2.94
...
·
 |
|---|
There are plenty of narrowband filters that work well at F2. Either mono or Nbz 2 LPara for OSC. I want to see people imaging broadbanding at F2 or native full speed, i know about narrowband, but i want to know why to stop down if it is a broadband data. |
4.06
...
·
|
|---|
Hi Tareq, with narrowband, you can use, as correctly remarked, high-speed filters that factor in the band shifting. An increase in f/ is not required if you adopt these filters. The reason for installing rings is a different one: fast f/ make difficult to eliminate defective peripheral stars. Therefore many chose adopt rings to increase the f/ and therefore to reduce the entity of the residual aberrations, at the price of much longer exposures! This is for broadband and narrowband. |
0.00
...
·
|
|---|
Tareq Abdulla:There are plenty of narrowband filters that work well at F2. Either mono or Nbz 2 LPara for OSC. Because in my case all the observed aberrations occured with broadband imaging. I haven’t tried using a narrowband filter yet, as I don’t have one suitable for fast focal ratios. I suspect, however, that with narrowband imaging, the aberrations might be less noticeable - though I could be wrong |
2.81
...
·
·
1
like
|
|---|
Tareq Abdulla: My problem, and I think everybody’s problem with the lens are the shape of the stars what is super difficult not only to controll but also tilt, backfocus and other image problems that are super sensitive at F/2. And apparently not all of this lens model comes perfect from the factory. But we need to have in mind that is a photographic lens been a lens, not an astrograph.. fully open at F/2, gives in bigger stars an lighthouse/hourglass effect resulted from the back retaining ring that secures the optical elements in place. This is in my case the real reason that some of us reduce the aperture down. It helps mitigate the star problems and some distortion in the image, etc. I don’t have mine fully corrected, despite compensate with a tilter and having the backfocus spot on in place, but I use her only with 4/3s sensors, which helps a lot in image quality. In my scopes despite that I have no problems working with sensors up to full-frame, for personal and financial preferences, I never wanted to go above the 4/3s size. For the FL that I’m use to work with, the 4/3s size seems to be the spot on size in many fronts. The solution presented on using the native aperture at F/2 for the DSO and the stepper down aperture for the stars seems a good fit, but needs more shooting time and the solution passes by not using the aperture internal blades of the lens, but using some stepper rings for attaching photographic filters in the front of the lens, to reduce the amount of light entering the lens that equals the aperture blades. This way avoids the star diffraction spikes caused by the internal lens blades. This is if you care with those intricacies. I do.. |
1.51
...
·
·
1
like
|
|---|
Michael E.:But at 2.8 the aperature is 48mm. Compared to being wide open, you are going to need to double your exposure time for a similar image. It might be worth considering taking a few frames to do a star overlay at 2.8 but the bulk of your imaging session at 2.0 would be the best bang for your buck. Correct, though in practice, this will not be ideal per se. Idk if this will work with your workflow, but another option would be to bin the image after stacking, and then stretch, and see if that helps with abberations. If sky-noise or read-noise limited, a 2bin will improve s/n 2x with a potential cost of spatial resolution. In post processing, you could then choose to keep the high resolution detail and then upsample the low signal areas. Then deconvolution sharpen high signal areas (I still use ImagesPlus for Adaptive Richardson Lucy deconvolution). You can also try drizzle and then bin, assuming you dithered, and see if star shapes come out better. Otherwise, as Tony references above, its all about the composition at 135mm focal length, as a result, most likely what you can do with ease to manage the abberations is downsize your final image display size. |
0.90
...
·
·
2
likes
|
|---|
Your pictures look great actually. AC is very controlled. The coma aberration can be easily fixed by blurX. Judging by the last picture you share at f2.8, the sensor could be a little closer than the optimal backfocus, that creates a radial pattern  If you want to do a final test on the back focus, attach the lens on your canon 6D then focus on a point at infinity (far away building for example). Without changing the focus of the lens, connect your dedicated camera, and see if the focus is correct. If not, don't change the focus, remove the dedicated camera, add (or remove) spacers rings and reconnect the camera to see if the focus match. Keep tweaking the spacers until the focus match. In my case, the spacers that came with my camera were not enough, so I added some paper spacers to test. I was off by almost 3mm. Unfortunately, the equipament say they have some thickness and etc, but small defects can add up. |
|
2.94
...
·
|
|---|
Ok, so all the comments are about the stars, is there another problem with f/2 or faster or native aperture of lenses? What will happen if i use 3nm filter standard speed with f/2 lens other than stars issue? |
|
4.06
...
