Astronomik 4 nm NB filters (released January 2026) vs. Antlia 3 nm - Forums

Topic starter

New contributor

1 day ago

Does anyone have experience with (or informed thoughts on) the new Astronomik 4 nm filters (released January 2026)? Astronomik claims “no halos” on their website:
https://www.astronomik.com/en/Narrowband-Filters/4nm-Filters/

I am currently deciding between the following options for my first SHO filter set:

Variant 1: S+H+O: ANTLIA 3 nm 2”
Variant 2: S+H: ANTLIA 3 nm 2” + O: Chroma 3 nm 2”
Variant 3: S+H+O: Astronomik 4 nm 2”

My typical focal ratio is f/4.9. Sometimes as bright as f/3.5. I am living under a Bortle 4.7 sky. Regarding variant 2: I don’t plan to use an OAG any time soon. So having SHO filters with different thickness should not be a problem.

My two main concerns are:

Halos that are hard to correct in post
Worse signal to noise with 4 nm vs. 3 nm

Well written Respectful Concise Engaging

Lynn K

21 hours ago

Are you asking about the Astromnomik “MFR" 4nm narrow band filter good down to F4.5 FL? I have the HA & OIII, but have not used them enought to evaluate. I have use the MaxFR Ha & OIII on a Tak FSQ106ED at F3.6 and a Officina Stellare RH200 at F3. There were no halos with those. Others have made similar claimes. I personaly have faith in Astronomik QE to full field its claimes. I only have a Antlia RGB narrow band OSC filter , but have not had an opportunity to use it.

The MaxFR 4nm filter have not been released.

Lynn K.

Tony Gondola

Frequent contributor

Exemplary contributor

20 hours ago

A 3 nm filter is going to give you considerable light loss at F/3.5, so much so it really negates the reason to even be at F/3.5

Well written Concise

andrea tasselli

Frequent contributor

20 hours ago

How do you know? Did you test them and with the OP’s gear too? Besides the disputable thinking of equating FR to SNR…

lumendo

Topic starter

New contributor

20 hours ago

Tony Gondola · Sep 19, 2026, 05:31 PM
A 3 nm filter is going to give you considerable light loss at F/3.5, so much so it really negates the reason to even be at F/3.5

On the ANTLIA website they state: “Blue shift data shows that Antlia 3nm Pro filters can be used with nearly all systems as fast as f/3 with minimal loss in emission signal and meets the requirements of fast optics.”

http://www.antliafilter.com/pd.jsp?fromColId=137&id=67#_pp=137_607

What is your opinion on that?

Well written Respectful

lumendo

Topic starter

New contributor

20 hours ago

Lynn K · Sep 19, 2026, 04:57 PM
Are you asking about the Astromnomik “MFR" 4nm narrow band filter good down to F4.5 FL? I have the HA & OIII, but have not used them enought to evaluate. I have use the MaxFR Ha & OIII on a Tak FSQ106ED at F3.6 and a Officina Stellare RH200 at F3. There were no halos with those. Others have made similar claimes. I personaly have faith in Astronomik QE to full field its claimes. I only have a Antlia RGB narrow band OSC filter , but have not had an opportunity to use it.
The MaxFR 4nm filter have not been released.
Lynn K.

Yes, I am referring tho the 2 nm MFR filters. I just saw that Astronomik states that they can be used with f/4.5 or slower. https://www.astronomik.com/en/Astronomik-OIII-4nm-2-M48/10206200

The 4nm MaxFR are probably the ones I would need for maximum SNR. But as you said, they are not yet released. I wonder how they realize the compatiblity with faster f- ratios. They state “Unlike other manufacturers' filters, which often require special “High-Speed” versions, Astronomik’s MaxFR filters maintain their transmission characteristics even with normal optical systems.”. (https://www.astronomik.com/en/Narrowband-Filters/MaxFR-Filters/) This sounds like they don’t just slightly blue shift the transmission window.

Did you youse the MaxFR 6 nm version?

Concise

Tony Gondola

Frequent contributor

Exemplary contributor

20 hours ago

I calculated the offset. I can’t take the time to look up the curves for all the filters mentions and plot the shifts. What I’ve calculated is based on the stated bandpass of the filters, the stated geometeric F ratio and a generous guess as to the refractive index of the glass. For instance, I for Ha I get a fequency shift of 1.64 nm. That gives a transmission loss of more than 50% with the 3nm Antila filter. What number do you get?

