There are quite a few narrowband filters on the market today designed to work with very fast optical systems, both for OSC and mono. I'm specifically interested in Astronomik's MaxFR filters, which come in 12nm and 6nm, for my Sigma Art f/1.4 lens. The 12nm is advertised to provide 85% transmittance at f/1.4 and 100% at f/1.7 while the 6nm is supposed to give 90% at f/2 and 100% at f/2.2 and slower. Given that I image under severely light-polluted skies, I'm trying to figure out if I should aim to take advantage of the lens' speed and go for the 12nm filter or try to maximize contrast and go for the 6nm at the cost of being a little slower. What do you think?
Speed or contrast - which to prioritize for narrowband?
Well Written Engaging
Under light polluted conditions I'd personally go for the 6nm. The increased contrast and reduced gradients will make processing easier.
Well Written Concise
I use the Astronomik MaxFR 6nm Ha and SII on a Tak FSQ106ED at F3.6. My skies are Bortle 7. The 6nm Ha and SII blocks the light pollution well. The background is dark and even. My experience over a decade of doing narrowband is the 3nm is needed for OIII to render an even dark background. I had to go with a Chroma F3 filter.
It has been my experience that 12nm will still produce gradients and poorer S/N dew to the noiser background. I fear any gain in light transmission from the 12nm will be unproductive dew to the fact that a percentage of that light will be unwanted signal from sky glow.
I personally discovered that 12nm will not sufficiently block Bortle 7 sky glow. I do however plan to use 12nm at a Bortle 2 dark site.
Lynn K
It has been my experience that 12nm will still produce gradients and poorer S/N dew to the noiser background. I fear any gain in light transmission from the 12nm will be unproductive dew to the fact that a percentage of that light will be unwanted signal from sky glow.
I personally discovered that 12nm will not sufficiently block Bortle 7 sky glow. I do however plan to use 12nm at a Bortle 2 dark site.
Lynn K
Helpful Insightful
Lynn K:
I use the Astronomik MaxFR 6nm Ha and SII on a Tak FSQ106ED at F3.6. My skies are Bortle 7. The 6nm Ha and SII blocks the light pollution well. The background is dark and even. My experience over a decade of doing narrowband is the 3nm is needed for OIII to render an even dark background. I had to go with a Chroma F3 filter.
It has been my experience that 12nm will still produce gradients and poorer S/N dew to the noiser background. I fear any gain in light transmission from the 12nm will be unproductive dew to the fact that a percentage of that light will be unwanted signal from sky glow.
I personally discovered that 12nm will not sufficiently block Bortle 7 sky glow. I do however plan to use 12nm at a Bortle 2 dark site.
Lynn K
Thanks. When I imaged at f/5 I found even a 3nm OIII let in too much LP so maybe I’m nuts for thinking a 12nm will be workable.
You can calculate exactly the effect on Signal and Signal to Noise in different situations:
Lets say your signal S=1 for f1.7 and 12nm filter with 100% transmission.
So the signal at f2.2 and 6 nm filter with 100% transmission is (1.7/2.2)^2 = 0.6. In both cases the filter is 100% efficient, so the filter has no effect on the signal.
First message is you lose 40% of your signal if you go to f2.2. In the brighrer parts of your target this is the only thing that matters, since the background signal is small compared to the bright parts.
The faintest parts of your target will be just above the noise level of the background signal, so for faint parts you worry about the S to N, where the noise is coming from your background, which is your LP.
Roughly speaking, the amount of background signal due to LP is proportional to the width of the NB filter. So 6 nm filter would cut down background signal by 2X compared to 12 nm. Now the noise from the background LP signal is the square root of the background signal, so our noise drops by 1/sqrt(2) going to 6 nm filter from 12 nm. Doesn't matter what your LP level is, this ratio will be the same.
So lets take the noise with your 12 nm filter to be 1, and your S/N at f1.7 is 1. Then the noise with your 6 nm filter will be 1/sqrt(2). S/N then at f2.2 and 6 nm is = 0.6*sqrt(2) = 0.86. So you loose 14% in the S/N in the faintest parts of the image by going to the NB filter. This would require a 1/0.86^2 = 35% increase in exposure time to compensate, a pretty big effect.
