oops, sorry Peter, I might have erred in my prior post. Thinking about how it probably works, you really need to subtract the total background, including skyglow - just taking off dark may not make much difference. Will PM.
I've done the dark subtraction. The results are confusing to me! The FWHM only changed a little, if at all. But, the aspects became terrible! Jeez, I hope they really are not that bad! I blinked two images - one dark subtracted, one raw. I could not detect anything by eye (other than a little bit of dark noise subtraction). I certainly couldn't see an aspect going from 8 into the 30s.
Good work on your experiments there. I find the problem in luminance rather than the blue channel (it tends to be in all channels).
So the theory its the colour correction of the scope is likely to be incorrect otherwise it would not bloat in luminance like it does.
Also I see the same issue at times with my CDK17 but that does have a corrector although chromatic aberration is minimal at the eyepiece end of scope as opposed to a refractor objective.
I'd say the problem is more too small wells. The other possibility is IR bloat which may affect all of the filters?
The other practical solution is to take a 10 minute luminance test exposure of a potential target and if you get bloated stars then pick another target. I have used the scope successfully on objects with no bright stars in the frame.
http://www.pbase.com/gregbradley/image/134561608 Microline 8300 and TEC180 bvery bright central star had a bright halo in red which was toned down in Photoshop but you can still see it somewhat. I put this down to small wells?\
I use Astrodon Gen 11 50mm square filters. They are top notch. Not sure how Astronomiks compare. I don't see blue halos specifically so perhaps they cutoff differently? Also I image with the TEC at a very dark site so that may also give a slightly different result.
One thing is for sure its only the Trius that does this. There are thousands of high end images from other cameras on TEC180/140's with no issue. So I'd say it points the finger at the Trius and perhaps filters if you are only getting the problem in the blue which is different to what I get.
What does work as a strategy on the CDK is to do shorter subs -5 minutes is my standard although sometimes I do 10mins if the target can take it. The CDK will actually burn out a galaxy core on 10 minutes on some subjects with the reducer!
Greg.
Last edited by gregbradley; 24-09-2014 at 06:07 PM.
Good work on your experiments there. I find the problem in luminance rather than the blue channel (it tends to be in all channels).
So the theory its the colour correction of the scope is likely to be incorrect otherwise it would not bloat in luminance like it does.
I'd say the problem is more too small wells. The other possibility is IR bloat which may affect all of the filters?
Greg.
Hi Greg, I would say that Peter's results and yours show quite clearly that the problem is a violet halo. Peter showed that a filter that removes the bottom 40 nm or so essentially cures the problem. The only 2 channels that can actually see the violet light are blue and luminance - which explains your lum result. This is not a criticism of the scopes - I am sure that they work exceptionally well - they just don't correct well enough in the violet (which is outside of the design range) to be suitable for use with a camera with small pixels and very high violet sensitivity.
Just to be clear, exactly what do guys mean by the term "bloated". If that just means bigger stars, then that is certain to happen - the Trius will produce equivalent stars that are 4x the area of those from the 16803 (or 1.4x the area of those from the 8300), simply because it puts 4x as many pixels under each star. Everything else in the Trius image will also be 4x the area including galaxies etc. So if you want to compare Trius images with 16803 images, you need to downscale the Trius images by 1/2 first. If you mean that the cores are saturated, that is due to to the combination of very high sensitivity and smaller wells of the Trius - saturation can easily be overcome by using short subs (eg a couple of minutes) and the Trius will allow you do that because of the low read noise. But fixing the saturation will not make the stars any smaller - it will just stop the cores of bright stars from losing detail and going white
I've done the dark subtraction. The results are confusing to me! The FWHM only changed a little, if at all. But, the aspects became terrible! Jeez, I hope they really are not that bad! I blinked two images - one dark subtracted, one raw. I could not detect anything by eye (other than a little bit of dark noise subtraction). I certainly couldn't see an aspect going from 8 into the 30s.
Check out the photo attached.
Thanks,
Peter
That confirms that I was wrong, as guessed - will have a look at the data with PI, which corrects for background.
