Darkfield Illumination

[garyjm]

I'm wondering if anyone has any experience with using a darkfield stop inserted into the existing condenser filter holder? How does this compare with using a dedicated darkfield condenser? Do they both achieve the same result? Advantages/disadvantages for each? You can see some discussion of this in the following video: https://www.youtube.com/watch?v=JnjoHUIlnzM&t=2s

Thanks,
Gary

[apochronaut]

All DF condensers or condensers modified to produce DF have limitations, so understanding those helps one to choose which method is most suitable. Obviously, using stops and an existing available condenser is the cheapest option.
Stops are o.k., even good with up to a 20X objective and possibly higher if they are really precisely made, you have good xy centering control over them and your condenser is perfectly axially aligned. With DF stops it is more the magnification and precision that determines their useability because the stop is a separate device from the condenser and needs accurate alignment to an accurate condenser to function properly. . Image quality wise, more highly corrected refracting condensers will give better DF but still not as good as a DF condenser. The limitations of stopped refracting condensers for DF are a combination of magnification and N.A.,with higher of each imposing increasing difficulty of set up.

Since DF condensers have precision built in stops, the N.A. not so much the magnification becomes the main arbiter of their potential, although lower magnification can impose a field coverage limitaton on them. . Most DF condensers are mirror condensers, so limit the aberrations to varying degrees based on the actual type of mirror system.
Dry DF condensers can be used up to about .70 N.A., so in practice are limited to 40X objectives or less. Dry DF condensers can cover a wider field , so can be used down to lower magnifications than oil DF condensers. Many down to 5X.

Oil DF condensers cover a narrower field , so usually can only be used from 40X objectives and up. Some designs are a wider field, so they can be used from between 10X and up.

High magnification DF requires a lot of light, so anything over a 60X objective requires an illumination system designed for high resolution DF or modified with that in mind. Oil DF condensers generally provide higher image quality than the two former systems.

All systems require that the objective being used have an N.A. lower than that of the condenser, about .20 N.A. lower but that can higher in very clean precise systems.

[garyjm]

Thanks apochronaut for the great information about this.

I have ordered some of these filters to play around with and compare with some of my dedicated darkfield condensers: https://www.ebay.com/itm/276297830862

The inclusion of the Circular Oblique Illumination (COL) Microscope Filter Set is particularly exciting.

Wow, so much to learn, so little time


Gary

[Scoper]

Excellent description Apo..thanks for posting it.

[Javier]

I have been usign home made DF filters for a few years now. I find they work fine up to 200x.

For best photographic results, set the brightness of your scope really high and leave the camera a bit underxposed.

I'm sharing some unedited examples with you. Of course, you can enhance a lot these results with some basic post-processing.

[apochronaut]

Those are very good. Curious what condenser and 20X objective you are using.

[Javier]

Those are very good. Curious what condenser and 20X objective you are using.

Thanks.

I'm using very basic stuff, just the stock condenser of an Amscope b120 (re-branded here in Argentina) and a 20 x achromat.

I noticed that you can tell right away that this is a cheap scope in bright field, but it does a decent job in dark field. I guess this happens because dark-field lowers resolution quite a bit, it is not that obvious that you are not using a high performance, high resolution scope. I remember asking Oliver once if I would get a better image making a filter with glass instead of plastic, and he answered me not to worry about that, because it was defocused light anyway.

IME the trick is to set the illumination really high (always at maximum on my 1w LED) and to set the phone camera to manual focus/exposure to avoid both noisy and burned images.

[apochronaut]

You are certainly getting your money's worth out of that scope/camera. I am curious why you say that DF lowers resolution?

[Javier]

You are certainly getting your money's worth out of that scope/camera. I am curious why you say that DF lowers resolution?

It was my understanding the BF achieved the highest resolution possible, while others illumination techniques helped to gain contrast at the expense of a lower resolution. But of course, I might be wrong. Just talking about my personal experience, DF allows my to see a some details and structures that are invisible in BF because of the increased contrast, but BF allows my to see the most fine details and structures of the specimen that are lost on DF.

I just found this old thread that will make an interesting reading very soon: viewtopic.php?f=25&t=7689&start=30

[Chas]

It was my understanding the BF achieved the highest resolution possible, while others illumination techniques helped to gain contrast at the expense of a lower resolution. But of course, I might be wrong. Just talking about my personal experience

This seems to have been the understanding and experience of most people back in the pioneering days of microscopy

See in here:
https://archive.org/details/journalofqu ... 0/mode/1up

( And pages starting 497 for W.B.Stokes and page 501 for Julius Rheinberg on the subject)

[apochronaut]

I agree with all that, Javier and you can mathematically show that DF should have lower resolution than BF but I am consistently finding fine details in DF that do not show up in BF. Part of that is probably caused by visual handicaps related to either the brains response to low light levels or possibly some differences in photopic vision compared to scotopic vision. Mostly, DF viewing would probably be characterized as mesopic vision( a blend of photopic and scotopic) but at some point with very low illumination from structures it will be scotopic, which requires a considerable adaptive time frame. I find my ability to see fine detail improves over time, viewing in DF. No room lights improves low light adaption and possibly a red light, low light adaption technique, such as is used by astronomers might be beneficial. I certainly sometimes see more in a well taken photo, than through the eyepieces but as you point out, photographing in DF requires a specific technique and I certainly don't always get there.