Pretty clear lower f-number trumps the narrower filter benefit in this case.
Now if f number were constant and both filters were equally efficient at that f number, then the 2X narrower filter gives you a 1.44X improvement in S/N!
One note, contrast is a tricky concept with LP. So digitally you can set your LP background to any value you want, whatever the LP is. Then contrast really only depends on the light signal. So you loose a lot of contrast by going to higher f number. For faint parts of the image it is S/N is the important thing, as faint signals are lost in the background noise. So for faint signals it is the background noise that limits your image contrast–and this is where NB filters help, except that in this case you lost signal due to the f-ratio change.
Hope this helps.
Rick
Lets say your signal S=1 for f1.7 and 12nm filter with 100% transmission.
So the signal at f2.2 and 6 nm filter with 100% transmission is (1.7/2.2)^2 = 0.6. In both cases the filter is 100% efficient, so the filter has no effect on the signal.
First message is you lose 40% of your signal if you go to f2.2. In the brighrer parts of your target this is the only thing that matters, since the background signal is small compared to the bright parts.
The faintest parts of your target will be just above the noise level of the background signal, so for faint parts you worry about the S to N, where the noise is coming from your background, which is your LP.
Roughly speaking, the amount of background signal due to LP is proportional to the width of the NB filter. So 6 nm filter would cut down background signal by 2X compared to 12 nm. Now the noise from the background LP signal is the square root of the background signal, so our noise drops by 1/sqrt(2) going to 6 nm filter from 12 nm. Doesn't matter what your LP level is, this ratio will be the same.
So lets take the noise with your 12 nm filter to be 1, and your S/N at f1.7 is 1. Then the noise with your 6 nm filter will be 1/sqrt(2). S/N then at f2.2 and 6 nm is = 0.6*sqrt(2) = 0.86. So you loose 14% in the S/N in the faintest parts of the image by going to the NB filter. This would require a 1/0.86^2 = 35% increase in exposure time to compensate, a pretty big effect.
Pretty clear lower f-number trumps the narrower filter benefit in this case.
Now if f number were constant and both filters were equally efficient at that f number, then the 2X narrower filter gives you a 1.44X improvement in S/N!
One note, contrast is a tricky concept with LP. So digitally you can set your LP background to any value you want, whatever the LP is. Then contrast really only depends on the light signal. So you loose a lot of contrast by going to higher f number. For faint parts of the image it is S/N is the important thing, as faint signals are lost in the background noise. So for faint signals it is the background noise that limits your image contrast–and this is where NB filters help, except that in this case you lost signal due to the f-ratio change.
Hope this helps.
Rick
Helpful Insightful
I don't have experience with that particular filter but I will agree go with the 6nm. I use and still have the IDAS NBZ filter with the 10nm wavelength and I then bought the Antlia ACP-T which is 5nm and the difference in gradients is significant! It will help your processing as @starfield mentioned. I will say there is a trade off in that you'll need to have longer individual subs, at least that's been my experience. I would have never thought I'd have to go 10 min subs but with that filter its becoming the norm even with my 12" f/5 newt.
Dale
Dale
Helpful
Rick Veregin:
You can calculate exactly the effect on Signal and Signal to Noise in different situations:
Lets say your signal S=1 for f1.7 and 12nm filter with 100% transmission.
So the signal at f2.2 and 6 nm filter with 100% transmission is (1.7/2.2)^2 = 0.6. In both cases the filter is 100% efficient, so the filter has no effect on the signal.
First message is you lose 40% of your signal if you go to f2.2. In the brighrer parts of your target this is the only thing that matters, since the background signal is small compared to the bright parts.
The faintest parts of your target will be just above the noise level of the background signal, so for faint parts you worry about the S to N, where the noise is coming from your background, which is your LP.
Roughly speaking, the amount of background signal due to LP is proportional to the width of the NB filter. So 6 nm filter would cut down background signal by 2X compared to 12 nm. Now the noise from the background LP signal is the square root of the background signal, so our noise drops by 1/sqrt(2) going to 6 nm filter from 12 nm. Doesn't matter what your LP level is, this ratio will be the same.
