 #...
·
 |
|---|
The following is something that I just posted with my last image, since it is relevant to the data generated and processed. I am partially reposting here, and changing where appropriate to those who may not see my image Rosette West Widefield - In OSC - Lesson Learned Regarding Light Pollution Filter and Short & Fast Optics! For roughly eight months now, I have been seeing rings/halos around some of the stars in images that I have generated with newer optics. These are a WO ZS61, a rented Sigma f1.4 105mm lens, and two Rokinon 135mm lenses, one of which I kept as a permanent arrow in my quiver. At first, I just assumed that it was brighter stars and optical defects or possibly having to do with backspacing or the like. It is much more rare in my refractor, though one I saw, I thought that it was an unknown planetary nebula! However, as you can see in the image below, which is a cropped part of my image from the post above, it is much more common in my Rokinon 135 and even more so in the rental Sigma f1.4: Figure 1. A clip from a mostly unprocessed master (the full stack), with STR and StarExterminator applied. StarX, does a good job of leaving behind the offending ring halos, even some slight arcs.  After seeing this repeated, independently of any changes in backfocus or other such adjustments I could make, I started to realize that there was no logical reason that some of these stars had the rings and some did not. Brightness did not matter, color did not matter, background did not matter. If it was just a simple optical issue, then it would be expected to be with all stars of a particular class, i.e. bright stars, red or blue stars, etc. No such luck. Long Period Variable Stars: I finally went onto Aladin and found some of these stars and most were listed in Simbad. In certain images, each and every one of the stars that had such ring halos were Long Period Variable Stars. In other images, almost every one was, but some were listed as S stars or some other classification that ultimately falls under similarities to LPV stars. I have since found a few that are listed only as stars with no indication of LPV. However, the correlation is so strong that I feel that these stars, if studied more deeply would be found to fall into this catagory. What is common to all of these stars is that if one looks at them in imaging with NIR/IR or Mid IR cameras (easy to do on Simbad), these stars all jump out as having very high light fluxes in the NIR-MIR. In fact, probably much higher in those wavelengths that in the visible. That goes for those stars that are listed as just stars. However, stars that are listed as Red Giants do not cause ring halos. Even though red giants do have strong IR signals. Hopefully that is interesting to anyone who cares! A listing of some relevant facts: 1. I use an L-Pro LPS filter for all my imaging and when I use my QHY268MC camera, it also serves as an IR/UV cut filter, since the optical window of this camera does not have this feature, unlike its sister camera the ZWO 2600 color. The L-Pro is an interference filter. 2. I have used my L-Pro filter for a number of years on different, larger scopes very successfully as a IR/UV cut filter. In fact, terrestrially, using this filter on my QHY camera yields very natural colors for trees, houses etc., which absolutely is required to have the L-Pro installed to do so. 3. To repeat, I have found that almost every star that is associated with the ring halos are classified as Long Period Variable Stars! 4. This artifact does not follow star brightness. There are many bright stars that have no ring halos and dim stars with very strong ring halos. 5. The artifact is not dependent of background brightness or color. (See faint ring halo just left of the center of the Rosette.) 6. Upon investigation of these stars in Simbad, and visualizing these stars in Aladin with a number of different wavelengths, I find a common thread that these stars are extremely bright in the NIR and Mid IR. Conclusions: 1. I am not certain, but I believe that at the center of the problem is likely the L-Pro filter. However, not to put full blame on a filter that has otherwise done very well for me, it is also the fact that with these optics: 2. very fast optics coupled with 3. very short focal lengths, means the very steep light cone (especially from the periphery of the objective) may well defeat the ability of the interference filter to eliminate the NIR signal, to which these CMOS cameras have significant sensitivity. The angle of incidence of light onto interference filters is critical for performance. 4. But, why then do these fast, short optics create a ring of illumination? Why wouldn't they just focus the NIR light onto the location of the star? I believe that is because they are not designed to do that. Especially for the SLR lenses in use today in Astrophotography, the market for these lenses is for terrestrial light or artificial light (designed for human visual sensitivities). And that brings us back to the fact that unmodified SLRs and mirrorless cameras come with a very strict IR cut filter. In fact, for those who use SLRs, modified for astro work, these filters are way more strict than to just NIR/IR. See below:  Note the red response in the Red Bayer layer of a color camera. We all know that SLRs, etc. are very poor at recording H alpha signal at 655nm, which is really well within the optical sensitivity of red light for humans. So those SLR filters are really cutting a lot of the red spectrum and not just the IR! I am not sure exactly why, but it may have to do with the origins of color film, and getting a "real" color balance in the product (film and digital). One thing is for sure, narrowing the width of the spectrum in the most common cameras in the market makes it a hell of a lot easier for lens manufacturers to focus all those wavelengths onto a single spot. Therefore, I now believe that these ring halos are "where" they are because the lenses are not designed to focus the wavelengths of light that cause the ring halos. 