Over the last six months or so I've tried many combinations of slits, straight edges, curved edges, offset phase rings, offset dark field condensers, intermediate lenses and so on. The results have sometimes been interesting. But to post all the videos I've made would take a huge amount of disk space so I've kept them to myself.

But I strongly suggest that a curved edge in the filter holder of the condenser* -- Phase or Bright Field -- can yield quite interesting and sometimes spectacular images. I find it more useful than off-setting the condenser and it's comparable to using an intermediate lens to guide the cone of light below the condenser, and, made from old floppy disks the inserts cost nothing. The method gives very good results with high magnification and I now use it exclusively for imaging the lines (and sometimes dots) of A pellucida.

* The Lomo Aplanatic with it's mechanism for rotating and offsetting the diaphragm is very useful for this. I leave the Iris wide open and use rotation and offset to move the knife edge.

Don W

Don,
Were your curved edges convex or concave? Could you show photos of the stops that were most successful? Also could you post some of the more spectacular pictures perhaps with comparison to other techniques?

Did you find any evidence of the enhanced resolution claimed by Piekos?

Selwyn

Selwyn,
I have been accumulating information (and images) available from DonW and also other microscopists that have independently been using the "DLC technique" with some very interesting results. With whatever time I have available over the next few weeks I would like to work with Don to use his videos as part of a webpage presentation of "DLC results".

This technique requires a convex edge plate(stop).

Enhanced contrast - yes. Enhanced resolution - maybe, and only in very special cases.

Btw, I finally got permission to publish the complete Axelrod paper describing his 'schlieren microscope':
  Link : http://www.kscitech.com/MicGroup/Temp/Axelrod/

Regards, Keith

Hi Don,
Could you put a couple of pictures in the photo section of the floppy disk inserts and their measurements.
I would like to give it a try.

Dan

Hi Don,
I have employed a similar shape but placed on the glass of the field light with the iris open. I also use a sort of fish hook shape which provides both convex and concave edges. I can then move this around by hand. I just cut the shapes out of thin plastic.

Salutations, Tim

Done. They're in the Don Williams photo album.

Also take a look and see how much desk space I have. Some of the stops go into the filter holder, others are taped (two tiny pieces of double sided tape) to the top of the filter holder. I had more stops and some slits but can't find them right now. Maybe they've been thrown away.

Don W

"Daniel Holloway" wrote:
> Could you put a couple of pictures in the photo section
> of the floppy disk inserts and their measurements.

Thanks Don, by the way aren't you up a bit early this AM.

Dan

I have added in the "files", "roseoptics" folder, a copy of a paper on Optical Shadowcasting of Living Cells. The techniques are very similar to those in this discussion.

Arthur

"scitech200" wrote:
> Btw, I finally got permission to publish the complete Axelrod paper
> describing his 'schlieren microscope'...

Keith,
Thanks for the Axelrod paper. I really look forward to seeing the presentation on DLC.

Selwyn

Don, not to be nosey, but I am interested in less-common microscopes. I think I recognize a Leitz Ortholux II on your back table, but what model is the handsome big microscope taking up your desk? Is it a research version of a Lomo?

Thanks, Greg

Hi Don,
I am always amazed at the images you get with web cam. It truly shows that you don't need high resolution on a microscope at high magnification. The photos are at:
  http://tech.ph.groups.yahoo.com/group/Microscope/photos/browse/586c
It saves a little time paging though to them.

Are you seeing anything beyond the resolution you expect with a microscope that could be resolution beyond the 1000 times the numeric aperture Abbe postulated?

Have you tried a stop that projects the same curve as the phase ring on the phase ring of Zernike phase objective? I some conversation some where the discussion was on Zernike phase contrast and crosses seem to be better for straight lines such as are found in semiconductors and rings better for cells and other curved biological features. I expect the same physics apples to oblique illumination. In the same conversation it was mentioned that a phase objective could be made by painting rings or crosses on the rear element of some objectives with shellac. I expect some modern plastics would work as well. I believe the rings or crosses should retard the light 1/4 wave length of 556nm green light. But if rings of shellac work it can't be to critical. Could it be that the curved oblique stop is just better for subjects that are curved than the straight and angular oblique stop historically used? Might that not explain some of the reason for the good performance of COL as well?

Using dull side floppy disk to the light addresses almost all of the failings I found with metal and plastic stops caused by reflection. I would like a thiner edge but that has its own problems of being to fragile or needing to be supported by glass that introduces two more reflective surfaces to reflect any problems cause by the stop 2 more times.

