Hi all,
For those interested in the schlieren imaging aspects of this technique, there is a fascinating article in "American Scientist" V94 N 1 Jan-Feb 2006 titled"High-speed Imaging of Shock Waves, Explosions and Gunshots".
  Link : http://www.americanscientist.org/template/AssetDetail/assetid/48547#48720

From the description of how the imaging works, there appear to be some similarities the DLC method, although it seems to me that a second knife edge should also be used above the objective. Incidentally, is this also how telescope mirrors are tested using the "Foucault test"?, as this also uses a knife edge.

regards, Matt

It only uses one curved edge not two or any other method to analyze the image.

Gordon

"Kevin" wrote:
> Does this "DLC" method use only one knife-edge, or two? Mention of
> knife-edges makes me think of the Schlieren method to image
> variations in the refractive index of transparent media.

Daron,
I am not supporting DLC - there is no room in the conventional physics of light for that resolution. I am playing with it and it is an oblique method. I haven't made a sharp enough edge on my stops to comment on diffraction. I probably never will be able to. The physics I learned doesn't include understanding the duality of light as a wave and particle at the same time. But it won't stop me from giving it a try.

The author of the piece on DLC uses a convex stop to [diffract] the light. Fritz Zernike made his phase contrast with both rings and crosses. Rings won out for biology as cells are round and it worked best with them and it is easier to make by a good deal. Semiconductor folks are said to prefer the crosses as it was supposed to work better with straight lines. That may play a part in DLC over a slit or a wire. Two wires in the condenser and one in the objective would make some kind of phase microscope. Slits could be substituted. Double edged razor blades smoked in a candle flame are as good as they come once set until they rust or the soot comes off.

Using a large sequin with the hole blocked works as surprisingly good COL stop on the glass of the 90 degree prism of my Zeiss using the field diaphragm to make the outside ring. I was surprised that worked as well as it did as the condenser has to be moved way down out of position.

I don't think the answer to the piece on DLC claims to much greater resolution than Herr Abbe proposed will stand up but it would be nice for amateurs if it did.

There are lots of things you can do to improve contrast of a modest scope not many to improve resolution but to step up and buy better stuff once you have your lighting. Unfortunately if you have any kind of scope at all you have to spend lots of money to get a little more resolution. A decent achromat will resolve almost all there is to see. It won't take good pictures or work well away from the center of the lens.

The method with wire seems the most harmless to try. I may try it with the curved stop with a curved wire.

Gordon

Hi Folks,
I have obtained permission from Barry Piekos, the author of the patent and article to post this link to a pdf of the Microscopy Today article:
  Link : http://pantheon.yale.edu/~wbp2/DLC.pdf
so everyone may have full understanding of his concept.

I'd be interested to hear if anyone achieves similar results by utilizing this technique and concurs with the hypothesis.

Salutations, Tim

"Tim Wilson" wrote:
> I'd be interested to hear if anyone achieves similar results by
> utilizing this technique and concurs with the hypothesis.

Tim,
I am only an amateur microscopist, so I won't bore you with my failures to reproduce the results.

However, from a theoretical standpoint...
I have read this article several times. It seems a careful piece of work, but I find the main point less than fully convincing. The article presents essentially three pieces of evidence for super-resolution.

Figure 1 shows an image of a high resolution reference standard. For this standard, the DLC setup resolves a 400 nm grating but not a 330 nm grating, while Köhler illumination barely resolves 500 nm. The article discusses this, arguing that the theoretical resolution of the grating is 390 nm (lambda / 2NA), versus the Abbe limit of 1.22 lambda / 2NA = 480 nm. This part seems fairly convincing, but we're only talking about a 20% difference, and that part supported by theory.

Figure 2 then shows an image supposedly showing 300 nm and 50 nm beads, and claiming that two closely opposed 50 nm beads are resolved as slightly overlapping circles. This is a significantly more aggressive claim, reaching something like 4X better than Abbe.

Figure 3, especially 3F, raises the bar again, showing images of a cell nucleus including features that approach 20 nm on the subject, judging from the scale bar.

There is a logical flaw lurking here.
The argument essentially says "The images contain fine detail, therefore (computing backward) extremely fine detail is resolved on the subject."
As a victim of floaters, I am painfully aware that the finest detail "shown" by my microscope actually belongs to my eyes. This and other experiences makes me leery of the computing backward argument. It is altogether too easy for images to contain features that arise more in the optical system than in the subject.