The last thing I want anyone to feel in these posts is that I am in any way criticising the TEC branded refractor. That is far from my intent. They are amazing scopes! But, it's well known that Yuri figures the optics for visual. Now if one puts a camera on the scope that perhaps mimics the frequency response of the human eye the blue channel will have a vanishingly small problem as witnessed by my own photos with the KAF8300 and Greg's many photos. But, put a camera on that has such a highly accentuated response to blue/violet/UV and we have an issue. As Greg has said one must match the camera to the optics. This is something I tried to do when making a purchasing decision about the Trius, but I failed to appreciate this one particular issue in blue. I spent a lot of time agonising over pixel size, etc. And, looking for similar posts such as this one...well, I couldn't find one which I find a bit odd given the popularity of the TEC140 and the Sony chip. I do think that this issue is likely to be raised more and more as the Sony chip takes on ever more popularity. Perhaps opticians will begin to rethink how they figure APO optics in reaction to this issue. Who knows? Time will tell.
I guess some people will take a swipe at the need for a -violet type of filter on an APO. My response is that's pretty much a garbage argument! I try to think of it this way: we don't blame DSLRs for having an IR cut filter do we (unless we are doing astrophotography)? We accept that a filter is necessary to compensate for the CCDs increased sensitivity (compared to the human eye). I think a -violet filter is exactly the same, just at the other end of the spectrum! Now, I suppose the argument isn't particularly helped when one considers that many reflecting type scopes can maintain critical focus into shorter wavelengths and thus manage to avoid blue fringed stars. But, I would argue that this light is no more "real" (by real I mean to the human eye in terms of colour) than IR. Maybe my argument is self serving. I don't know, but I do know that I'm not stressed about using a filter if in the end I can take a great photo with tight stars!
That confirms that I was wrong, as guessed - will have a look at the data with PI, which corrects for background.
Ray, I will crop out the data you requested and put it on Dropbox. I'll send you the link privately. I really appreciate your willingness to take a look at this data and to help me figure it out!
Peter
PS: In your other post tonight your point is well taken about the "magnification" effect of the smaller pixels. Yes, one must compare apples to apples so to speak. Well, this is exactly what started me down this path: I take luminance at .88 arc-sec on the TEC180 with the G2-8300. The colour from the Trius on the TEC140 at .95 arc-sec. It is during the L-RGB combine (where the Trius image is scaled up to match the .88 arc-sec luminance) that I notice that many of the stars coming from the Trius are just larger than from the KAF8300 chip. Of course without a special filter blue was/is huge. But I notice it with other coloured stars as well.
A picture is worth a thousand words. Perhaps Peter's and mine setups are slightly different as when I see the problem its in all colour channels and luminance not just blue. When I do colour combines and process I don't necessarily get blue halos but often red sometimes it is blue.
So the conclusion that its the high violet performance of the camera I would question as it does it in every channel. Hence my conclusion its the small wells as I don't see it on a deep well camera.
I also see evidence of hazy bloated stars in others' images sometimes using other brand cameras too so it obviously is something to consider when setting up a system. I don't see it in Mike's images despite using a corrector and I don't see it your images Ray using a Newt with a corrector. I do see it to a much lesser extent on the CDK with corrector but again in all channels not just blue.
Not sure what you were referring to about blue/violet halos in the TEC180/Trius/Astrodon Gen 11 images I posted in my last post as none of them have that except one with rather minor bluish halos.
Below are luminance, red green blue subs usin TEC180 Trius 694 Astrodon Gen 11 filters. As you can see the bright stars are bloated and are much the same in all 4 channels with no one channel much better than the others. I think these were 10 minute subs.
I am interested in seeing the results from different filters like minus violet etc but don't expect it to handle based on my experience. Shorter subs, perhaps a UV/IR block may help. Perhaps its stronger response in UV/IR that is doing the damage?? The bloated stars look a bit like they can go when I use my clear filter with no UV/IR block on it.
I'd also love to see some images from a FLI Microline 694 to see if there is any evidence its really the camera and its electronics - like gain settings, quality of the ADC, quality of the coatings on the CCD window and other supporting electronics that could influence the final image.
Or is it simply the small wells. I see SX quoe 18,000 electron well depth yet FLI quote 20,000+ so there is some difference right there.
Your photos are superb and if I could get stars like yours I'd be quite happy.
Now, I believe that your Astrodon filter cuts at 400 nm, but the Baader I am using cuts at 380 nm. I use Astronomiks on my G2-8300 and I get good results in blue with that camera. I put that more to the chips QE in blue than to the particular cut off of the blue filter.
A friend loaned me an Astrodon blue filter last night. It produced slightly tighter blue stars than the Baader. But, not good enough to satisfy me. Adding the Hutech IDAS significantly helped as it cuts at 410 nm. Both filters showed improvement with the IDAS. A bit more cutting and I think I will be there.