·
·
1
like
|
|---|
@Tareq Abdulla Exactly: we are talking two different topics, even if both originated by fast lenses. Let's leave the star shape question apart, and let's focus (oops!) on the usage of narrowband filters with fast f/. Yes, you can use narrow narrowband filters at f/2, provided that their pass-band range has been appropriately pre-shifted: some producers (Antlia for instance) call these filters "High Speed". For instance, the "normal" Antlia ALP-T is a (wonderful) 5 nm dual band, but should not be used at f/2. Instead its brother, the 5 nm ALP-T HighSpeed, is specifically designed for f/2 focals and can be used. Hope this helps. Ciao, Mau |
|
2.81
...
·
·
1
like
|
|---|
Tareq Abdulla: *The problem regard fast system does not related only to the stars, but also the image in general. A small tilt in a fast system can blur a part of the sensor frame regarded to the stars, but that doesn't affect only the stars itself, also the nebula, galaxy, setc. behind it. Because one part will be focused, the other blured. That's tilt and backfocus off. Things get worst when you go up to F/2 and even F/1.5, because in lenses for example the center of the lens could be in total focus and the corners out of it. In cases of lenses and refractors the center of the cell will be the sharpest point of the image, but the corners won't, even if you have an almost corrected optical system. At F/2 everything is amplified. When you talked about filters, you have what is called wavelenght band shift. And that happens always to the left part of the visible spectrum, unless you use specially designed filters for those highspeeds. If you close on the bandpass of the filters such as 3nm, the tolerances will be even shorter. Imagine that your shooting using an Antlia 3nm "normal" Ha filter in an fast system. The FWHM will be 1.5nm for each side of the 656.3nm bandpass of the Ha. In a fast system with such narrow FWHM, insted of your shooting Ha, you'll be shooting NII.. This is why for example Baader, have narrowband filters for highspeeds in ranges such as F/1.8 to F/2.8, F/2.8 to F/3.5.. etc. because the FWHM is too short. In the end of the day, if you try to use a filter that is not designed for a certain range of aperture/ f/stop, you'll have bandshift to the left side of the spectrum or to the blues, as some people call it. Bottom line: if the wavelenght is not properly centered or is not in range of the FWHM, you'll end shooting another emission rather the one provided by the filter, and with that you also loose in the case of the FWHM be in some range of the central emission line of the ionized gas, you loose transmission of it. I don't know if this helps to understand more about not only the behavior of the fast systems but also the filters. Regards, Cesar |
8.11
...
·
·
1
like
|
|---|
Because of the way camera lenses form the final image, I wonder if the F ratio rule and narrow band filters still applies? Meaning that it's not a simple optical system and it's possible that the angle of the light converging to focus does not match what you would expect from the F stop.
|
|
0.00
...
·
|
|---|
Thank you all for helping to quantify the star shape problem and exploring the possibilities for mitigating or working around it. I will definitely also try the "focus-with-Canon6D-and-adapt" solution, and I’m curious to see how much my backfocus was off. Now that this topic is shifting towards the use of narrowband filters on fast systems, there is one thing I fail to understand: => How can one filter be successfully used on a fast system and manage the rays approaching the filter at a "flat" angle (originating from the edge of the lens or mirror) while simultaneously dealing with the rays coming from the center parts, which enter the filter at a more perpendicular angle, as would correspond to a normal narrowband filter? The only explanation I can think of is that the filter coating is not evenly applied but instead follows a radial pattern. Is that the case? |
|
2.81
...
·
|
|---|
Michael, I'm not an expert in the way those filters are manufactured, to be honest. What I know is the light cone in fast systems, from lenses to mirros, is shorter and larger. Filters for fast systems (I don't know the way they are coated beaside the use a multiple layers with dialectric coatings deposited on them), manufactures pre-shift the filters to the red/right part of the spectrum, in order when used in fast systems, the central wavelenght falls back into the right transmission line. The shifting of the band happens always to the left part of the system, toward the blues, I don't know why, but it does in any filter. The angle of incidence resulted of the shorter and wider light cone, leads also to a loss of transmittion, which usually reduces contrast and efficiency compared to a normal narrowband filter, but since we are gathering a lot of light we don't see it because we don't look for it. Our focus is mainly in capture the details. But if you compared a master file made on the same target, using the same sensor, but with a fast system and a normal/slow system and their correspondent filters, you see that exact problem. You'll have more signal (but not necessary ionized gases) in one rather than the other, but you also see the loose of contrast, and in my personal opinion (detail). This is how my knowledge goes in this matter, beside what I've talked above. I'm not a speciallist or optical engineer. Maybe someone with deeper knowledge can help. Regards Cesar |
|
0.00
...