Gordon

Yes that's the big Lomo. I've not seen another anywhere. I've asked people around Finland if they've seen, or have any, but the answer has been in the negative. It's a fine solid stand but needs a few parts for the fine focus mechanism to make it perfect. I'd need to make them and don't have the facilities. But in general the microscope is good.

The thing in the background is a Diavert -- an incomplete upside down microscope. It has no phase rings for the condenser and is only useful in bright field. Also -- it needs 170 mm objectives of which I have only two. But I use it now and again and there's nothing much wrong with it.

Don W

I forgot to mention that the stage on the Lomo is Leitz. The same as that used on the Diavert and Diaplan. Also the lighting system has been altered to take a 100W halogen in place of the original filament lamp socket. A few other things have been added. The frame provides both transmitted and incident lighting.

Don W

Hi arthur,
Any chance of getting that paper in some format other than .mdi which can only be opened by MS Office 2003 or later which I don't have.

Ted

Ted,
If it helps, you can get a free reader at
  http://www.hot.ee/mdiviewer/

Regards, Tom W.

Hi Gordon,
Floppies are made of Mylar (I think) and are thin and flat. It's very easy to cut a decent edge with a sharp pair of scissors. I suppose a craft knife would do as well but I like scissors and the resulting edge is always clean. I haven't tried very small radii (the size of the phase rings) simply because it hadn't occurred to me up to now. They'd have to be punched anyway and I don't have the tools. But by displacing the phase ring of the condenser to give a bright semi-circle of light -- oblique phase -- quite impressive results can be achieved. I tried it on a few amoebae, but I don't think any more detail was revealed. The best images I have of amoebae (internal organelles) were made on film (B&W very grainy and messy) they showed good detail of the nuclei. I also used C42 and they're not too bad -- but also very grainy.

I've removed some old pictures from the Don Williams photo album and added one more -- showing how really cluttered things can become here.

The Webcam has become an essential part of my setup. I have a lot of trouble these days seeing fine detail through the binoculars. There are floaters in my eyes the size of a pickup truck. I need to find a way of fitting one of these cameras to one side of the binocular so that I can put the Pentax on the top tube and still be able to see the image. I have a very light camera that gives up to 1280 x 960 at 15fps and once I get an adapter made it will make things easier. But I can't make it on the school lathe because it requires boring in a rather smaller diameter than I could manage. The school machine is nine feet long and has a swing of 24" and it's also rather buggered. The smallest boring bar is about an inch in diameter!

Don W

The paper about shadowcasting posted by Arthur (roseoptics) raises many fascinating questions about why this and related techniques pop up in the literature from time to time and seem soon to be forgotten in the mainstream where phase contrast and DIC have become dominant. The paper below cites the shadowcasting paper. It is of interest because the authors present many images of the same fields in phase contrast and shadowcasting. The comparisons beg questions and I should be interested to hear the views of professionals in cell biology in this group regarding the relative merits of the two forms of contrast with respect to the features they elucidate and artifacts they may introduce.

Giraldo et al. Human sarcomas in culture.
  http://www.jem.org/cgi/reprint/133/3/454.pdf

Selwyn

For anyone that can not yet access the .mdi file:
  Link : http://www.kscitech.com/MicGroup/Temp/ShadowCasting.htm

-Keith

"selwynstleger" wrote:
> The paper about shadowcasting posted by Arthur (roseoptics) raises many
> fascinating questions about why this and related techniques pop up in
> the literature from time to time...

Thanks for posting that. I wanted to post that as a PDF, but was having problems with the computer/scanner.
Sorry if it caused problems with some members.

Arthur

"scitech200" wrote:
> For anyone that can not yet access the .mdi file:
> http://www.kscitech.com/MicGroup/Temp/ShadowCasting.htm

Hi Dan (and all other DLC experimenters),
I would be very interested to see what the results are with a 'DLC stop' as shown in Fig.1 :
  Link : http://www.kscitech.com/MicGroup/Temp/DLC_stop.htm
for some protists.

The stop (18mm diameter) is cut from black masking tape and is stuck to a blue filter. A dime happens to be a ready made template for the ol' xacto knife to follow! The sketch shows a setup whereby the stop is moved on an arc by swinging the filter holder. But you have the more sophisticated Lomo setup, so you can very precisely move the stop left-right.

Have fun!

-Keith

"Daniel Holloway" wrote:
> Hi Don,
> Could you put a couple of pictures in the photo section of the floppy
> disk inserts and their measurements.
> I would like to give it a try.