From my standpoint, a more convincing demonstration would show that for the same subject, the same overall structure and fine irregular details are resolved both by DLC and by some conventional high resolution imaging technique such as scanning electron microscopy.
A cursory Google search finds Barry Piekos identified as "supervisor of the electron microscope laboratory in the Department of Molecular, Cellular and Developmental Biology" at Yale (http://yalemedicine.yale.edu/ym_su00/scope/etcetera.htm).
This would seem to be an ideal position from which to produce a convincing direct comparison.
I presume that only a shortage of time has prevented appearance of that comparison thus far, and I eagerly await the publication. It would be fun to see long-standing theory so thoroughly shaken up. :-)

-Rik

Hi Rik,
Thank you for your report. I'm assuming you saw no enhancement of images whatsoever.

I'll pass your suggestion on to Dr. Piekos.

Salutations, Tim

Or should I have entitled this thread with "How to get famous in the community with irreproducible results"?

I spent some time on "reproducing/understanding" a recent claim published in "Microscopy Today". Now, I read about it on the front page of Micscape Magazine.
Before some of the community of amateur microscopists is led too deep into the dark and starts to entertain the thought that a 10-fold increase [in resolution] can happen just like that, I want to add my two cents to this discussion.

Maybe you have already heard about DLC as published in the November issue of Microscopy Today. If not, take a look at the recent message in the Micscape Magazine. The editors of Micscape Magazine wrote:
----begin
Diffracted light contrast: Improving the resolution of a basic light microscope by an order of magnitude' by W. Barry Piekos, Yale Univ. This article in the November 2006 issue of 'Microscopy Today' has, not surprisingly, prompted interest and discussion amongst microscopy enthusiasts, both in the public forums, i.e. see Yahoo Microscopes forum thread and personal communications. Essentially a very simple stop at the plane of the field lens iris and defocused with condenser is reported to offer considerable benefits. As it's a technique the enthusiast can in principle readily reproduce why not have a try with your own system?
----end

Why would an amateur be able to "readily reproduce" this? I tried for many hours without success. I tried to find any possible reason to explain why this method brings benefits beyond oblique illumination. I found nothing. Nothing at all!

My experience in Physics tells me that the claim is false. It's a joke at best. My experiments tell me that it is wrong. It does not even compete with a carefully established circular oblique illumination (COL) because the images are full with unwanted interference.

Fourier optics clearly explains the theories of how a spatially coherent light source illuminates a specimen, even when said source is generated by a convex-shaped stop located somewhere close to the field diaphragm. It is also perfectly clear why for an ordered substrate a spatially coherent light source offers the BEST resolution. But that's well documented in the literature since many years. For instance see Slayter et al in "Light and Electron Microscopy", chapter 9 "Resolution", Cambridge University Press (2000).

In the future, I really hope that the editors of Microscopy Today and Micscape Magazine distance themselves a little from this kind of claims. Let the scientific community decide whether or not this method 10-folds the resolution of a standard LM. Such claims should first be filtered by experts in the field before being published or mentioned in decent journals for the microscopists. I am very sure that this claim is certainly not "readily reproduced". It is irreproducible.

Your comments are appreciated.

Thanks, Gregor

Perhaps you should have read some recent posts on this forum prior to authoring this post. There are a few comments regarding this.

Salutations, Tim

BUT . . .
It is a rare thing that a hobbyist can read a paper, try to fathom it, and then turn around and try to replicate the results using a 'lowest common denominator' scope.

Why shouldn't we give it a try? What [do] we really have to lose? Wouldn't one learn something in the process? Finally, why would anyone wait for 'experts' to do an experiment he or she could readily do for themselves.

'Try and see' seems to be easy enough in this case doesn't it? And suppose in trying one learns about their scope, or finds a derivative method, or re-discovers some obscure oblique technique known to the 'ancient world' . . . might come in handy some day :-)

As to 'fame' . . . only if the method works . . . and in this case I get to be the judge of that for myself -- so can you too!

FWIW, Steve

Dear all,
I think I mentioned before that I was not convinced that the published results proved the increase in resolution the author claims.

So I started over to see what could be done. There is a video on Gordon's site that shows the progress of a convex edge (Mylar -- radius 66mm) across the beam in the position recommended by the author. Naturally it's a bit shaky because the edge was moved by hand:
  Link : http://www.science-info.net/pages/EDFWilliams/VIDEO/656DX.avi

Step 2: Without changing focus the light was re-centered. The filament image was focussed carefully on the condenser diaphragm and then both it and field diaphragms were opened fully. Naturally no detail whatsoever could be seen. This is obvious from the start of the video.