So lets take the noise with your 12 nm filter to be 1, and your S/N at f1.7 is 1. Then the noise with your 6 nm filter will be 1/sqrt(2). S/N then at f2.2 and 6 nm is = 0.6*sqrt(2) = 0.86. So you loose 14% in the S/N in the faintest parts of the image by going to the NB filter. This would require a 1/0.86^2 = 35% increase in exposure time to compensate, a pretty big effect.
Pretty clear lower f-number trumps the narrower filter benefit in this case.
Now if f number were constant and both filters were equally efficient at that f number, then the 2X narrower filter gives you a 1.44X improvement in S/N!
One note, contrast is a tricky concept with LP. So digitally you can set your LP background to any value you want, whatever the LP is. Then contrast really only depends on the light signal. So you loose a lot of contrast by going to higher f number. For faint parts of the image it is S/N is the important thing, as faint signals are lost in the background noise. So for faint signals it is the background noise that limits your image contrast--and this is where NB filters help, except that in this case you lost signal due to the f-ratio change.
Hope this helps.
Rick
This is very helpful, thanks. I guess my follow-up question is when to prioritize gradient management over raw SNR. A 12nm filter is going to have a substantially worse gradient than, say, a 4nm filter. Gradient removal software has gotten a lot better, but in my experience it is still an uphill battle to manage severe gradients from wider filters.
Baader's 4nm OIII filter apparently gets ~87% transmittance at f/1.8 and ~95% at f/2. I'm thinking it might make more sense to just stick with my 6.5nm Baaders for Ha and SII and go to their 4nm for OIII simply to make processing less of a headache. The downside of the Baaders is that they're 2mm thick rather than 1mm like the Astronomiks, so spacing will be tough.
Well Written Helpful Insightful Engaging
So at f1.8 with a 4 nm you would only lose about 15% on the signal due to the transmission loss (ignorning a very slight f ratio changes), and background noise would go down by sqrt(3)=1.7X now compared to 12 nm, so your S/N would improve in the fainter parts of the image by 0.85x1.7 = 1.47X, so this seems like a really good option as far as I can see. You should be able to increase your spacing, the filter effect will be about 2X being 2X thicker, or about 1/3*2mm = 2/3 mm typically (depends also on RI). You can get fine delrin spacers to make this up, Baeder for example has a 0.6 mm T2 spacer which should be close enough?
As for the other filters, up to you how to prioritize. i'm a bit puzzled why you are struggling with LP gradients with a NB filter. I don't really have any issue with my L-eNhance which is 24nm/10 nm wide for Ha/OIII and certainly not for my L-eXtreme, which is 7nm/7nm. The only exception to this being a full moon, or perhaps if I am trying to target something close to a streetlight. So I just plan my work away as much as possible from those sources. I use Startools module to clean the background, not an issue that I can recall. Actually I have far more problems with bad flats causing gradients than LP when doing NB. Do you have a long dew shield, this helps to block stray light? I also have a small makeshift barrier to block a neighbors light when he leaves it on. Others have made a full surround out of sunshield mesh, to more fully block light, and as a bonus wind too.
Good luck with whatever you choose!
CS
Rick
As for the other filters, up to you how to prioritize. i'm a bit puzzled why you are struggling with LP gradients with a NB filter. I don't really have any issue with my L-eNhance which is 24nm/10 nm wide for Ha/OIII and certainly not for my L-eXtreme, which is 7nm/7nm. The only exception to this being a full moon, or perhaps if I am trying to target something close to a streetlight. So I just plan my work away as much as possible from those sources. I use Startools module to clean the background, not an issue that I can recall. Actually I have far more problems with bad flats causing gradients than LP when doing NB. Do you have a long dew shield, this helps to block stray light? I also have a small makeshift barrier to block a neighbors light when he leaves it on. Others have made a full surround out of sunshield mesh, to more fully block light, and as a bonus wind too.
Good luck with whatever you choose!
CS
Rick
Helpful
Rick Veregin:
So at f1.8 with a 4 nm you would only lose about 15% on the signal due to the transmission loss (ignorning a very slight f ratio changes), and background noise would go down by sqrt(3)=1.7X now compared to 12 nm, so your S/N would improve in the fainter parts of the image by 0.85x1.7 = 1.47X, so this seems like a really good option as far as I can see. You should be able to increase your spacing, the filter effect will be about 2X being 2X thicker, or about 1/3*2mm = 2/3 mm typically (depends also on RI). You can get fine delrin spacers to make this up, Baeder for example has a 0.6 mm T2 spacer which should be close enough?