There is also the elusive understanding of the difference between reflected light and transmitted light in microscopy. DF begins with very coherent light transmitted to the viewed object through total internal reflection in the slide. If you use a laser as a light source, you can trace it bouncing along the inside of the slide. I can't see that the diffractive nature of the illumination beam in DF is equivalent to that in BF.

[apochronaut]

It was my understanding the BF achieved the highest resolution possible, while others illumination techniques helped to gain contrast at the expense of a lower resolution. But of course, I might be wrong. Just talking about my personal experience

This seems to have been the understanding and experience of most people back in the pioneering days of microscopy

See in here:
https://archive.org/details/journalofqu ... 0/mode/1up

( And pages starting 497 for W.B.Stokes and page 501 for Julius Rheinberg on the subject)

Those older discussions diverge into camps, each one resorting to the use of even older somewhat divergent theories of resolution, one based on self luminence, the other on diffraction. They are kind of at loggerheads, with each one coughing up an idea only of why DF can't really resolve as highly as BF. They are beliefs really.
What does come out of their discussion is that rather than DF resolution being limited by a maximum objective N.A. which is calculable as a percentage of the peak condenser N.A., there is a not very understood consistent increase in objective N.A. based on the potential maximum aperture of the objective and not strictly limited by the condenser's N.A. limit. That was shown empirically and then a theory to explain that surprising result went missing. All of a sudden we have .85 or thereabouts when Conrady predicted a preposterous maximum of .47.
The working N.A. of the objective however, doesn't provide all the answers
The fact that in the object luminence, refraction plays a significant role and I would also add reflection, is dabbled with in the discussions and by inference escapes to a large degree a resolution discussion based on diffraction theory alone.

[Javier]

It was my understanding the BF achieved the highest resolution possible, while others illumination techniques helped to gain contrast at the expense of a lower resolution. But of course, I might be wrong. Just talking about my personal experience

This seems to have been the understanding and experience of most people back in the pioneering days of microscopy

See in here:
https://archive.org/details/journalofqu ... 0/mode/1up

( And pages starting 497 for W.B.Stokes and page 501 for Julius Rheinberg on the subject)

Thanks you!

I agree with all that, Javier and you can mathematically show that DF should have lower resolution than BF but I am consistently finding fine details in DF that do not show up in BF. Part of that is probably caused by visual handicaps related to either the brains response to low light levels or possibly some differences in photopic vision compared to scotopic vision. Mostly, DF viewing would probably be characterized as mesopic vision( a blend of photopic and scotopic) but at some point with very low illumination from structures it will be scotopic, which requires a considerable adaptive time frame. I find my ability to see fine detail improves over time, viewing in DF. No room lights improves low light adaption and possibly a red light, low light adaption technique, such as is used by astronomers might be beneficial. I certainly sometimes see more in a well taken photo, than through the eyepieces but as you point out, photographing in DF requires a specific technique and I certainly don't always get there.

There is also the elusive understanding of the difference between reflected light and transmitted light in microscopy. DF begins with very coherent light transmitted to the viewed object through total internal reflection in the slide. If you use a laser as a light source, you can trace it bouncing along the inside of the slide. I can't see that the diffractive nature of the illumination beam in DF is equivalent to that in BF.

Interesting.

Never occurred to me that scotopic vision was even possible while observing through a microscope. Given the low resolution nature of scotopic vision, my first reaction would be ample the signal to get a comfortable view on photopic vision to get both a high signal, high contrast image. But maybe there is something to it. I agree that I get the best DF images in a dark room, but I always attributed this results to the absence of stray light.

My background in general science is very poor, so I cannot get technical on discussion involving nature of light and optics, but from my experience, I can tell that I don't see more in one technique or the other, they seem to bring different features and structures of the specimen.

[Chas]

The fact that in the object luminence refraction plays a significant role and I would also add reflection, is dabbled with in the discussions and by inference escapes to a large degree a resolution discussion based on diffraction theory alone.

Yes, I think that you are right.
Also most structures we look at are not repeating ( grating/rulings) and I think that is why a lot of those microscopists moved (back) towards considering the size of Airey's discs which was developed further by Gordon into the idea of 'Antipoints'.

The article that triggered Conrady's is quite an interesting one:
https://www.biodiversitylibrary.org/ite ... 7/mode/1up

Some things repeat , some things refract ,some things reflect,somethings 'light up' & some things are mounted in water.

In one way 'contrast' is at the heart of which ever resolution criterion you care to use.

[StargazerX5]

Conrady (1912) made use of Abbe's theory of image formation, which as he noted was the only theory sufficiently well developed at that time. Abbe was mostly using off-axis pencil beams investigating diffraction effects. Using that approach Conrady presented his result that the condenser must have not less than three times the NA of the objective.

However it would be much later when the theory of image formation would be better understood. Today the approach to the fuzzy concept of 'resolution' is better explained as a signal to noise (SNR) problem, as apochronaut suggested in his darkroom method.

Rayleigh (1888) wrote when discussing his result for double line resolution :
"This rule is convenient on account of its simplicity; and it is sufficiently accurate in view of the necessary uncertainty as to what exactly is meant by resolution."

Glenn

[Phill Brown]

The patch is na specific but the aperture width effects the contrast.
A narrow gap of around 1mm will improve the resolution.

[Chas]

Javier, those are very nice darkfield photos, btw

As a side note; I found Traviss's instructions for making his adjustable expanding darkfield stop :
https://archive.org/details/journalofqu ... 6/mode/2up

I cant quite follow them but maybe one needs to make a large one out of paper first and then it might all become a bit clearer.

[Javier]

Thank you, Chas!