Step 3: The curved edge of a piece of Mylar cut from an old 5.25" floppy disk was slowly moved across the surface of the optical glass cover over the mirror in the base of the frame, just as the author describes. The longitudinal axis of the A pellucida valve was 'normal' to the edge (if such is possible with a curved surface) at a certain stage the lines become visible on the monitor -- but they can't be seen through the binocular because the contrast is too low. The lines were not visible when the curved surface was 'lined' up with the longitudinal axis of the specimen. I was looking for dots since this method is supposed to improve resolution. I saw none.

I moved the condenser down a little then up again and shifted the edge a little and took a few still pictures, several are in the photo album -- 849 to 852. I could not get as good an image as could be obtained with simple deflected light. I'm not sure if I'll go on with this because it's a terrible fiddle. A mechanism needs to be made to move the edge accurately across the beam and rotate it as well. I'm not sure I haven't been using this method with an offset Aplanatic condenser anyway. Except in this case the edge is convex and according to the paper less useful. When the light cone comes in at an angle, whether it's because it was tilted by being offset, deflected by an extra lens, or cut down by an off-centre diaphragm, there are edges all over the place and the final image might well be a stew of physical effects.

I have a few extra Aplanatic condensers and if I remove the lens from one it might be possible to use it as a translation mechanism for a DLC edge. I'll look into this today.

D

After more careful experimentation and the use of a phase telescope to see how much of the beam is cut off by the edge, I have come to the conclusion that what I am seeing is the result of rather inefficient oblique lighting. When the lines are most well defined, the edge masks off just about half the light. By the way I've tried a straight edge and there is no visible difference, but there is hardly anything to see anyway. The edge is sharp in the telescope and so it's in the right place according to the description. And, according to the laws of optics, there must be diffraction. But whether diffraction is seriously involved in the imaging of the lines is something I cannot decide upon right now.

However, (in a few seconds) by moving the mirror up to the left (about 10:30 O'Clock) until (through the telescope) it's a tiny bright spot at the very edge of the field, I can get a clear contrasty image of the lines, with a perfectly centered condenser.

D

Some of you may remember last year I nearly tore my hair out trying to see lines - let alone dots in Ap. The recent papers and topic have renewed my interest.
I have two Kemp Ap slides, one old NBS strew and a Realgar mount kindly lent me by Carel Sartory.

Last night With all four slides I first used B/F, D/F and Phase. I used Zeiss x100 Neofluar 1.3 and KPL x10's. O.9 Zeiss phase condenser (I have achro 1.4's to try later, but as immersion of the condenser is not possible with the adjustment required by the suggested techniques 0.9 - 1.0 is all that's available in air). Also tried Optovar x2 in all situations. Nothing! Beautifully defined Ap, not a sign of lines.

I then went through the suggested techniques with every slide. I used convex,concave,straight edges at every angle and height possible. I introduced spare lenses into the substage position, on axis, offset etc. Nothing - although I got a lot of very nasty diffraction.

I took two photos using a Canon 300D with mirror lockup on aperture priority using a Zeiss x12.5 KPL goggles relay (I've left my photo eyepieces elsewhere). Zeiss trino, exposures about 3 seconds.

Originally 6 Megs reduced in photoshop. The photos are in my folder (Ray Sloss) [Photo Ap2 is shown here ].

I presume this was a sortof 'jury rig' offset illumination.

 
I then searched the slide (this was the Realgar slide). I was surprised to find only a few of the Ap's exhibited line structure, most appearing entirely clear within the main body structure.

I then spent two hours searching the NBS and two Kemp slides using the same technique - nothing. Can it be that some preparation methods of Ap destroy the internal structure?

I was about to embark on a few days of trying better objectives and condensers when my halogen bulb decided it had had enough. When I've collected my spares from elsewhere I'll post some more photos using COL.

Merry Christmas and happy new year to everyone - and may the wavelength of light shorten for our convenience!

Ray

Hi Ray, thanks for description of your experiments! I remember something like using Risley prisms on Surirella gemma (who was it again??) during the Grayfield wars on resolution. Very tricky indeed.

> I then searched the slide (this was the Realgar slide). I was surprised to find only
> a few of the Ap's exhibited line structure, most appearing entirely clear within the
> main body structure.

I assume you check orientation of the striae with oblique direction of illumination?!

> I then spent two hours searching the NBS and two Kemp slides using the same technique - nothing.
> Can it be that some preparation methods of Ap destroy the internal structure?