Concise

Tony Gondola

Frequent contributor

Exemplary contributor

20 hours ago

Dominik Floess · Sep 19, 2026, 06:08 PM
Tony Gondola · Sep 19, 2026, 05:31 PM
A 3 nm filter is going to give you considerable light loss at F/3.5, so much so it really negates the reason to even be at F/3.5
On the ANTLIA website they state: “Blue shift data shows that Antlia 3nm Pro filters can be used with nearly all systems as fast as f/3 with minimal loss in emission signal and meets the requirements of fast optics.”

http://www.antliafilter.com/pd.jsp?fromColId=137&id=67#_pp=137_607

What is your opinion on that?

I guess you can test that for yourself. Grab the plot for the filter and plot a line at 654.6 nm. That’s the position of the Ha line in an F/3.5 light cone.

Tony Gondola

Frequent contributor

Exemplary contributor

19 hours ago

If anyone wants to dig into the fast optics, filter bandpass question, you might enjoy this write up by Jim Thompson:

http://karmalimbo.com/aro/reports/Article%20-%20Narrowband%20Filters%20&%20Fast%20Optics_Nov2020.pdf

andrea tasselli

Frequent contributor

19 hours ago

Tony Gondola · Sep 19, 2026, 06:17 PM
I calculated the offset. I can’t take the time to look up the curves for all the filters mentions and plot the shifts. What I’ve calculated is based on the stated bandpass of the filters, the stated geometeric F ratio and a generous guess as to the refractive index of the glass. For instance, I for Ha I get a fequency shift of 1.64 nm. That gives a transmission loss of more than 50% with the 3nm Anila filter. What number do you get?

Assuming it’s me you’re replaying to, did you calculate the integral over all the FRs weighted against the relative areas? Because if you didn’t you didn’t get the right answer. My calculation for the filter in question (here Ha) is that the effect is 8% transmission loss.

lumendo

Topic starter

New contributor

19 hours ago

Tony Gondola · Sep 19, 2026, 06:22 PM
Dominik Floess · Sep 19, 2026, 06:08 PM
Tony Gondola · Sep 19, 2026, 05:31 PM
A 3 nm filter is going to give you considerable light loss at F/3.5, so much so it really negates the reason to even be at F/3.5
On the ANTLIA website they state: “Blue shift data shows that Antlia 3nm Pro filters can be used with nearly all systems as fast as f/3 with minimal loss in emission signal and meets the requirements of fast optics.”

http://www.antliafilter.com/pd.jsp?fromColId=137&id=67#_pp=137_607

What is your opinion on that?
I guess you can test that for yourself. Grab the plot for the filter and plot a line at 654.6 nm. That’s the position of the Ha line in an F/3.5 light cone.
Tony Gondola · Sep 19, 2026, 07:35 PM
Dominik Floess · Sep 19, 2026, 06:55 PM
Tony Gondola · Sep 19, 2026, 06:22 PM
Dominik Floess · Sep 19, 2026, 06:08 PM
Tony Gondola · Sep 19, 2026, 05:31 PM
A 3 nm filter is going to give you considerable light loss at F/3.5, so much so it really negates the reason to even be at F/3.5
On the ANTLIA website they state: “Blue shift data shows that Antlia 3nm Pro filters can be used with nearly all systems as fast as f/3 with minimal loss in emission signal and meets the requirements of fast optics.”

http://www.antliafilter.com/pd.jsp?fromColId=137&id=67#_pp=137_607

What is your opinion on that?
I guess you can test that for yourself. Grab the plot for the filter and plot a line at 654.6 nm. That’s the position of the Ha line in an F/3.5 light cone.
📷 image.pngSo the rays with the largest angle of incidence have ~70% of light tramsmission. All the other light rays have a smaller AOI and thus higher transmission. So if I neglect any fabrication tolerance on the central wavelength of the transmission window, there seems to be still some significant light gathering benefit of f/3.5 over f/4.9. Although, as you mentioned, I would not get the full amount of light potentially available by f/3.5.
I might be wrong but I belive most of the energy is coming from the outer area of the apature, where most of the area is. I belive the write up I posted the link to talks about that. As you go further off ideal frequency it’s like a reduction in apature in terms of radience because the steeper rays are attenuated the most.
Tony Gondola · Sep 19, 2026, 07:35 PM
Dominik Floess · Sep 19, 2026, 06:55 PM
Tony Gondola · Sep 19, 2026, 06:22 PM
Dominik Floess · Sep 19, 2026, 06:08 PM
Tony Gondola · Sep 19, 2026, 05:31 PM
A 3 nm filter is going to give you considerable light loss at F/3.5, so much so it really negates the reason to even be at F/3.5
On the ANTLIA website they state: “Blue shift data shows that Antlia 3nm Pro filters can be used with nearly all systems as fast as f/3 with minimal loss in emission signal and meets the requirements of fast optics.”