Dan

Maybe it's just me, but it's getting confusing to follow the threads on some of these lighting techniques. In some cases, there seem to be quite different methods being lumped together.

First there is "oblique" illumination (or "off-axis" illumination) which has been around forever. Here the condenser aperture is off-set from its centered position, and the subject is seen in some relief. The effect can be striking. All manner of differently shaped "stops" can be placed at or near the position of the condenser diaphragm and a variety of effects are obtained. There are numerous references that discuss this tried-and-true technique.

The Mathias Arrow is one of the more novel oblique methods, and in an excellent article Walter Dioni discusses it here:
  Link : http://www.microscopy-uk.org.uk/mag/artoct03/wdoblique.html

The article that Arthur recently made available to us here:
  Link : http://www.kscitech.com/MicGroup/Temp/ShadowCasting.htm
by Hlinka and Sanders, "Optical Shadowcasting of Living Cells" appears to be discussing relatively straightforward oblique illumination.
--------
Then there is the quite different technique that, in the Daniel Axelrod paper, is referred to as schlieren illumination:
  Link : http://www.kscitech.com/MicGroup/Temp/Axelrod/
While this method also employs an opaque edge stop at the front focal plane of the condenser (somewhat like oblique) a second properly oriented opaque edge stop is required at the back focal plane of the objective.
It is well explained in the article. It would appear to have some similarities with Hoffman Modulation Contrast. This page is also good reading to get a better grasp on modulation contrast:
  Link : http://www.olympusmicro.com/primer/techniques/hoffman.html

A very detailed paper that was also brought up here a while ago:
  Link : http://www.opticsexpress.org/DirectPDFAccess/....
by Yi, Chu, and Mertz, "Graded-field microscopy with white light" would probably be a close relative to this technique. Also very much worth reading (but with this ridiculously long URL it might take some careful cut and pasting to get there!)
---------
Then there seems to be a third approach outlined in a paper by Barry Piekos, "Diffracted Light Contrast":
  Link: http://pantheon.yale.edu/~wbp2/DLC.pdf
This might be the article that got this discussion going, since the claims, in the title, of: "Improving the Resolution of a Basic Light Microscope by an Order of Magnitude" seemed to raise a few eyebrows.
---------
It's great to discuss all this, but it seems we should be careful to differentiate these methods...
"oblique", "schlieren", and "DLC"(as per Piekos).

Some may feel the "Graded-Field" discussed in the above reference is sufficiently different to be placed by itself.

Hi Charlie,
You are right, a discussion of illumination techniques can get to be very confusing. That's why I made an attempt to present some information in a summary format via 'The Microscope Group Archives':
  Link : http://www.kscitech.com/MicGroup/Dir_01.htm,
in particular for all topics related to the Piekos DLC technique.

Of course hobby time sometimes delays keeping this completely up to date! For instance, a preliminary webpage:
  http://www.kscitech.com/MicGroup/Msg_14

However, it seems that the majority of the techniques that have been under discussion can indeed be accurately described as some variation of the "old" and well proven 'oblique illumination'.

Regards, Keith

"Charlie" wrote:
> Maybe it's just me, but it's getting confusing to follow the
> threads on some of these lighting techniques.

In a previous posting I alluded to the curious manner in which deviant contrast techniques (i.e. those not widely used now by researchers) pop up in the literature and then get forgotten. These are oblique, COL, DLC and combinations/generalisations thereof. I offer some further comments and speculations.

We all know that light microscopy was brought to near perfection by the end of the Victorian era. Subsequent developments (e.g. phase contrast, DIC, and confocal microscopy) are incremental changes but not revolutionary (c.f. the original Ford motor with the present day car.)

The matter of contrast is very interesting. DF, oblique, (probably) COL and other fairly cheap techniques were all known in the early days. Yet,those interested in visualising living cells hailed phase contrast and, later, DIC as the ideal (setting aside fluorescence which is not just about optics.) It seems as if the simpler techniques were largely ignored even though several authors demonstrated their worth. Why was this?

One explanation (tentatively suggested) is that the increased contrast found by sticking a finger (or pencil) between mirror and condenser was distrusted. Everyone knew (or was taught) that the effect of shutting down a diaphragm to increase contrast in bright field introduced diffraction artifacts, so why trust other "dodgy" techniques?

Another explanation is best expressed in terms of the philosopher of science Thomas Khun's idea of "paradigms". Within what Kuhn termed "normal" science there is a paradigm that encapsulates the following: the area of study, methods of research thought appropriate, key questions to be researched, how research and researchers are to be valued, etc. I suggest that mainstream techniques of contrast enhancement (phase and DIC) became valued within the paradigm of cell biology more then others.