5. Why are these ring halos not colored, such as red? The answer is if you look a the response curve above with the Bayer throughput overlayed, I have highlighted where the Bayer filters lose their ability to absorb light of certain wavelengths. Essentially the Bayered sensor becomes agnostic to any light above 770nm and that begins even before that. Because the light in the ring halo is long wavelength, all pixels see it and the halos will be some shade of grey. I see that all the time when I use my NIR camera, which actually is a color camera, but when I put a 950nm cut filter in place, the result is mono (gray). 6. The stars that cause the issue are not just red stars. Many red stars, even known red giants, do not leave these ring halos with my fast/short optics. I believe that the Long Period Variable stars are particularly rich in the longer wavelengths, vs just red stars. 7. Why did I not see this issue in my f2.2 RASA? Well my RASA was not that short a focal length, but Celestron clearly states that the RASA 11 and 8 are designed to not pass much light beyond 700nm. And that understood, is the cutoff that Celestron designed its focal abilities/capabilities for the scope. The 13 inch RASA is specifically designed to have an extended useful spectral range from 700-900nm. And that may have been more easily achievable for its longer focal length. I have a very speculative theory on why LPV stars may have this issue. It seems that these stars are near the ends of their lives and they are variable because they are ejecting large quantities of gas and dust. Maybe like what is going on with Betelgeuse right now. This means that there is a large amount of dust in the local region of these stars and because dust is close, it can get heated enough to emit very large amount of infrared light. Much more than a red giant star can alone. How or why this is so unique to this type of star is beyond my current understanding, since I would have thought that situation could be met by other types of stars. It may just be that LPV stars are much more common. In any case, for these new short, fast lenses, I am going to need a more conventional IR/UV cut filter! And that bums me out, because I have come to rely so much on the light pollution reduction of that class of filters, which all seem to be interference-based construction. It should be noted, that if I also was using or interested in any of the other multiband interference filters that are all the rage right now with OSC cameras, one should be looking to see if there is an issue in their images with these stars. If anyone who is new to using these short, fast optics and are having these same problems, I hope that this is helpful. And other than knowing to choose the correct type of IR/UV cut filter for their OSC camera, if anyone knows of any other reasons or similar issues, it would be nice to get some feedback. Finally, if I were really into finding and studying LP Variable stars, this would be the tool for hunting them down! Anyway, I will repost this on the Forum, but if anyone else has struggled with this issue, let me know. Also, if I am a novice bonehead, and "everyone already knows about this", please don't hesitate to let me know as well! Best, Alan |
4.40
#...
·
·
1
like
|
|---|
| Interesting. I do not see this with my IDAS NBZ and an IMX571 camera with AR window (no separate UV/IR cut). And, of course, also not with my LRGB-filters (Chroma, Astrodon, Astronomik) that are also interference filters. Would be interesting to hear other users of the L-Pro chime in. I tried searching a bit, and could find no other example of this, so it might be a peculiarity/defect in your sample (see here, for example: https://www.astrobin.com/full/jec3tv/0/?mod=&real=). |
#...
·
|
|---|
Torben van Hees: Actually, Torbin, your example from @Riedl Rudolf 's fine image actually demonstrates a couple of fine examples of Variable Star Halos resulting from the L-Pro. It therefore confirms that my filter is not the only one to show this effect. Though mine, unfortunately, may be a prime example! Maybe Riedl will have more to say from his experience with this filter/lens combination. The most dramatic is the star T-Cep, a Mira variable (long period, by the way). I copy it here to show you along with the Aladin image in NIR to show you how very bright these stars are in NIR: Riedl's T-Cep (Edit to add: Note the brightness of T-Cep in Riedl's image is unnaturally much too bright. I noticed that all such stars in my images are also much too bright.)  DSS color image from Aladin: (Note T-Cep, in the cross-hair, is not a particularly bright star.)  J-band image from 2MASS:  Ka-boom! You can see just how strongly these stars emit in the NIR, and even IR. You can also find another in VV Cep in the image, a very interesting variable of red giant, blue giant binary star. I think your experience with IDAS NBZ is a hopeful sign that this is not necessarily a common or general occurance for the dichoic interference filter. No doubt, the complexity of these filters, in how they are designed, and also how they are manufactured, could lead to variability in performance, batch-to-batch. In this case, I can believe that the performance is well within the spec and purpose designed, but where the variability comes into play is outside the performance window of operation, such as doing its job with such short optics. This may have been unexpected from the designers of this filter to be using f/2 and faster lenses. I would not expect this phenomenon to cause issues with