http://www.antliafilter.com/pd.jsp?fromColId=137&id=67#_pp=137_607

What is your opinion on that?
I guess you can test that for yourself. Grab the plot for the filter and plot a line at 654.6 nm. That’s the position of the Ha line in an F/3.5 light cone.
📷 image.pngSo the rays with the largest angle of incidence have ~70% of light tramsmission. All the other light rays have a smaller AOI and thus higher transmission. So if I neglect any fabrication tolerance on the central wavelength of the transmission window, there seems to be still some significant light gathering benefit of f/3.5 over f/4.9. Although, as you mentioned, I would not get the full amount of light potentially available by f/3.5.
I might be wrong but I belive most of the energy is coming from the outer area of the apature, where most of the area is. I belive the write up I posted the link to talks about that. As you go further off ideal frequency it’s like a reduction in apature in terms of radience because the steeper rays are attenuated the most.

I had to remove the graph in my last post as I found out that the transmission spectrum I shared might actuall not be correct. It is surprisingly hard to find an official spectrum for the antlia 3 nm H-alpha. However, I understand your reasoning behind the wavelength shift and the transmission loss.

Concise

Tony Gondola

Frequent contributor

Exemplary contributor

19 hours ago

andrea tasselli · Sep 19, 2026, 06:48 PM
Tony Gondola · Sep 19, 2026, 06:17 PM
I calculated the offset. I can’t take the time to look up the curves for all the filters mentions and plot the shifts. What I’ve calculated is based on the stated bandpass of the filters, the stated geometeric F ratio and a generous guess as to the refractive index of the glass. For instance, I for Ha I get a fequency shift of 1.64 nm. That gives a transmission loss of more than 50% with the 3nm Anila filter. What number do you get?
Assuming it’s me you’re replaying to, did you calculate the integral over all the FRs weighted against the relative areas? Because if you didn’t you didn’t get the right answer. My calculation for the filter in question (here Ha) is that the effect is 8% transmission loss.

What was your result for the shift in nanometers?

Tony Gondola

Frequent contributor

Exemplary contributor

18 hours ago

Dominik Floess · Sep 19, 2026, 06:55 PM
Tony Gondola · Sep 19, 2026, 06:22 PM
Dominik Floess · Sep 19, 2026, 06:08 PM
Tony Gondola · Sep 19, 2026, 05:31 PM
A 3 nm filter is going to give you considerable light loss at F/3.5, so much so it really negates the reason to even be at F/3.5
On the ANTLIA website they state: “Blue shift data shows that Antlia 3nm Pro filters can be used with nearly all systems as fast as f/3 with minimal loss in emission signal and meets the requirements of fast optics.”

http://www.antliafilter.com/pd.jsp?fromColId=137&id=67#_pp=137_607

What is your opinion on that?
I guess you can test that for yourself. Grab the plot for the filter and plot a line at 654.6 nm. That’s the position of the Ha line in an F/3.5 light cone.
📷 image.pngSo the rays with the largest angle of incidence have ~70% of light tramsmission. All the other light rays have a smaller AOI and thus higher transmission. So if I neglect any fabrication tolerance on the central wavelength of the transmission window, there seems to be still some significant light gathering benefit of f/3.5 over f/4.9. Although, as you mentioned, I would not get the full amount of light potentially available by f/3.5.

I might be wrong but I belive most of the energy is coming from the outer area of the apature, where most of the area is. I belive the write up I posted the link to talks about that. As you go further off ideal frequency it’s like a reduction in apature in terms of radience because the steeper rays are attenuated the most. Adding a central obstruction makes it worse.

lumendo

Topic starter

New contributor

6 hours ago

I will mostly capture with f/4.9. So transmission loss should be OK.

