosigma angulatum im Hellfeld - vorausgesetzt die Kondensorblende ist weit genug geöffnet - gut auf (man muß allerdings etwas angestrengter gucken); im Phasenkontrast (hier PH2) aber nicht mehr so ganz.
Ich denke, etwas verallgemeinernd und scherzhaft formuliert kann man sagen: Das Produkt von Auflösungsvermögen und Kontrast ist konstant; und da man von letzterem immer ein gewisses Minimum braucht, kann man Ersteres i.A. nicht voll ausreizen. Allzu genauen Zahlenwerten sollte man deshalb skeptisch gegenüber stehen; die klassischen Berechnungsmethoden bringen wegen der dabei gemachten Idealisierungen im Grunde - größenmäßig sicher vernünftige -  Anhaltswerte.
Auch die verschiedenen Ansätze mit ringförmigen Aperturen, entweder im Objektiv oder im Kondensor ("Col") bringen nur einen sehr begrenzten Gewinn an Auflösungsvermögen, i. A. wieder gepaart mit Kontrastverlusten. Vielleicht ist für Standardisierungen normales Hellfeld das Einfachste.
So sehe ich es jedenfalls. Freundliche Mikrogrüsse
H. Husemann

Egbert Lange

Zitat von: Rene in Dezember 12, 2008, 16:32:33 NACHMITTAGS
Can I call Anoptral a phase contrast method? I do not think the sooth ring in the objective is a retarder, but part of the method might be based on 'pseudo' phase effects like with annular lighting (COL). This would mean objectives from other manufacturers might give different results.

Hallo Rene,

nach verschiedenen Autoren wurde für das Anoptral-Verfahren der Firma Reichert aus technologischen Gründen kein Ruß eingesetzt. Der vermutete Schichtaufbau ist in Françon, 'Progress in microscopy' auf Seite 79 dargestellt.   (nach 'progressinmicros00fran' googeln  ;) )

Die PhA-Objektive der Firma PZO hatten einen Ruß-Ring, für die PLUTA wohl die experimentellen Vorarbeiten erstellte.

Ausführliche Beschreibungen, einschließlich Bemerkungen zur Auflösung des Nomarski DIC, finden sich bei  PLUTA im Vol. 2 von 'Advanced Light Microscopy'.

MfG E.

Rene

Thanks Egbert, I completely forgot about Francon! I dug up my copy (in Deutsch!) and I am impressed with the detailed information on all these different phase variants. Can't remember to have seen that in Michel's Mikrophotographie. Anyway, I have found information about the thickness of the phase ring, but not really about the width of the ring compared to the objective aperture. So I'm still stuck with the question what image an objective would be able to give if the phase ring is build around the edge of the aperture.

Hallo Herr Husemann, of course resolution and contrast are biting eachother. People would use phase contrast because you can easily spot the lightest of silicified diatoms which tend to be underestimated in brightfield. Nevertheless, annular lighting (in quite a simple way) shows the striae of A pellucida clearly whereas most phase contrast sets are not even able to resolve F saxonica. I seem to remember Zernike calculated his annulus to be around 0.7x the objective aperture in order to fulfill the dogmatic 1.22lambda/2NA limit (but I cannot remember where I got this from and I have to admit I couldn't follow the maths). The Heine objectives at least work on this edge, but I can't understand why nowadays I cannot choose between, say 2 different sizes of phase ring? Surely there must be a use for it in for example material microscopy?

Rene.

Gunther Chmela

Zitat von: hinrich husemann in Dezember 12, 2008, 19:16:28 NACHMITTAGS
Auch die verschiedenen Ansätze mit ringförmigen Aperturen, entweder im Objektiv oder im Kondensor ("Col") bringen nur einen sehr begrenzten Gewinn an Auflösungsvermögen, i. A. wieder gepaart mit Kontrastverlusten.

Lieber Herr Husemann,

was aber, wenn ich bei COL definitiv eine Kontraststeigerung sehe (neben der Steigerung der Auflösung) ? Jedenfalls ist das bei sehr feinen Strukturen von Diatomeenschalen der Fall.

Beste Grüße!
Gunther Chmela

Rene

Hallo Herr Chmela, if you permit me to answer on this topic, if in COL with a perfect objective you increase the annulus size, you will see a decrease in contrast. The Heine condenser is quite a perfect tool for this. A Wild fluotar lens is a good lens to start investigations with.
Any increase in contrast while enlarging the COL aperture, I would attribute to the 'pseudo' phase effect of a not-fully corrected objective. This is actually a painful theorethical point in my recommendation for COL, which I have problems with to sell to professionals. They generally do not understand and surely do not like the idea that a more perfect lens on their blingbling microscope would actually give worse results.

Best wishes, René.

Gunther Chmela

Zitat von: Rene in Dezember 14, 2008, 00:21:59 VORMITTAG
Hallo Herr Chmela, if you permit me to answer on this topic, if in COL with a perfect objective you increase the annulus size, you will see a decrease in contrast. The Heine condenser is quite a perfect tool for this. A Wild fluotar lens is a good lens to start investigations with.
Any increase in contrast while enlarging the COL aperture, I would attribute to the 'pseudo' phase effect of a not-fully corrected objective. This is actually a painful theorethical point in my recommendation for COL, which I have problems with to sell to professionals. They generally do not understand and surely do not like the idea that a more perfect lens on their blingbling microscope would actually give worse results.

Best wishes, René.

Lieber René,

meine Erfahrungen basieren auf Beobachtungen mit den wirklich besten Objektiven (Zeiss Neofluar 40:1, 0,75, Zeiss Planapo 40:1, 1,00). Auch mit "schlechteren" Objektiven (Zeiss Achromat 40:1, 0,65 u.a.) habe ich dieselbe Erfahrung gemacht. Alles bei optimaler Einstellung des Strahlengangs, der Zentrierung aller optischen Elemente etc.

Ich will Ihre Aussagen wirklich nicht anzweifeln, doch - zumindest bei mir - steht hier Lehrmeinung gegen Praxis!

Vielleicht sieht der Rest der Welt etwas anderes als ich?  ???

Beste Grüße!
Gunther Chmela

Rene

Hello Gunter,

I will have to try out the 40/1.0 myself to see what it is doing. But the design is quite old, which sounds promising for COL experiments. The correction I am talking about is sferical abberation that often drops off steeply at the edge of the aperture. At least the relatively modern Olympus splanapo 60/1.4 is corrected too much to be useful, in our lab we're using a Leitz NPl 100/1.3 fluotar routinely. All achros I have tried thus far do fine as well; I have however no data on modern infinity lenses. The Wild Fluotar 100/1.3 is amazingly well corrected as to be useless with high NA annular lighting. At the time I was starting out with annular lighting, and I was surprised to see this lens not performing like all other lenses I did try out until then. So I searched high and low and in the end I found 10 of these lenses, both phase and non-phase. They all behaved the same in this respect which I couldn't understand. Later on, Sterrenburg pointed out to me that the Wild fluotars are corrected without compromises (they surely didn't do much on Bildfeldwelbung for example).

Best wishes, René.

Rene

Thomas, just to be sure, I am talking about contrast differences at high NA-COL between different lenses. If the Nikon planapos are not older then the 90's, I do expect to see a difference in contrast with the others. There is still always contrast difference between COL and brightfield, because you are blocking considerable amount of non-contributing light. But the phase effect is a stronger contrast mechanism.

Hope that clarifies things, René.