anyone doing narrow band imaging or at least LRGB imaging. Those seem to be more tightly designed, since they have a simpler job, in a sense. They also seem to incorporate other elements other than just dichroic. However, I think anyone working with the fast camera lenses should be aware of the issue and know when to reject a bought filter if they see the problem. My simple solution is that for my 135 f/2, I have a normal UV/IR cut on the way. I will post if it resolves the issue. I will keep using my L-Pro for my longer optics as an LP/UV/IR filter. Some say it is a waste to use an LP filter in Bortle 4 skies, but I have some light domes around me and find that it helps with gradients, or when sky clarity is not perfect, which is often for me. The "halo" artifact makes perfectly good sense to me now, after giving this issue some thought. The defect in the L-Pro is that it fails to filter NIR from the high angle component of the image, which in a ray trace diagram comes from the part of the objective nearest the outer periphery. Because we focus optimally for the visible bands, this IR image is just the bloated IR image of the star. It looks like a ring because the center of the star image is a tight focus of the visible star, and the IR component that would have filled in the center is deleted by the functioning of the L-pro in that region of the star image (that region has no IR signal because that part of the defocussed star image comes from the center of the objective. Those who use Newtonians and SCT know that a defocussed star image shows the central obstruction clearly). Hence the ring! If I ever get rid of my L-Pro, I must remember not to throw it in the trash, but rather donate it to science for anyone who wishes to use it to discover the most interesting variable stars! |
#...
·
|
|---|
@Torben van Hees, your comment spurred me to look into the frequency of this issue, which I hadn't thought to do at the time I first posted. I used the search function to search for images using a 135mm optics and LP filters (generic) and narrowed the search to the last year. I did find other instances. My search could not specify many of the particulars. I started off looking and basically avoiding ZWO cameras, with the thought that most of them had UV/IR cut built in. And I did find these Long Period Variable halos in some images. I then, just for the heck of it looked at some ZWO cameras and surprizingly found halos when the camera listed was a ZWO ASI294MC Pro. Only then did I learn that camera only has only an anti-reflective window. So in fact I could even use the presence of halos in images to tell whether it had a UV/IR cut filter built in. I'm not going to link specific images since I think that may be inappropriate to do here. Troubling was that I found some halos in images that used other filters than the L-Pro. If I recall, one was a duo-band filter and another, the L-enhance. I am not sure I will spend much more time looking for these, since I am not sure that anyone notices (even the makers of the images) or if anyone cares. If it was just a tight halo around each affected star, few and far between, then I would probably also not care. However, the fact that these otherwise often faint "stars" show up in these images as super bright stars really bothers me. Since it looks like stars that should not be there, never mind the halo. Next thing is users of fast optics are going to start reporting supernova from their images  |
|
4.40
#...
·
·
1
like
|
|---|
| You‘re right. I did not think to double-check the star brightnesses with visual magnitudes. I don‘t find it easy to see on heavily postprocessed images. Your geometric reasoning seems sound. Actually, it testifies to the quality of the Samyang 135: The visible spectrum is tightly contained, as only the IR shows a bloated star image. Many, even more expensive, refractors already show the halos in the red, especially with dual-narrowband filters. In those cases, the additional IR bloating might be harder to detect. |
#...
·
·
1
like
|
|---|
|
I also am becoming more impressed with the 135 as I continue to work with it. I have no doubt that when I get my UV/IR Luminance filter, the problem I am having will be one less issue (probably the final one) to deal with with that lens. I do not like the halos in my image even though at a normal viewing size they are not so easy to see. But the fact that some of these variable stars show up so bright cannot be avoided, even if no one would recognize them as such. But as you can tell it bugs me, so it just costs me extra time to correct them during processing. I have no doubt that others have seen issues with this lens, and my looking for them in some of the other images posted here on AB may have been defeated by the fact that those images have been "repaired" by their authors. Just as I mostly have done repairs on the few images I posted. Related to this effect, I now believe that a related issue that I had when using a rented Sigma 105mm f/1.4 Art lens may have been caused by a similar effect when using the L-Pro. With the 105, I most certainly find halos around LPV stars just like with the 135 when using the L-Pro. However, with the f/1.4, I also had large blue white halos around bright blue stars only with the 105 (not the 135 f/2). I am going to stick my neck out, but I bet that the very steep image cone of the f/1.4 somewhat also defeated the ability of the L-Pro to filter out the strong UV from those stars. I very much blamed the Sigma 105 lens for the problem. Likely, all along, it was my fault! Torbin, I appreciate your input in this discussion. It has helped me to much better understand what is going on, and I hope that it will help me get best performance from this lens and similar lenses I might use in the future. Best, Alan |