Well, damage by silica dissolution occurs during storage before mounting, some damage of the finest details for SEM/TEM by 'rough' oxidation (peroxide bubbling!).
Anyway, examination of the oxidized Kemp material in the SEM there's a lot of crap visible in the valves. The same material mounted in resin shows far more useful valves, so the crap must be transparent. Not sure what effect realgar has on that transparency. But if it has been made some 100 years ago, who knows the purity of the reagents used...

Rene.

Tim,
Why shouldn't an amateur be able to reproduce this if it works. I have better optic and 35 mm cameras equal to what Barry used he makes no claims of any special conditions or tools I don't have. I do need to get the subjects he used. I can borrow the slide an I hope to have mica partials that size on the way today or tomorrow. With Mica I can also tinker with polarized light. My setup is far from the best I have seen. And is far more versatile than any on campus. Some of the "amateurs" that have worked on this ran microscope facilities for entire countries before they retired.

This group has great deal more depth than shows up in the on line posts. There are a number of professionals that follow the discussions and don't get in the fray for one reason or another. Some of their discussions come my way. Some of the list members discuss things in private they don't want to discuss on list and have to explain the physics or other details of the discussion that many of the group ask about and don't have the background to understand the answer leading to hard feelings all around.

Following his patent things worked out to my surprise about like he claimed. I don't think that is is a good a the best illumination but it is pretty good for a condenser as far out of place as I had it in some configurations. I believe he tossed out the claim of super resolution without a theory to support in a journal with no review as an observational discovery for others to explore if they wished. A bit more information about the nature of illumination with diffracted light would help my own experiments show the nature of the edge has large effect on the light.

I don't feel I have explored the are of illumination with diffracted light enough to from an opinion based on observation and I can find no theoretical base to guide me so I need to design an experiment to explore it by observation if I am to proceed much further. An of the cuff list is a group of slides that include latex spheres the same size used in the paper, diatoms every on knows, a target to measure distortion and some live biologcal samples made up every time. Then set up a microscope with a optical bench for light train so the stops can be inserted anywhere in the path from the conventional positions in and below the condenser to the field stop and and in between. Then make a set of stops including COL, darkfield Barry's and others with various thicknesses edges form knife edge to coin sized. Compare every set up against the best image possible rating them with a transparent indexing method. Its a very complex project to undertake for a project with at the very best shaky theoretically basis. The promise of high resolution at low cost has lured people to it since the observation of diffraction of light of pascals too small to resolve in the early days of darkfeild microscopy leading to the field of ultra microscopy over 100 years ago.

Gordon

Gordon,
You must have me mixed up with someone else. I do think amateurs can attempt to reproduce this and suggested as such in my previous post. I now have a range of photos which I am attempting to post which show some of the DLC (?) effect but have not been able to access the photos section from my other computer yet.

Salutations, Tim

Sorry Tim, for my misunderstanding.

Gordon

Hi Folks, (sorry if this post appears twice)

I have started a photo album 'Tim Wilson' in the photos section and have posted a series of photos using a convex edge lying on the field light glass. I used a 1/4 section cut out of a Canadian penny coloured black.

All photos are bright field with both apertures wide open. I used a 20X and 63X lens.

First of all I find the enhancement subtle. For someone with a cheap bright field scope this subtlety (sp?) could be a real bonus, especially if they are just looking for the presence of certain organisms.

Second, at least in my case, the effect conveys very poorly from scope looking down the eye pieces to camera to computer screen.

The series of photos show the specimens (testate amoebae) in bright field adjusting for Köhler using the convex edge. There may be some effect from the edge at this setting but very little. Then there are varying degrees of the condenser raised above (red) or below (blue) Köhler. You may see that some lines and details become more visible as the condenser is set at various levels.

I have no clue what is the cause of this subtle enhancement (oblique illumination or DLC) but I have yet to see the effects claimed by Barry Piekos. This does not mean he is not correct.

In my world of just trying to help folks with cheap bf microscopes discern bacteria and protozoa in compost extract soil amendments, I prefer the hook shaped object (mentioned in a previous post) I cut out of plastic or the concave edge of a flat washer cut in 1/4 or 1/2.

Salutations, Tim

I have posted a picture of a knife edge test on a telescope mirror that did a few years ago. This is to stir some possible thoughts about the recent posts on resolution.

The mirror is 16 inches diameter and the camera used to capture the pic is simple video cam mounted half way behind the knife with the light source behind knife edge half way covered. Both placed at the radius of curvature (R.O.C) of the mirror tested.