A third explanation relates to the ease of use of the various techniques. Following simple rules anyone can set up phase contrast or DIC correctly. Most graduate researchers and faculty members want to get on with their observations as easily as possible. They are not interested in the microscopal technology per se. Why should they be? The non-standard techniques (oblique, COL and now DLC) require a lot of fiddling about and skill in setting up the optics correctly; also they are not so readily reproducible as the likes of phase contrast.

This leads to the economic imperative. Manufacturers are sensitive to market needs. Researchers want reliable, easy to set up, microscope configurations for various purposes. Profit margins for high tech solutions are greater then those for simple solutions. Researchers get what they want (ease and reliabilty of use) and manufacturers make the profits the market allows (perfectly fair). I understand that modern high end 'scopes do everything for the user (focusing, etc.) apart from making the tea during the work break. Fair enough, a research lab wants productivity.

That does not mean that simpler contrast techniques are inferior; which is why I am fascinated by the resurgence of interest in oblique illumination, COL, and DLC (if it is new).

Selwyn

Selwyn,
Your post was most enjoyable on this weekend.

107 years after a particular meeting of  'The Quekett Club'  we are discussing variants of 'oblique illumination'.

Here is a glimpse of a discussion following an address given by A.A. Merlin F.R.M.S., "On the Resolution of Amphipleura pellucida with a Dry Lens and Axial illumination" (November 16th, 1900):
"Abbe's law, which was based on mathematical considerations admitting of no controversy, declared that, with the smallest possible beam of truly axial illumination, the number of lines to the inch capable of being resolved = lambda x NA where lambda is the number of wavelengths to the inch of illuminating light actually used. Putting this into actual figures, seeing that there are about 47,500 to the inch for visual light, it was found that NA = 0.95 gave 45,125 lines to the inch as the theoretical limit. But with oblique light this formula was doubled and became 90,250 to the inch. It was evident, then, that Mr. Merlin could not have seen the lines [of A.pellucida] without oblique light."

It deals more with resolution, than contrast, but the complete address and following discussion is a great read and will be available at:
  Link : http:/www.kscitech.com/MicGroup/Temp/Quekett_01.htm,
courtesy of a group member across the pond from Masachussetts.

You have to wonder about how fulfilling it must have been to be able to sit back, think about newly available experimental data and then discuss such fundamental topics - whereas today we sit at a computer and rely upon MathCAD (or whatever) software to resolve some menial, incremental aspect...

Best regards, Keith

PS. The Quekett Club  (founded in 1865) website is at:
  Link : http://www.nhm.ac.uk/hosted_sites/quekett/

"selwynstleger" wrote:
> We all know that light microscopy was brought to near perfection
> by the end of the Victorian era. Subsequent developments (e.g.
> phase contrast, DIC, and confocal microscopy) are incremental
> changes but not revolutionary....

Ted, Keith, Arthur,
Here are some more papers on Shadow Casting.
  Link : http://www.jem.org/cgi/reprint/133/3/454.pdf
  Link : http://jcs.biologists.org/cgi/reprint/12/2/453.pdf

Gordon

In order to demonstrate the various methods of oblique illumination I've started making a series of videos.

The first, using a convex knife edge (after Piekos), is on science-info.net:
  Link : .../pages/EDFWilliams/VIDEO/test_DLC_01.avi

[ Right-click on 'Link' to download the video.
The DivX codec (free) is required for viewing this video. ]

The specimen is a three month old Vaseline prep from a mixed algal/protist culture. The video starts with the knife edge out of the field and it is moved across slowly, allowing the camera to compensate for the loss of light. There is a point just before the wide colour fringe appears where the image contrast is maximized. If the condenser is raised right up oblique dark field can be achieved.

While the final image through the binoculars is completely black, the camera is sensitive enough to see the faint light that gets through the mylar! The next video will use the same device, but at high magnification.

Don W

Don,
An outstanding demonstration movie!

Once again, thank you for sharing your work. I have updated the summary webpage:
  http://www.kscitech.com/MicGroup/Msg_14/

One question: what was the diameter (approx) for the convex edge?

Best regards, Keith

"Don Williams" wrote:
> In order to demonstrate the various methods of oblique
> illumination I've started making a series of videos....
>
> The convex knife edge (floppy disk fragment)...

