A first comparison of the uncoated 2.05mm thick Andrea-U filter with the Baader-U2 filter (2", new type) brought the following results:
I) Transmittance
A spectral scan results in the following transmittance graph (@ zero and 45 degrees angle of light):
The results of the Baader-U2 filter is also shown, to be able to compare both filters. Andrea-U has some leakage into the violet/deep blue to about 410nm, but there is no sign of IR-leakage (up to 850nm).
II) Reflected UV Photography Test
(using UV sensitive camera, Cerco 94mm lens; Rudbeckia hirta as a target, Xenon flash; identical settings for aperture, flash energy, exposure)
- leftmost: using Baader-U2 filter, using my standardized "standard false UV color palette"
- rightmost: using Andrea-U filter, using my standardized "standard false UV color palette"
- middle: using Andrea-U filter, using a "special false UV color palette" (developed to match the results when using the Baader-U2 filter)
The resulting Andrea-U image has an exposure which is about 1/4 stop less that that of the Baader-U2. There is no sign of IR-leakage (up to 850nm); the somewhat warmer tone of the middle image was caused by the special palette used. All images were processed identically otherwise.
III) Summary
The Andrea-U filter is mechanically well made (the description is printed on foil and taped on the rim, which might eventually come off after long use - I would have preferred it printed or engraved of the metal, but that would certainly have increased cost).
Personally I would rate it a quite useful filter for the 375nm range, especially if used with older "normal" UV transmitting lenses since in the 375-400nm range it has a higher transmittance as the Baader-U2 filter (25% higher = 0.5EV at 390nm).
IV) Incorrect maker statements
The maker of the Andrea-U filter states that "The leading Venus filter (i.e. the Baader-U filter) is a dichroic filter" which is not correct. It is further stated that "This sensitivity to the angle of incidence (AOI) of the light results in some ultraviolet light being rejected that should have passed to the sensor... in other words, a false representation of the object being photographed in the ultraviolet is created." Also that statement is not correct and I will in the following course prove why:
The Baader U filter is an ionic colored filter that defines its properties mainly and it has a two sided dichroic overcoating that only has to block unwanted VIS and NIR leakage. I measured the Baader-U filter at 0-degrees and 45-degrees and my findings were 1) a decrease in transmission of about 0.2 stops (that effect also any pure ionic colored filter has (I show UG11 as an example and also show that under I) above for the Andrea-U filter), as the transmitting light has to pass simply a thicker glass passage - although the Andrea-U filter also shows that effect, it is not mentioned on the maker's site) and 2) a widening of the FWHM of about 12nm.
Both results are irrelevant for practical photographic work based on my experience.
Baader-U (ionic colored filter, with dichroitic overcoating to block NIR leakage):
UG11 (ionic colored filter, that has NIR leakage):
I don't understand why these possibly misleading statements were made by the maker, as his Andrea-U filter is a good one and certainly finds a use for reflected UV photography. But I leave any interpretation of that to the interested reader...
Stay tuned, more will follow on that fascinating subject...
More info on this very interesting field may be found on my site http://www.pbase.com/kds315/uv_photos