The amplitude of the brighter streaks ( tool marks) on the surface are about 10 nm. There is no way the camera could resolve this but we are able to see this feature easily. Is this possibly what is occuring with the curved knife on the microscope, that the amplitude is resolved in relation to some reference focus point? Just a thought.

The picture is under "My other scope" listed as "Knife edge test"
Still zealous! Ron

After putting the KK slide away for a while I tried DLC again. This video is about the best I've managed so far. I'm still not sure it's all diffraction contrast. It looks rather like oblique lighting to me. But the knife edge is where it ought to be and the field, at low magnification (440X), is uneven as the theory suggests. But by clicking in the 2.6X lens in the magnification changer I could choose an area that seemed to be more evenly illuminated.

The video is a bit long but shows some interesting action:
  Link : http://www.science-info.net/pages/EDFWilliams/VIDEO/669DX.avi

The edge is convex -- a piece of Mylar cut from a 5 1/4" floppy disk. Objective Leitz EF 40X dry; Condenser 1.4 Aplanatic centered -- Köhler; photo eyepiece 10X Periplan; mag changer 2.6X. DivX compression.

D

Don,
Do you have any idea the diameter of the flagella on the organism?

An earlier paper by Piekos states the radius of the concave edge to be 32 mm.

While I haven't taken any pictures I get results about like yours. I haven't seen anything worth taking a picture of. I made COL stop that uses the field iris to make the outer circle to use as a standard to test against. I haven't found any results that were any better than I got with the COL lighting. I think by comparing the results to COL the effects of oblique lighting can be accounted for to some degree.

Here are some observations I have made on stops. I don't have any idea what the ideal stop is. The dull side of a floppy disk is probably a good place to start. I have found the construction and reflectivity of the stop to have a considerable effect on the nature of the light I don't know how much it matters yet. A razor sharp edge that has been smoked with soot produces the least diffraction. When it was shiny before it was smoked it was one that scattered the most light. I was able to see double image using smoked knife edge (a X-Acto #23 knife blade ). I believe it a reflection of the edge on the glass. I haven't seen it with any other stop but none of them scatter so little light as the smoked knife edge. I was hoping this blade with its curve edges would act as an improved Mathias arrow. I didn't see a great deal of improvement. The next step for me is to put the stops in a condenser at the historically proper place and see how they act.

Working with copper leaf a few microns thick glued to glass it makes a difference if the glass or copper side is up leading me to question if the stop should be placed on glass at all.

I am surprised how well the method works with the light wide open and the lighting improperly adjusted. I am using a simple Abby condenser so there is considerably more aberration in the light from the stop on my system than yours. I need to mount a better condenser. I haven't seen anything that looks like super resolution.

This has peaked my interest in how the edges of stops should be finished.

I left a page out of Axelrod's paper on the Zero Cost Schlieren microscope I have correct that. If I sent it to someone contact me for a corrected copy.

Gordon

"Don Williams" wrote:
> The edge is convex -- a piece of Mylar cut from a 5 1/4" floppy disk.

The following is based on additional information about "DLC" published in Micscape Magazine.
From: Cover of Micscape Magazine from December 2005 (issue 134)
URL: http://www.microscopy-uk.org.uk/
Information obtained from the above URL on 12/21/2006 at 11:48PM PST
--begin--
Dec. 21st 'DLC' update. The author Barry Piekos has kindly sent us a recent image of the well known test diatom Frustulia rhomboides taken with DLC. Barry remarks: 'the striae in F. rhomboides have a periodicity of ~300 nm and they are easily resolved using my DLC technique. Indeed, it appears as though the dots are resolved in the upper left corner of the diatom'. Details supplied: Leitz plan fluotar 40x NA0.70 objective. Leitz Diaplan microscope. 40 mm opaque stop used at field iris plane as described for DLC. Image shared with author's permission.
--end--

The limit of resolving power is given by:
  d[µm] = 0.61*lambda[µm]/NA(Objective),  if NA(Condenser) > NA(Objective).

[ A table shows the resolving power, d[um], for a range of illuminating light wavelengths, lambda[nm], as compared with a typical diatom frustule line spacing : ]

Summary:
1) If DLC improved resolution beyond COL (or DF), we must see all lines clearly. Otherwise, it does not uphold the author's claim of a ten-fold resolution improvement. It is unclear whether the author used monochromatic filters.