In order to demonstrate the various methods of oblique illumination I've started making a series of videos. The second, using the same convex knife edge as used for the first test (after Piekos), is on science-info.net:

  Link : .../pages/EDFWilliams/VIDEO/test_DLC_02.avi

  Link : .../pages/EDFWilliams/VIDEO/test_DLC_02f.avi

[ Right-click on 'Link' to download video. ]

The specimen is a strew of A pellucida -- KK material. The video starts with the knife edge out of the field and it is moved across slowly, allowing the camera to compensate for the loss of light. There is a point just before the wide colour fringe appears where the image contrast is maximized.

The second shorter clip is a through focus clip with the knife edge at or near its optimum contrast position.

Don W

I used the outside of the 3.5" floppy disk as it comes out of the housing. The radius is 42.65mm.

Video of test 02 is ready and waiting. I'll probably take longer to do the others as I have a few chores in and around the forest.

Don W

"scitech200" wrote:
> One question: what was the diameter (approx) for the convex edge?

Don,
How do you think it compares with a properly illuminated image?

Gordon

"Don Williams" wrote:
> In order to demonstrate the various methods of oblique illumination I've
> started making a series of videos. The first, using a convex knife edge
> (after Piekos),...

Thanks Dr. W, and thanks to all of you who share so much.

I use 'scopes both professionally and as a hobbiest.
The discourse here is the most valuable that I have encountered: always my first reference!

interlofer

Barry has been following the group discussion in Topic summary format:

"What a pleasant way to start the week; I'm delighted to see Don's images.

I do have a few comments/suggestions. Don appears to be using the filter holder that's attached to his condenser. I'm still not sure if the condenser is at Koehler or not. As I state in my paper, the increased resolution I obtained was, at least in part, due to the fact that condenser was not at the Koehler position. When I perform the technique, with the edge plate near the level of the field stop and physically separated from the condenser, I'm forced to alter the position of the condenser to obtain the broad band of diffracted light. Also, I noticed that he was stopping down his field stop. When I perform the technique, I keep all apertures fully opened so as to minimize other sources of diffracted light. In order to truly replicate my results, Don (or anyone else) should be performing the technique as I taught it. It may be that on Don's scope, the configuration he's using is best, but I think it may be interesting to compare his approach to mine."

Don, in your estimation, is it fair to state that using the convex edge stop mounted on the Lomo filter holder gives enhanced contrast results that are superior to all other configurations you have tested? My understanding is that any improvement in resolution is a seperate issue. Your setup is relatively easy to implement, with the edge stop in close proximity to the aperture diaphragm, but is distinctly different from Barry's preferred configuration as described above.

Regards, Keith

PS. I have updated the current Topic summary at:
  http://www.kscitech.com/MicGroup/Msg_14

"interlofer" wrote:
> Thanks Dr. W, and thanks to all of you who share so much.

Hi Keith and Barry,
I don't fiddle around looking for broad bands Barry because I'm not actually trying to reproduce your results. I'm using the edge to enhance contrast with oblique light which is all it does for me.

I forgot to mention that I always set the microscope up for Köhler before I begin. The field stop is opened to fill the field of view in the traditional way -- the way I always set things to start with. I have tried focussing the filament image on the knife edge itself -- instead of using the telescope to focus it in the back plane of the objective -- and below and above (normally I focus it on the condenser iris) but neither procedure seems to affect the contrast with the specimens I'm using. And this goes for the A pellucida test as well. I'll try focussing the lamp back and forth after setting up for maximum contrast with the video camera running and see what that does. But I really don't think it'll make an awful lot of difference.

By the way the knife edge (floppy disk fragment) is about 8mm or so below the condenser iris.

I've already tried reproducing your method to resolve the dots on A pellucida which, according to what you've found, ought to be possible because it's supposed to improve resolution enormously. Yes? It doesn't help me to resolve the dots. But with certain specimens lying at the right angle I can see them here and there just using the knife edge to obtain oblique light. But I can try again and will. With this microscope I can image the filament anywhere I want it.

Using a knife edge (floppy fragment) instead of an intermediate lens, or shifting the condenser iris, to obtain oblique light is easier and quicker than anything I've used before. I can also move it much further than I can the intermediate lens or iris. And so I would say that it does give the best results so far.
But its a long time since I tried a whole lot of other stops, points, wedges and so on so I can't be absolutely sure. One thing though -- I'm not about to start that all over again. The knife edge works nicely as does an intermediate lens.

There is one of the pointy things in the picture I posted of the knife edges by the way towards the side. It was in the box with the others.

Don W