As for the other filters, up to you how to prioritize. i'm a bit puzzled why you are struggling with LP gradients with a NB filter. I don't really have any issue with my L-eNhance which is 24nm/10 nm wide for Ha/OIII and certainly not for my L-eXtreme, which is 7nm/7nm. The only exception to this being a full moon, or perhaps if I am trying to target something close to a streetlight. So I just plan my work away as much as possible from those sources. I use Startools module to clean the background, not an issue that I can recall. Actually I have far more problems with bad flats causing gradients than LP when doing NB. Do you have a long dew shield, this helps to block stray light? I also have a small makeshift barrier to block a neighbors light when he leaves it on. Others have made a full surround out of sunshield mesh, to more fully block light, and as a bonus wind too.
Good luck with whatever you choose!
CS
Rick
Thanks. Re. the spacing issue - I do have the 9-pc set of spacers from Blue Fireball and will try them with my Baader filters. I've done a fair bit of testing different spacer combinations with my IDAS filter (2.5mm thick) but haven't had much luck. What I'm thinking of doing is shooting the stars separately at, say, f/3, which should make the stars should be round across the frame, using StarXTerminator to remove the stars from the f/1.8 image, and then replacing them with the f/3 stars. StarXTerminator seems to do an excellent job of removing even very distorted stars so I am optimistic that this method should work.
I have always suffered with gradients in OIII, even when I used a 3nm Astrodon filter at f/4.9. I live in a white zone in an ocean of strip malls and major highways. I'm hoping I can combat this with the very fast focal ratio and 4nm filter and careful gradient removal.
Helpful Insightful Respectful Engaging
Sounds like a plan–it is certainly challenging at that fratio to get good stars. Your LP must be even worse than mine, I really feel for you. I'm seriously thinking of never doing wide band with my LP, only NB filters make it bearable.
All the best
Rick
All the best
Rick
I am in Bortle 8/9, so i need 1nm band not just 3nm or 6nm, so i aimed to Ha 5nm in the beginning when i started for NII which was amazing anyway then added Chroma OIII/SII 3nm [all at 1.25" size], all of them are beautifully working for me.
Later when i bought larger sizes sensor as IMX571 APS-C then i added two more filters, Antlia Ha 3nm and then Optolong SII 3nm, now i have to decide between either OIII 3nm or dual band filter of 3nm for OSC camera to get that OIII 3nm from there, i can't look back to wider band than 3nm for OIII and even SII, surprisingly i got more gradient in the past using Optolong SII 6.5nm over Optolong Ha 7nm, that was a big surprise really as i thought that SII is further reddish than Ha and yet it is more noisy, so i made sure to have all 3nm as much as i can and i never regret that really, so if one day i will choose from Astronomik MaxFR i will definitely go with 6nm over 12nm no matter what for my sky Bortle, i also have lenses from f/1.4 up to f/2.8 wide open.
Later when i bought larger sizes sensor as IMX571 APS-C then i added two more filters, Antlia Ha 3nm and then Optolong SII 3nm, now i have to decide between either OIII 3nm or dual band filter of 3nm for OSC camera to get that OIII 3nm from there, i can't look back to wider band than 3nm for OIII and even SII, surprisingly i got more gradient in the past using Optolong SII 6.5nm over Optolong Ha 7nm, that was a big surprise really as i thought that SII is further reddish than Ha and yet it is more noisy, so i made sure to have all 3nm as much as i can and i never regret that really, so if one day i will choose from Astronomik MaxFR i will definitely go with 6nm over 12nm no matter what for my sky Bortle, i also have lenses from f/1.4 up to f/2.8 wide open.
My L Enhance has become mostly unusable over the last two years in my bortle 7 yard, sadly. The LP swamps so much signal that the subs become unusable. The 7 nm L Extreme has been fantastic! 3nm would be optimum. I do have the Antlia duo band 5 nm and contrast is great on that filter wide open on an f 2.8 lens.