2) Since decades, microscopists are able to resolve subresolution particles by means of special contrast methods (such as DF and oblique illumination). However, since the observed size of these subresolution particles is just around twice the resolving power d, it becomes very difficult to clearly distinguish the frustules of Frustulia rhomboides with NA(Objective) = 0.70.

3) Although the provided image is impressive when obtained with such an objective, it clearly shows that the method employed is diffraction limited and hence seems to be comparable with COL. From my experiments, however, I cannot confirm that DLC works as well as COL.

Gregor

Yep, I agree with you guys, nothing extraordinary discovered (yet). I just hope Piekos didn't look to far into the noise to search for structures he recognised. That would be a typical Rife-like trap he should know only too well as EM tech. He should.

Anyway, I have uploaded a pdf in the File section by Michael Bingley claiming super resolution by interference effects with critical illumintion. Rife back and fro.

I still think it's a good thing for Micscape to get amateurs involved. Anybody new starting playing with illumination that discovers contrast enhancement by simple methods is a bonus.

René.

> 3) Although the provided image is impressive when obtained with
> such an objective, it clearly shows that the method employed is
> diffraction limited and hence seems to be comparable with COL. From
> my experiments, however, I cannot confirm that DLC works as well as COL.
> Gregor

Of course, RRR and friends are just around the corner. I only hope that this does not trigger our yearly episode about superresolution with standard compound microscopes ;-)

Thinking about oblique illumination, when using an objective with an NA of 0.70, how much would be its "effective NA" with oblique illumination? Oblique illumination moves higher-order diffraction patterns into the lens and hence "increases" the lens' light gathering capabilities for higher orders. From a simple experiment, I tend to think of an increase of as much as 0.15 (NA 0.70 almost "becomes" NA 0.85 with oblique illumination). Hence oblique illumination together with a decent blue filter and a recording device that is sensitive enough to record waves at 450nm (blue), Frustulia rhomboides should be partially resolved with a good objective that has a "brightfield performance" of just 0.70. - Just a thought.

Merry Christmas!
Gregor

Hi,
I am new here so perhaps a word of introduction is merited. I am recently retired and have taken up microscopy as an amateur. I live in the UK.

Barry Piekos' DLC is intriguing and certainly gives good contrast. I am sceptical about claims for enhanced resolution. I suggest that those hearing those claims fall into one of two camps: those thoroughly understanding optical theory who thus "know" that resolution cannot be pushed to the level Barry suggests and hence are not interested in pursuing the matter further and those rooted in empiricism who take the view that theory must be tested by observation. I belong to the latter group. I fully accept that observations such as those others and I present may well be explicable by current theory and there may be sound arguments to demonstrate that these are optical artifacts. Yet I firmly believe that the ball is now in the court of the theorists to explain the observations.

I have uploaded three images into a photo album called "Curiosities". These are taken from a Klaus Kemp diatom test plate. The optics employed were as follows:
- Zeiss fluotar 40x phaco 0.75 NA objective.
- Zeiss DIC/phaco 1.4 NA condenser. This does not have a true brightfield position but the subcondenser polarizer and the analyser were removed from the optical path.
- Nikon Coolpix 4500 with Lietz compensated eyepiece as relay lens.
Very crude DLC plate which has reflective surfaces (cardboard cut out covered with cooking foil). I had thought of sooting it but didn't want the mess around the optics.
- Green filter.
- Camera set to B&W imaging.
- All diaphragms fully open.

The putative details on the diatoms were visible in single shots but, as when enlarging images to the extremes, it was questionable whether they truly differed from noise. Thus, I took the arithmetic mean of four images using 32 bit arithmetic in Image J. These pictures are not ideal as image averaging was not my initial intention. The photos were taken at slightly differing focus because I was seeking one good image. However all four showed much the same with the reservation noted about noise.

The images have not been further processed beyond using the built in simple contrast/brightness adjustment offered by Image J. I have messed around with deconvolution but this doesn't add anything helpful in this case. As an aside, I wonder what the deconvolution settings should be if DLC truly enhances resolution.

Image 1: F. Rhomboides
Image 2: Amphipleura
Image 3: Wider shot from which the two enlargements were made. This shows part of Pleurosigma on the right and helps get a feel for overall scale.

The dots on Pleurosigma were imaged in excellent contrast as would be expected from a 0.75 NA objective but these don't show up here because the image isn't displayed at adequate size in the album.

I find the F. Rhomboides image pretty convincing.

Amphipleura is intriguing. Could the apparent lines be a shadowing or offset image effect?

I have communicated these findings to Barry Piekos.

Selwyn