·
|
|---|
Thanks, Cesar. So now I understand it in this way: The design of narrowband filters for fast systems is actually a trade-off. Due to the pre-shift, they perform well toward the edge of the aperture but suffer from some transmission loss toward the center. However, since the area of the aperture increases with the square of the radius, the outer part contributes more area and thus more signal overall. This makes the central area act almost like a central obstruction, with its lower transmission also contributing to the loss of contrast, as you observed. I also wonder how much of an effect various optical designs have on the angle of the light cone. I could imagine that a Petzval system, a lens like the Samyang with 11 elements, or even just a field flattener might produce more parallel rays compared to a single lens group system. But then, not all systems would behave the same in this regard. Who knows, maybe there are f/2 systems that behave more like f/4 systems in terms of the angle of incidence? |
9.89
...
·
|
|---|
There are plenty of narrowband filters that work well at F2. Either mono or Nbz 2 LPara for OSC. No there aren't and they are all rather expensive. |
|
9.89
...
·
|
|---|
Michael E.:But at 2.8 the aperature is 48mm. Compared to being wide open, you are going to need to double your exposure time for a similar image. It might be worth considering taking a few frames to do a star overlay at 2.8 but the bulk of your imaging session at 2.0 would be the best bang for your buck. I find the above statement about BXT rather puzzling as I can correct worse aberrations without leaving this much of a trace. Are you sure you run the latest AI? |
|
0.00
...
·
|
|---|
@andrea tasselli I'm not sure. BXT version 2.0.0, AI version 4, with default settings.
|
|
9.89
...
·
|
|---|
Same as mine. Can you make a sample available so that I can check?
|
|
2.81
...
·
|
|---|
Michael E.: I can't say Michael, how they are coated the the process itself. I don't know if they perform the way you say.. But I imagine the transmission is evenly in all the filter, not the way that you're describing, because if that was the case, made all the sense for coating circular filters, but not so much for square 50mm and the new ones which they are a in a rectangle shape. I think the factory coatings and transmission are all the same through the entire filters for fast systems. But like I said before.. My knowledge before what I've talked about, is limited because I don't work and didn't designed any filter up to this day |
7.19
...
·
|
|---|
Tareq Abdulla: I stop down with mine to f2.8, look at todays IOTD, a samyang stopped to f4. It is an inexpensive lens, a $250 telescope. Stopping down on most copies of this lens is reasonable and practical. |
|
0.00
...
·
|
|---|
andrea tasselli: Happy to do so, but I guess I need to crop it heavily to not exceed the size limit here and am unsure if the effect will remain the same. I guess you are talking about the stacked raw data in 32-bit? 16-bit in linear mode is certainly also not ideal. Otherwise, shall I provide a download link? |
|
9.89
...
·
|
|---|
Download link would be ideal.
|
|
2.94
...
·
|
|---|
I asked or posted i it is all about stars because i was thinking about what if we use this lens for just NB signals and remove stars, will we see issues also? I was planning to use another optic for stars without stopping down for just RGB, so if i shoot say Ha with this lens at f/2 using any regular Ha filter not high speed and i remove the stars, is there any issue left then? |
1.20
...
·
·
1
like
|
|---|
Michael E.:But at 2.8 the aperature is 48mm. Compared to being wide open, you are going to need to double your exposure time for a similar image. It might be worth considering taking a few frames to do a star overlay at 2.8 but the bulk of your imaging session at 2.0 would be the best bang for your buck. So I faced something almost similar but not black gaps, after applying BlurX, big stars left a huge ring around them. So what I did is apply StarX first then apply BlurX on the stars image alone, and the starless alone and it worked. So you can try this, StarX before BlurX, then apply BlurX two times on stars and the starless. I know this is not how it should be done, but it worked for my case. |
|
0.00
...
·
·
1
like
|
|---|
Thank you @Mohammed Hussain Khalifa , interessting precedure - I'll try and report back...
|
|
0.00
...
·
|
|---|
andrea tasselli: I just realized that the samples I provided were processed on drizzled data, where the effect is much more pronounced. If you'd still like to give it a try, here is a download-link to a significantly cropped (for file-size reasons) raw stack of the drizzled data: https://www.swisstransfer.com/d/926cfb25-1411-487d-9197-10bbf2e2fbb6 |