I agree with Dale: I never thought I'd see the day I was taking long exposures on my 13" newt using nb
I agree with Dale: I never thought I'd see the day I was taking long exposures on my 13" newt using nb
Helpful Insightful Respectful Engaging
@Rick Veregin I have a Nikon D810A and a 600mm F4 lens, and 2" SVBony 7nm Ha and OIII filters. My home town is coastal and I evaluate the skies as Bortle 7 when humid to a good Bortle 6 when dry, so not too bad but yeah humidity's a pain.
Seeing the previous messages about the Optolong filters, what are the gains if tomorrow I start imaging with a L-Ultimate (or Extreme) at F4 over the 7nm H/O? Is it "just" that a single duo band session doubles the imaging time with obvious gains in terms of integration, and better background "management", or there's something more subtle going on which I have not grasped?
Also, I do notice a small difference in focussing between Ha and OIII and wonder if a L-Ultimate would impact "sharpness" by forcing me to focus "good enough" for both B and R channels at the same time?
Thanks
Seeing the previous messages about the Optolong filters, what are the gains if tomorrow I start imaging with a L-Ultimate (or Extreme) at F4 over the 7nm H/O? Is it "just" that a single duo band session doubles the imaging time with obvious gains in terms of integration, and better background "management", or there's something more subtle going on which I have not grasped?
Also, I do notice a small difference in focussing between Ha and OIII and wonder if a L-Ultimate would impact "sharpness" by forcing me to focus "good enough" for both B and R channels at the same time?
Thanks
Engaging
Ben:
@Rick Veregin I have a Nikon D810A and a 600mm F4 lens, and 2" SVBony 7nm Ha and OIII filters. My home town is coastal and I evaluate the skies as Bortle 7 when humid to a good Bortle 6 when dry, so not too bad but yeah humidity's a pain.
Seeing the previous messages about the Optolong filters, what are the gains if tomorrow I start imaging with a L-Ultimate (or Extreme) at F4 over the 7nm H/O? Is it "just" that a single duo band session doubles the imaging time with obvious gains in terms of integration, and better background "management", or there's something more subtle going on which I have not grasped?
Also, I do notice a small difference in focussing between Ha and OIII and wonder if a L-Ultimate would impact "sharpness" by forcing me to focus "good enough" for both B and R channels at the same time?
Thanks
I am assuming you mean that you currently use separate Ha and OIII filters with your Nikon camera, and are wondering the benefit of going to a duo band, where you collect both Ha and OIII at once.
Say you do one hour each for Ha and then OIII. And say the signal from the target is the same for Ha in R, and OIII in B and G, say 1000 photons/hr.
With the Ha filter in place you get R pixels=1000 photons, but G=0, B=0. With the OIII filter in place, R=0, G=B=1000 photons. After two hours R=G=B=1000 photons.
Say you capture the same two hours with L-eXtreme duo-band (about the same efficiency and contrast as you Sony filters as they are all 7 nm bandwidth).
You get R=G=B=2000 photons. So yes, you get twice the signal, and your S/N improves by sqrt(2)=1.41! Your background noise is thus improved dramatically. All this is the benefit you get.
However, your background itself is just as bright relative to the image signal, you are collecting twice the background as well as twice the signal. So this won't help contrast and gradients. Contrast comes from how narrow your filter is to block LP. I find that the L-eXtreme for me in Bortle 8 gives a nice background generally, no bad gradients, most of my gradients then come from poor flats. and I'm good now with my light-panel flats, I have a procedure that works well.
There is now an L-Ultimate which is 3 nm duo band, which will improve contrast and background LP. But at f/4 it might start cutting into your signal, as the bandpass changes as f-ratio decreases, so one can lose some signal. So not sure how to advise you there, it may still be better than the L-eXtreme for contrast but you will lose at least some signal, just not sure how much you lose for your setup. And it is pricy, I don't see myself going that way, happy at L-eXtreme.
L-eNhance is much broader bands, so contrast is degraded, though you get the same throughput advantage. I find this works well as a simple LP filter, for broadband targets like galaxies for example.
The difference in focusing is because your two filters don't have exactly the same refractive index and thickness, small variations there change your focus slightly when you change filters--the filters are not parfocal with each other. The duoband is one filter, so you don't have to worry about this effect. If your lens is poor quality with poor color focus, that would of course show up as a problem, but assuming you have a good quality lens, I don't see an issue.