I don't mean to beat a dead horse so to speak, but I think it's quite easy to mix up two different problems into one.
I'm quite convinced that in my system, for whatever reason, that I have a real focus issue, and large blue stars with halos, due to excessive violet out of focus light being picked up by the Trius and its high QE in that spectral region. Otherwise, how can I explain that I have so much better focus and smaller haols (in blue) using a filter that cuts at 410 nm? As noted earlier, my Baader filter passes to 380 nm which is 20 nm more than your Astrodons. And as I noted in the previous post the IDAS also improved the Astrodon filter I borrowed last night.
Problem 2 is probably completely unrelated to problem #1. I would also guess that the large stars one sometimes sees in other colours might be due to small well size. It is this second problem that I'm now trying to come to terms with. For me, though, it is much less of an issue than the unusable blue I was generating.
Strange you are only getting bloat in the blue. That would seem to point towards the filter or the scope.I can't say I have ever seen poor performance of my TEC180 in the blue channel though. Blue is often hazier than the other channels due to light scatter but seems no different to any other scope. As Ray points out this chip is also quite small so the magnification factor is quite high. Again not sure why that would be a problem but it may be magnifying whatever that something that is non optimum here.
I am curious though if my AP140 shows any difference. At least that would nail down the possibilities a bit more.
Another possibility for the bloat may be the microlenses on the sensor. I wonder if they don't like F7 and the angles the rays hit them? A bit of a long shot though.
The UV/IR block aspect may be worth an experiment. I have a Baader UV/IR block filter I could tape on the CCD window to see if it changes anything. Another little test next time I have it out.
I was not too worried about it really as my other cameras suit the style of images I want with this scope and my CDK does better than the TEC180 for galaxies although the TEC180 does an excellent job. The CDK at 17 inch aperture simply is better.
Greg.
Yes, Greg, I agree with you. It's quite important to keep the TEC140 and TEC180 discussion quite separate. I make no "claim" about the TEC180 as I've never tried my Trius on that scope. It may very well have a different colour correction than the TEC140. In fact, I'm unable to find any colour correction data on the TEC site about the TEC180 at all. Do you happen to have anything?
And again, I stress that the blue issue I see on the TEC140 is perhaps due to the way it is corrected AND exaggerated by the super hi QE of the Trius. At least my own tests seem to indicate this. However, the large star issue is probably something quite different once blue is taken care of. It may be something seen by both scopes and probably has much more to do with the camera and the way it is used rather than either scope.
Your eventual test of the AP130 will be most interesting!
Oh I misunderstood Peter. I thought the blue issue was with the 180.
And you are getting no problems in the other channels? Sorry you probably have mentioned all this in previous posts on this thread. You refocus after each filter change? In my experience Baaders are nowhere near parfocal despite their claim. Astrodon's are.
There is a difference in flourite and colour correction and fluorite is chosen because of that. But I would have thought it was more to be able to get the same colour correction with a faster F ratio rather than
better colour correction at the same F ratio over ED glass. There is better scatter which could account for better blue performance (but as you see in my samples not really only if the star is not too bright).
Good luck with your experiments. I also think it may be useful to post on the SX Yahoo Group to see what others are finding. Or use your 8300 camera for colour and the Trius for luminance.
I certainly do focus with each filter change. My Baaders do seem pretty close to parfocal. The whole issue came to light when I just couldn't find a focus in blue that looked anywhere near what the FWHM values indicate. In other words FWHM of say 2 but looks terrible on the actual image.
Yes, perhaps I could switch cameras on the scopes but I would lose the fairly lucky nearly same FOV and similar resolution I get with the cameras set up this way. Also, with the sensitive Trius gathering colour I think it's a win win situation. Once I get the blue tamed I should really be good to go. The resolution of the 8300 chip on the TEC180 is great!
The 8300 is quite a good match for the TEC180 but you still have to watch out for the occasional very bright star and exposure times. That 180 aperture lets through a lot of light fast.
I think this little picture gives this discussion a bit of perspective!
I have 2 filters coming via Gerd Neumann at Astronomiks - the GG420 and the GG435. He is happy for me to keep whichever performs the best in my setup. He also tells me that Astronomiks is developing a set of RGB filters to exactly address this blue issue in refractors. I will resurect the thread once I have a result!
