The coatings that make these narrow-band filters work are marvels of modern engineering, sometimes dozens of layers thick. They are interference coatings, so the thicknesses of these layers must be precisely controlled to 1/4 wavelength tolerances (sound familiar?). Lumicon states that their O-III filter has an 11nm bandpass, meaning that most of the energy passed by the filter consists of only wavelengths about 5.5nm either side of the center of the band. For O-III the lines we are interested in are at 496nm & 501nm, so 11nm encompasses both nicely if centered properly. A finicky feature of interference filters is that they are choosy about the angle of incidence of the light passing through them. This makes sense, as the angle of incidence increases, the light must travel through a longer path in the coating material, thus interfering at a longer wavelength when rebounding. There is a chance that Rich's filter is optimized for a fast Newtonian (at a higher angle of incidence)and Jims is happier at a longer focal ratio such as a "stock" SCT might work at (lower angle of incidence). Food for thought. Chuck __________________________________ Do you Yahoo!? SBC Yahoo! DSL - Now only $29.95 per month! http://sbc.yahoo.com
Chuck, The scope we tested this in has 100 inches of focal length at f/6.25; Jim's 10-inch I believe is f/7 or more which he also had no success with; I wouldn't call either of these scopes particularly fast. On the other hand, maybe we should try the TeleVue filter in a faster scope, say an f/4.5 and see what happens. But the bottom line was this -- the nebula was dramatically enhanced by the Lumicon filter, and only barely so (if at all) by the TeleVue, made even more obvious by my filter slide (both filters were loaded at the same time, and it was a simple matter of sliding each under the eyepiece (35mm Panoptic) without taking my eye away from the image. There was little if any difference between the TeleVue and the view sans filter. My Lumicon did come with transmission specs (in fact, I bought it from a dealer who had 3 of them sitting on the table, and was able to pick out the "best" of the 3). Jim, did your filter come with similar information? It would be interesting to repeat this test with other filters and/or filter manufacturers (anyone out there with a thousand oaks O-III in a 2-inch format?) and in different telescopes of varying focal lengths. -Rich --- Chuck Hards <chuckhards@yahoo.com> wrote:
The coatings that make these narrow-band filters work are marvels of modern engineering, sometimes dozens of layers thick. They are interference coatings, so the thicknesses of these layers must be precisely controlled to 1/4 wavelength tolerances (sound familiar?). Lumicon states that their O-III filter has an 11nm bandpass, meaning that most of the energy passed by the filter consists of only wavelengths about 5.5nm either side of the center of the band. For O-III the lines we are interested in are at 496nm & 501nm, so 11nm encompasses both nicely if centered properly.
A finicky feature of interference filters is that they are choosy about the angle of incidence of the light passing through them. This makes sense, as the angle of incidence increases, the light must travel through a longer path in the coating material, thus interfering at a longer wavelength when rebounding.
There is a chance that Rich's filter is optimized for a fast Newtonian (at a higher angle of incidence)and Jims is happier at a longer focal ratio such as a "stock" SCT might work at (lower angle of incidence).
Food for thought.
Chuck
__________________________________ Do you Yahoo!? SBC Yahoo! DSL - Now only $29.95 per month! http://sbc.yahoo.com
Half an f/stop could be enough of an angle deviation to shift the central bandpass several nanometers, effectively killing the filter for that wavelength. What I'm trying to get across is that interference filters aren't like garden-hose accessories that one can simply attach to any other hose and get similar performance, and I think many amateurs lose sight of that. Another example more people might be familiar with are the etalon solar-filters for H-a viewing. They are very temperamental about primary f-ratio. Shift that ratio, and you shift the wavelengths being passed by the filter. I'm trying to avoid the conclusion that the TeleVue filter is truly a dog- the tests should be more comprehensive before we can come to this conclusion. Another possible explanation is a reflectivity drop-off in Jim's mirror coating that coincides with the O-III lines, alternatively your mirror may have a reflectivity spike at that wavelength. Chuck --- Richard Tenney <retenney@yahoo.com> wrote:
Chuck,
The scope we tested this in has 100 inches of focal length at f/6.25; Jim's 10-inch I believe is f/7 or more which he also had no success with; I wouldn't call either of these scopes particularly fast. On the other hand, maybe we should try the TeleVue filter in a faster scope, say an f/4.5 and see what happens.