Hope this helps.
Rick
Helpful Insightful
Thanks Rick, much appreciated. I thought I understood most of the previous posts but you kindly confirmed it all. 
When looking at the L-Ultimate last night, I did wonder about signal loss but wasn't sure if it would start becoming an issue at F4.
If budget is there, I could consider buying both L-Ultimate and L-eXtreme and spend a couple good nights doing comparisons before selling the least suitable one … I'll see if I can log this expense under "astro mid-life crisis" and convince my partner it's better than spending 20 times more on a sports car!
And noted re focusing, I had a "Duh …" of hindsight when reading your reply … 2 pieces of glass can indeed not be the exact same.
The lens is old, is not apo and lacks modern glass elements like FL, HRI, etc … but I did not notice much fringing on RGB images, at least in the centre half. I should double-check so if duoband imaging brings a "focus compromise", I won't be surprised.
Thanks again.
When looking at the L-Ultimate last night, I did wonder about signal loss but wasn't sure if it would start becoming an issue at F4.
If budget is there, I could consider buying both L-Ultimate and L-eXtreme and spend a couple good nights doing comparisons before selling the least suitable one … I'll see if I can log this expense under "astro mid-life crisis" and convince my partner it's better than spending 20 times more on a sports car!
And noted re focusing, I had a "Duh …" of hindsight when reading your reply … 2 pieces of glass can indeed not be the exact same.
The lens is old, is not apo and lacks modern glass elements like FL, HRI, etc … but I did not notice much fringing on RGB images, at least in the centre half. I should double-check so if duoband imaging brings a "focus compromise", I won't be surprised.
Thanks again.
Respectful Engaging
On the matter of fast F vs. NB filters characteristics, by only skipping through slides of the linked presentation, it's unlikely I need to go narrower than 6nm at F4 (and certainly not at F6.7) and should expect a ~1.2nm blue shift in Ha, and less than that in OIII. Based on Optolong low-res transmission graph, I think it equates to ~10-12% loss in signal (let's say 15% to be safe) from peak transmission.
Now I wonder which one between F4 and F4.5 is the sweeter spot when considering blue shift, signal loss (whatever the source), overall image quality, and ease of processing ... I guess I should buy the L-eX and find out?
https://www.youtube.com/watch?v=rtFX92dZE_w
Now I wonder which one between F4 and F4.5 is the sweeter spot when considering blue shift, signal loss (whatever the source), overall image quality, and ease of processing ... I guess I should buy the L-eX and find out?
https://www.youtube.com/watch?v=rtFX92dZE_w
Ben:
On the matter of fast F vs. NB filters characteristics, by only skipping through slides of the linked presentation, it's unlikely I need to go narrower than 6nm at F4 (and certainly not at F6.7) and should expect a ~1nm blue shift in Ha, and less than that in OIII. Based on Optolong low-res transmission graph, I think it equates to ~10% loss in signal (let's say 15% to be safe).
Now I wonder which one between F4 and F4.5 is the sweeter spot when considering blue shift, signal loss (whatever the source), overall image quality, and ease of processing ... I guess I should buy the L-eX and find out?
https://www.youtube.com/watch?v=rtFX92dZE_w
That as great link, thanks, very informative! As a retired imaging scientist, this is the right approach. Use the data available to choose the most promising path, then experiment to optimize it. A slight increase in f-ratio will not be a huge factor in signal, but could improve any chromatic aberration as well as coma, and giving you nicer stars as you go off axis. But, these subtle optimizations need experimentation. One thing I'm sure of, you will be really happy with the impact of using the duo band filter, only way to go with a full color camera.
Good luck and CS
Rick
Helpful Insightful Respectful Engaging Supportive
I have an f 4.5 and the LX is perfect for my bortle 7. It's a huge difference from the enhance, but YMMV. Contrast is spot on. My usual sub is 300". I am tempted to try the L Ultimate, but am not up to the spending yet, haha!
I thought I would update this thread now that I've done some tests of my own with my 105mm f/1.4 lens and both 12nm and 6nm filters. The results aren't too conclusive so far.
