Haze Amplification in Virginia

[OlDoinyo]

A few days ago I had the chance to pass through Shenandoah National Park for the first time in nine years. The Skyline Drive that traverses the park usually has an abundance of appealing views. Haze is an everyday fact of life in eastern North America; in summer, visibilities frequently hover in the 5-8-km range even for visible imaging. This time, in November, it was considerably higher, about 20-30 km. the chief component of the haze, according to the National Park Service, is hygroscopic ammonium hydrogen sulfate particles from burnt coal which are suspended in the lower levels of the atmosphere. In the humidity of summer, the particles engorge with the humidity and produce heavy Rayleigh scattering of the ambient light. In the drier atmosphere of the off-season, the particles shrivel down to a smaller size and so produce much less scattering. Ultraviolet imaging, however, allows amplification of the remaining haze to higher levels than would visible imaging, and allows one to play with aerial perspective of receding ridgelines. Some altitudinal gradient in the haze is faintly visible. Some ridges as far away as 10-15 km are visible even in UV, although this does not match the clarity I have seen in the Rockies in winter. A bit of sheen from the bark of bare tree branches is also present. The images are near-monochromes; they were produced with the Sony A900, a Steinheil 50mm lens set at f/16, and a Baader U2 filter. Display intent is BGR.

 

"Blue Ridge Haze" (looking south along the main range's eastern flank:)

 

"Gunmetal Vale:"

 

"Spillover:"

[Andrea B.]

Clark, I am really enjoying these layered photographs - along with the interesting explanation of why it is always so "smoky" in this area. I had forgotten about the coal. (Note to non-US readers: The Appalachian area is an important coal mining area in the US.)

 

Anytime of the year Skyline Drive offers a beautiful drive and ample photographic opportunities. How nice to see it in UV for the first time!!

 

Just out of curiosity, is it the Steinheil which makes the vignetting or is it the BaaderU and step rings which cause it? I have had vignetting with my current BaaderU setup and would like to try a rear mounted filter setup to cure that. I just haven't had time to get it done.

[Alaun]

nice tranquil aesthetic pictures!

[OlDoinyo]

As an aside: there are multiple roads with the name "Skyline Drive" in the US; aside from the one in Virginia, one of the most notable is an unpaved road in central Utah which reaches over 3,300 meters' elevation in places and from which one may see views such as this and this.

 

Concerning the matter of "vignetting," that is actually a grab-bag term for a number of effects, which I will try to discuss below. As I often do contrast-boosting as part of my work-up, these things often get accentuated:

 

Image-circle truncations:

 

These tend to manifest themselves as a fairly abrupt transition from full image brightness to black at the edges of the image, especially in corners.

 

Primary ICT is when the image circle of the lens is insufficient to cover the sensing medium. As the Steinheil was designed to cover a 24mm x 36mm frame, this is unlikely to be the case here.

Obstructive ICT is when something extraneous to the lens (e.g. adapters or lens hoods) causes the image circle to be smaller than designed. In my case, there is no direct line of sight from the edge of the filter or stepper ring to the front element of the lens. However, I have had a bit of an issue with the inner rim of my home-made adapter clipping the corners of some of the frames--I need to carve that out a bit. In some cases, filter hardware could cause this; I do not know the geometry of your set-up, so I cannot comment more. You see a bit of that in some of these frames, but that is not primarily what is evident.

 

Optical falloff effects:

 

These manifest themselves as a more uniform decrease in brightness from center to edge. Filters have nothing whatever to do with these effects.

 

Cosine falloff is a fundamental property of all rectilinear optics, and especially wide-angle optics. As the incidence angle of the incoming light departs from the axis by angle a, the effective light-gathering area A of the front element changes approximately as

 

A = A₀ cos a

 

which means that image areas farther away from the central axis will be progressively dimmer. A little bit of what you see in my images may be attributable to cosine falloff, though a 50mm lens would not be expected to show this effect strongly.

 

Well falloff is a thing peculiar to digital sensors--film does not exhibit it. Each sensel of a digital sensor is in effect a tiny well, with a microlens and a reseau filter in front of the sensing surface. As the light departing the rear of the lens arrives at increasingly off-axis angles, less and less of the incoming light strikes the sensing surface directly, and more losses occur when the light bounces off the sides of the well, making the sensing process less efficient. Modern normal and wide-angle lenses have extra oversized collimating elements in the rear to help ameliorate this issue by reducing the departure angle; but old film lenses such as the Steinheil do not, and as it is a slow lens with a tiny rear element, the angle of departure to the corners of the frame is relatively steep. I think most of what you see in my images is most likely well falloff. Longer telephoto lenses have naturally small departure angles and do not suffer appreciably from this issue.

 

Filter-specific effects:

 

Filter cosine falloff stems from the consequences of Beer's Law, which describes the absorbance A of a substance at concentration c in a light-path of length l:

 

A = εcl

 

where ε (the extinction coefficient) is a property of the absorbing substance at a given wavelength. Note that as light traverses a filter of thickness l₀ and refractive index i at angle a from the perpendicular,the path length is given by

 

l = l₀/cos (arc sin((sin a)/i))

 

which makes the filter effectively thicker toward the edge. How important this is depends on the actual shape of the filter's transmission curve; one would expect some darkening toward the corners, but my use of the Baader filter with film cameras has shown little of this. Whether this effect (if important) could be reduced by moving the filter to the rear depends on lens geometry: if the angle of projection from the rear lens element to opposite frame corners is less than the lens's field of view, some benefit might accrue; if the reverse is true, moving the filter would actually amplify the effect.

 

Dichroic angle degradation results from the fact that dichroic coatings function optimally at their designed angle (usually perpendicular for filter applications) and less so at angles diverging from such. For the Baader filter, the net result would be increasing IR and visible leakage toward the edge of the frame, resulting in brighter corners (were this the only effect.) The same considerations as above apply to the question of moving the filter from front to rear.

[Andrea B.]

I think most of what you see in my images is most likely well falloff.

 

Very interesting!

 

And thank you for the review of how vignetting may be produced. I will try to use the proper terminology - so cool to know it! I was aware of all this, although I did not have good phrases for it. I do know that on some of my lenses I am experiencing OICT because it goes away when I use a slightly different setup for the 48mm Baader filters. "-)

• Primary Image Circle Truncation
  
• Obstructive Image Circle Truncation
  
• Cosine Falloff
  
• Well Falloff
  
• Filter Cosine Falloff
  
• Dichroic Angle Degradation
  

This is one of those excellent posts which should be split off into its own subject, yes?

[Shane]

Nicely compiled "vignetting" explanation Clark