Nevertheless, for conventional microscopy, the information from the sample resides in the diffraction from the samples structure. In transmission microscopy, the diffracted information competes with unscattered background light that greatly reduces contrast in the absence of staining. Why is dark field microscopy not the standard approach?
There were several responses so far, but they do not really answer my question. One response is:
>>>>>>>> An expansion upon your second paragraph may help here. The finer structures diffract the light to a greater degree off the incident (illuminating) path. Imagine a ray that strikes a cell membrane. Some of the light is attenuated but passes on undiffracted - this is the 0th order light. finer and finer structures causes the same incident ray to be diffracted more and more. Many of these rays are collected by the objective lens. The more diffracted rays travel a longer path. At the intermediate image plane (where the reticle would be inside the eyepiece) however these rays recombine causing interference. The detail we see is the result of the interference of all of the diffracted and undiffracted light. Without these different waves competing as you call it, most specimens, even unstained ones would yeild little info. Darkfield throws away all except the most diffracted light accentuating only the finest structures at the expense of moderate sized features. <<<<<<<
This, more or less, agrees with what I said and also suggests that dark field microscopy can reveal detail without the zero order light dazzling the viewr or otherwise reducing contrast. I will disagree that moderate sized features are suppressed with dark field microscopy. What happens is that the larger features have small diffraction angles. As a result many more diffraction orders are produced. These are collected by the objective. Because of the small diffraction angles, a smaller objective allows information about these larger features to obtained. In ordingary transmission microscopy, however, the first order is also collected. In dark field microscopy that large first order from the light source is eliminated, but the orders containing the information are collected just as in transmission microscopy.
Another response was:
>>>>>> Darkfield microscopy is good for small things that scatter well but don't absorb. If there are many small things it becomes hard to tell them apart. Darkfield does not use the full numerical aperture available since the objective must have smaller NA than the illumination system. I think darkfield makes pretty pictures. I have noticed that people who want to see small things that don't absorb tend to use edge detection optics such as differential interference contrast, Hoffman contrast, or single side-band egde enhancement. These methods are combined with video for contrast enhancement. SSEE and DIC use the full available aperture at least in one direction. The original Hoffman didn't use the full available aperture but may have evolved to do so by now. It turns out that scattered light and diffracted light are essentially the same. Darkfield makes no use of the phase information in the signal. The other methods mentioned apply interferometry to generate contrast based on refractive index gradients. SSEE optics can be configured to yield contrast based on refractive index differences similar to phase contrast. <<<<<<<<
Most of the techniques described use more than just simple transmission. Although many dark field systems may use some of the solid angle available to collect information for illuminating the object, that is not truly necessary. Edge detection, I would guess, avoids the use of low orders of diffraction, including the zero order, and uses the higher diffraction orders. Phase contrast and interference microscopy, useful as they may be, fall outside the domain of my question.
For simple microscopy, no phase information is used directly either except for the recombination of orders. There is no fundamental reason why phase information could not be used in dark field microscopy. There are Fizeau interferometers using reflected light (sort of like dark field) and Twyman-Green interferometer for transmitted light.
William Buchman