Astronomik states that their 12nm MaxFR filter gets about 85% of the maximum transmittance at f/1.4 and 100% at f/2. Max transmittance varies from filter to filter, but is usually about 96%.
First I tested the 12nm Ha filter. As noted earlier in this thread, I live under severely light polluted skies, about Bortle 7-8. I figured the best initial results could be obtained with Ha for this reason. The 12nm filter suffered under my skies. With a full night of data at f/1.4, the background was nigh unmanageable and there was poor contrast between the nebula and background.
After being disappointed with these results, I reached out to Gerd Neumann to see if he had a 6nm Ha filter that could be used at f/1.4. He found one that had optimal (90%) transmittance between f/1.8 and f/2.8 but would work at f/1.4. He did not say what the actual transmittance at f/1.4 was.
The weather has been terrible here lately, and I've had just one night with poor transparency to test it out. I did one 120s exposure of NGC7000 at f/1.4 and one at f/2. ASI294MM at gain 120 in "bin1" mode.
f/1.4 has less noise both in the background and nebula, but it's not a huge difference. PI's Noise Analysis script gives me 1.09 for f/1.4 and 1.12 for f/2. f/2 has noticeably better contrast. Fainter signals in the outer parts of the nebula pop out better at f/2. f/2 also has the benefit of sharper stars. I don't see any difference in nebula sharpness.
I'm still not sure if I understand the results. f/1.4 has about 2x the light gathering ability with this lens as f/2. Does the loss in transmittance at f/1.4 translate to worse contrast and only slightly better noise?
What does seem certain is that 6nm is better than 12nm under poor conditions.
The below images are autostretched with default ABE settings applied.
f/1.4:
f/2:
Astronomik states that their 12nm MaxFR filter gets about 85% of the maximum transmittance at f/1.4 and 100% at f/2. Max transmittance varies from filter to filter, but is usually about 96%.
First I tested the 12nm Ha filter. As noted earlier in this thread, I live under severely light polluted skies, about Bortle 7-8. I figured the best initial results could be obtained with Ha for this reason. The 12nm filter suffered under my skies. With a full night of data at f/1.4, the background was nigh unmanageable and there was poor contrast between the nebula and background.
After being disappointed with these results, I reached out to Gerd Neumann to see if he had a 6nm Ha filter that could be used at f/1.4. He found one that had optimal (90%) transmittance between f/1.8 and f/2.8 but would work at f/1.4. He did not say what the actual transmittance at f/1.4 was.
The weather has been terrible here lately, and I've had just one night with poor transparency to test it out. I did one 120s exposure of NGC7000 at f/1.4 and one at f/2. ASI294MM at gain 120 in "bin1" mode.
f/1.4 has less noise both in the background and nebula, but it's not a huge difference. PI's Noise Analysis script gives me 1.09 for f/1.4 and 1.12 for f/2. f/2 has noticeably better contrast. Fainter signals in the outer parts of the nebula pop out better at f/2. f/2 also has the benefit of sharper stars. I don't see any difference in nebula sharpness.
I'm still not sure if I understand the results. f/1.4 has about 2x the light gathering ability with this lens as f/2. Does the loss in transmittance at f/1.4 translate to worse contrast and only slightly better noise?
What does seem certain is that 6nm is better than 12nm under poor conditions.
The below images are autostretched with default ABE settings applied.
f/1.4:
f/2:
Helpful Insightful Respectful Engaging Supportive
Interesting, thank you.
i also have and use a 105 1.4 (nikkor in my case) for deep sky. Never have used it with a narrowband filter.
i also have and use a 105 1.4 (nikkor in my case) for deep sky. Never have used it with a narrowband filter.
Interesting, thank you.
i also have and use a 105 1.4 (nikkor in my case) for deep sky. Never have used it with a narrowband filter.
I’m beginning to wonder if it serves any purpose at all from my skies. It seems like it’s better suited for exclusive dark sky work. My intention was to be able to capture Ha from home and then go deep on OIII and RGB at a dark site but that’s starting to seem like a fool’s errand.
I expect the shift @/1.4 to be very high, especially for the further marginal rays and I'm wondering whether you'd better served by moving the filter to the front and accept that the price to be paid for tight NB is to limit to the speed to about f/2.2.