I have finally completed some preliminary tests using various UV/blue blocking filters in combination with my TEC140/Trius combination. It appears I need to do a bit more as the data I have so far is not conclusive. Perhaps it will be of interest to others. And, certainly I'm open to suggestions for improvement of these experiments. There are so many variables to consider.
I have 3 filters to consider:
1. Shottt GG420
2. Shott GG435
3. Baader Fringe Killer
I previously tested the IDAS LP filter which gave some improvement in blue resolution and fringing (cuts around 410 nm). The Baader filter has a complex cut (see image below) and the Shotts cut at 420 and 435 nm)
I decided to image NGC1977 (Running Man) with my Trius using the Baader blue filter (which passes down to 380 nm). In front of the SX filter wheel I installed an Orion 4 position 2" wheel containing the 3 filters. I decided on a series of 10 x 2 min exposures for each filter. After collecting that data I had a quick look at the average adu in an area known to be blue. After a bit of trial and error I settled on an exposure of 1.5 min for series of RGB without any UV filter in place. My reasoning (perhaps incorrect) is that a 25% time reduction in blue without a filter would approximately equal the blue transmission with the UV blocking filters in place and lead to a hopefully close to correct colour balance. It appears that I somewhat underestimated the reduction, particularly for the Baader Fringe Killer.
First, a quick comparison of resolution shows that all three filters give a significant visual improvement to raw frames with the Baader doing the best job, though not by much.
I registered all the images at once and then stacked and mean combined for each filter. I then created 4 RGB images, one for each filter and one that used no blocking filter. The results were quite surprising as I fully expected to see a difference in resolution given the bloated appearance of the blue channel in the sub frames without the blocking filter. I cannot detect a difference in the final result in this experiment. I even went back and only registered the blue results that used no blocking filter. Again I saw no difference visually though the FHWM reported by CCDStack was about 10% worse than when I registered all the images at the same time. In any case. with this particular imaging target, it is difficult to conclude that I should be using any blocking filter at all even though with a different target I struggled with quite a lot of blue fringing on medium to bright stars. This time I simply don't see the fringing. But, I do see quite a lot of reflection halos in the sub frams as soon as I use the blocking filters. Again the halo effect goes away in the RGB combine, I guess, because there are no halos in G and R to reinforce. It makes me wonder if I saw the blue fringing in my earlier experiments only because I had not removed the blocking filters shooting R and G and consequently all the channels had refection issues.
I exported the RGB images from CCDStack after slight background adjustments to eliminate blue clipping that was showing with CCDStack background left at default. The images from CCDStack are what comes out of the default settings and no colour adjustment was made to the background. I only adjusted the background to eliminate clipping when exporting to Photoshop. I used Photoshop CS5 to align and balance black points of RGB. By far - to my eyes - the RGB image with no blocking filter looks the best in terms of blue (these were all 1.5 min subframes). As the strength of the blocking filter increases I see progressively more and more green caste which decreases somewhat when the RGB black points are aligned. However, even after alignment the blocking filter images do not look as nice in blue as the image using no filter. I have gone back and measured adu in the images and conclude that perhaps the Baader image should have been 132 seconds rather than 120 seconds. The GG filters exposure times needs perhaps only a bit more time. Here I am assuming a linear CCD response. In any case it looks with the use of these filters needs at least ca 30% increase in exposure time and the Baader perhaps 40% or more. In the next round of experiments I will increase the exposure time in blue to the calculated amount but also do a series with perhaps double the amount of time. It would be interesting to see if any amount of extra blue exposure time would result in as good an image as seen without the blocking filters.
Use a different camera. The KAF8300 cameras have no such issue and your off and running. Some setups simply don't play well together and that's part of the trick of astrophotography. Which items work well with what items.
The 694 chipped camera is great for galaxies, narrowband and small objects but its never going to be a good wide field chip. 11002, 16803 are the kings of that arena and will be for some time to come I predict (no new decent CCDs for years now and none I am aware of on the horizon either). The TEC140 is a premium widefield astrograph not so much a galaxy scope anyway.
With Kodak's ex-imaging division being sold around its unlikely they will produce a new hot CCD as it seems its a struggling business. Sony seems to be hot in development of cameras CMOS sensors but not so much in large CCDs where we need them.
There are thousands of excellent TEC140/11002/16803/8300 images.
Using a filter simply degrades the image and is unlikely to ever work well without compromise. You'd better off selling it (its still a great camera and apparently best suited to fast reflector scopes) and get another KAF8300 camera or an 11002/16803 camera.