But the bottom line was this -- the nebula was dramatically enhanced by the Lumicon filter, and only barely so (if at all) by the TeleVue, made even more obvious by my filter slide (both filters were loaded at the same time, and it was a simple matter of sliding each under the eyepiece (35mm Panoptic) without taking my eye away from the image. There was little if any difference between the TeleVue and the view sans filter.
My Lumicon did come with transmission specs (in fact, I bought it from a dealer who had 3 of them sitting on the table, and was able to pick out the "best" of the 3). Jim, did your filter come with similar information?
It would be interesting to repeat this test with other filters and/or filter manufacturers (anyone out there with a thousand oaks O-III in a 2-inch format?) and in different telescopes of varying focal lengths.
-Rich
--- Chuck Hards <chuckhards@yahoo.com> wrote:
The coatings that make these narrow-band filters work are marvels of modern engineering, sometimes dozens of layers thick. They are interference coatings, so the thicknesses of these layers must be precisely controlled to 1/4 wavelength tolerances (sound familiar?). Lumicon states that their O-III filter has an 11nm bandpass, meaning that most of the energy passed by the filter consists of only wavelengths about 5.5nm either side of the center of the band.
For O-III the lines we are interested in are at 496nm & 501nm, so 11nm encompasses both nicely if centered properly.
A finicky feature of interference filters is that they are choosy about the angle of incidence of the light passing through them. This makes sense, as the angle of incidence increases, the light must travel through a longer path in the coating material, thus interfering at a longer wavelength when rebounding.
There is a chance that Rich's filter is optimized for a fast Newtonian (at a higher angle of incidence)and Jims is happier at a longer focal ratio such as a "stock" SCT might work at (lower angle of incidence).
Food for thought.
Chuck
__________________________________ Do you Yahoo!? SBC Yahoo! DSL - Now only $29.95 per month! http://sbc.yahoo.com
_______________________________________________ Utah-Astronomy mailing list Utah-Astronomy@mailman.xmission.com
http://mailman.xmission.com/cgi-bin/mailman/listinfo/utah-astronomy __________________________________ Do you Yahoo!? SBC Yahoo! DSL - Now only $29.95 per month! http://sbc.yahoo.com
Chuck, Your point about h-Alpha filters' performance changing raises a concern: If you change what wavelength it passes, does it then become capable of passing light that can damage vision? Thanks, Joe Joe Bauman science & military reporter Deseret News bau@desnews.com (801) 237-2169
No worries, Joe. H-a filters must use an "energy rejection" pre-filter that acts much like a standard white-light mylar filter. It removes the intense heat and blinding light, broadband, before it enters the H-a filter. Many of you who have used older H-a filters have had to "tune" the filter in order to bring-out the prominences and other H-a details. What you actually did was center the filter's passband on the H-a line. Like tuning a radio, only with light instead of radio waves. You gotta be tuned-in to the right station to get the broadcast. You won't get harmful levels of sunlight if not passing the H-a line, you just won't see the details you're after. I realized as soon as I walked away from the keyboard on my last post that mirrors with differnt reflectivities didn't affect Rich & Jims test, since both filters were used on the same scope. Here's the real test for f-ratio sensitivity: Repeat the test using the same scope and filters just to verify the first run's results. Now, stop-down the same scope to f/10 or longer and repeat the test. The image will be dimmer because of the reduced aperture, but we are comparing the filters relative to each other at 2 f-ratios, not the view before and after stopping-down, so ignore it. If the filters still behave the same, then I'm afraid that the TeleVue isn't doing what it's supposed to do. The only other thing I can think of is if Jim actually got a narrowband LPR instead of an O-III, placed in the wrong cell. Narrowband LPR filters still pass the O-III lines, but have a bandpass of dozens of nm in that region, not a mere 11nm. Contrast will be much better with a true O-III filter on certain objects. If another TeleVue filter can be obtained for comparison, that might shed some light on the apparent discrepancy also. Chuck --- Joe Bauman <bau@desnews.com> wrote:
Chuck, Your point about h-Alpha filters' performance changing raises a concern: If you change what wavelength it passes, does it then become capable of passing light that can damage vision? Thanks, Joe
__________________________________ Do you Yahoo!? SBC Yahoo! DSL - Now only $29.95 per month! http://sbc.yahoo.com
participants (3)
-
Chuck Hards -
Joe Bauman -
Richard Tenney