I am thinking about filters that i can use at F1.8 up to F2.8 because i have lenses from F1.4 to F2.8 wide open as fastest lenses i have, i don't care about F4-F6 lenses, and i don't want to stop down the lenses as it is no point to go slower a bit, but in this case i have two choices if i will never stop down the lens:
1. Image in RGB only so the shift isn't an issue, if i understand correctly that the shift in so fast optics is only with narrowbanding and not RGB, correct me.
2. Use a narrowband filter that is design at F2 up to F4, better if from F1.8 even, without much issue, because no matter what i will always or only use RGB stars, i really don't know why people keep insisting to use NB stars rather than RGB stars, so if i will use RGB stars then no matter and whatever issues i will see from NB filters regarding stars will be gone.
So i want to know what it should be here, i do have dual band filter L-Ultimate which is 3nm Ha/OIII and also small sizes filters of Ha 5nm and OIII/SII 3nm in addition to 2" size Ha/SII 3nm filters, but i assume those are only better for maybe F3.5 or F4 and slower, and i live under Bortle 8/9 with LED streetlight only by North direction which is where Polaris and Ursa Major targets are, most targets i want to capture are by East or West which are less LP, South is covered by my house, some targets are almost passing the Zenith which is the lowest LP i see, so not sure if i will have any luck with broadband RGB signals too.
I have mono camera 4 as cooled and one OSC cooled camera, so i am flexible, and if things getting better in future sooner i hope then i might add like 2 more cooled camera, could be both OSC so i can use them both for RGB with separate setup, i really will work on having RGB stars more than NB stars, i hate to do SO LONG processing correcting everything mainly stars, so if the stars are and always the matter thing then i will buy a capable best setup to get RGB stars alone at best, so i will never worry how bad stars from narrowbanding data.
1. Image in RGB only so the shift isn't an issue, if i understand correctly that the shift in so fast optics is only with narrowbanding and not RGB, correct me.
2. Use a narrowband filter that is design at F2 up to F4, better if from F1.8 even, without much issue, because no matter what i will always or only use RGB stars, i really don't know why people keep insisting to use NB stars rather than RGB stars, so if i will use RGB stars then no matter and whatever issues i will see from NB filters regarding stars will be gone.
So i want to know what it should be here, i do have dual band filter L-Ultimate which is 3nm Ha/OIII and also small sizes filters of Ha 5nm and OIII/SII 3nm in addition to 2" size Ha/SII 3nm filters, but i assume those are only better for maybe F3.5 or F4 and slower, and i live under Bortle 8/9 with LED streetlight only by North direction which is where Polaris and Ursa Major targets are, most targets i want to capture are by East or West which are less LP, South is covered by my house, some targets are almost passing the Zenith which is the lowest LP i see, so not sure if i will have any luck with broadband RGB signals too.
I have mono camera 4 as cooled and one OSC cooled camera, so i am flexible, and if things getting better in future sooner i hope then i might add like 2 more cooled camera, could be both OSC so i can use them both for RGB with separate setup, i really will work on having RGB stars more than NB stars, i hate to do SO LONG processing correcting everything mainly stars, so if the stars are and always the matter thing then i will buy a capable best setup to get RGB stars alone at best, so i will never worry how bad stars from narrowbanding data.
andrea tasselli:
I expect the shift @/1.4 to be very high, especially for the further marginal rays and I'm wondering whether you'd better served by moving the filter to the front and accept that the price to be paid for tight NB is to limit to the speed to about f/2.2.
I suspect you're right. Given there's no reliable method for iris control on this lens without using a DSLR, it would seem backyard mono imaging with this lens is just a pipe dream. The Rokinon 135 proves to be the winner again, it seems.
Well Written
andrea tasselli:
I expect the shift @/1.4 to be very high, especially for the further marginal rays and I'm wondering whether you'd better served by moving the filter to the front and accept that the price to be paid for tight NB is to limit to the speed to about f/2.2.
The Dragonfly Array Spectral Line Mapper uses this front filter approach. It allows them to use 1nm wide, tunable filters.
Baader sells standard 6.5 nm filters up to 65mm, you will probably need custom filters for the 105 mm f/1.4?
Helpful