Most concerning to me is the formation of halos that are difficult to correct in post. Which of the three variants ist most sensible?

Variant 1: S+H+O: ANTLIA 3 nm 2”
Variant 2: S+H: ANTLIA 3 nm 2” + O: Chroma 3 nm 2”
Variant 3: S+H+O: Astronomik 4 nm 2”

Concise

lumendo

Topic starter

New contributor

3 hours ago

Tony Gondola · Sep 19, 2026, 07:35 PM
Dominik Floess · Sep 19, 2026, 06:55 PM
Tony Gondola · Sep 19, 2026, 06:22 PM
Dominik Floess · Sep 19, 2026, 06:08 PM
Tony Gondola · Sep 19, 2026, 05:31 PM
A 3 nm filter is going to give you considerable light loss at F/3.5, so much so it really negates the reason to even be at F/3.5
On the ANTLIA website they state: “Blue shift data shows that Antlia 3nm Pro filters can be used with nearly all systems as fast as f/3 with minimal loss in emission signal and meets the requirements of fast optics.”

http://www.antliafilter.com/pd.jsp?fromColId=137&id=67#_pp=137_607

What is your opinion on that?
I guess you can test that for yourself. Grab the plot for the filter and plot a line at 654.6 nm. That’s the position of the Ha line in an F/3.5 light cone.
📷 image.pngSo the rays with the largest angle of incidence have ~70% of light tramsmission. All the other light rays have a smaller AOI and thus higher transmission. So if I neglect any fabrication tolerance on the central wavelength of the transmission window, there seems to be still some significant light gathering benefit of f/3.5 over f/4.9. Although, as you mentioned, I would not get the full amount of light potentially available by f/3.5.
I might be wrong but I belive most of the energy is coming from the outer area of the apature, where most of the area is. I belive the write up I posted the link to talks about that. As you go further off ideal frequency it’s like a reduction in apature in terms of radience because the steeper rays are attenuated the most. Adding a central obstruction makes it worse.

Chroma shares CSV files for the transmission data of their 3 nm filters and different focal ratios (https://www.chroma.com/products/sets/27106-sii-ha-oiii-3nm/). I plotted them in the following graph:
📷 image.pngI assume they calculate the effective transmission over the full bundle of light rays. These filters appear to produce almost no transmission loss for telescopes as fast as f4. For f/3 the effective transmission goes down to ~70% for Ha and SII but stays higher for OIII.

I would really like to see such a graph for antlia 3 nm filters.

Well written Helpful Respectful Concise Engaging

Ralf Dinkelmeyer

2 hours ago

I have measured those filters with my highresolution spectrometer: spica-observatory.de/astrofotografie/schmalbandfilter-im-schnellen-strahlengang-monobandfilter-am-hα-band/

📷 grafik.png📷 grafik.png📷 grafik.png

Brian

1 hour ago

Hi @Ralf Dinkelmeyer

you tested the Astronomik maxFR 4mn in your Artikel on your website.
But where you bought the filter?

On the Astronomik webpage it says “4nm filters with MaxFR coating are currently in development”.
It seems that you can only buy the normal MFR series at the moment.
Or did you get the filters for testing from Gerd?

The filters are looking quiet interesting 😄

Greetings from Germany
Brian

Respectful

lumendo

Topic starter

New contributor

26 minutes ago

@Ralf Dinkelmeyer Great Article! I really enjoyed reading it (we share the same mother language). Thanks for your effort and for sharing! Such resources are really valuable.

I have the following questions:

Do you have data on the Antlia 3 nm filters? On the Antlia Website they provide some charts that suggest that the emission lines are in the middle of the transmission windows (see https://app.astrobin.com/forum/topic/255193?post=376952). This would not make any sense as they would not be pre-compensated for any angles of incidence larger than 0°. Do you think that they show the effective transmission spectra (taking all the different angles of incidence into account and assume a certain focal ratio)?
Because of this data I am hesitant to buy those filters as I do not want to purchase them with such a design flaw. The fact that ANTLIA ist not really proactively showing these spectra their website makes me also doubt their validity (they are quite hidden on their website and only provided for 50 mm unmounted filters).
Halos: Based on your extensive experience with the topic, could you assess which of the following combinations would be a sensible combination for obtaining only minor halos (meaning straigtforward removal in post processing)? Ideally, could you rank them in therms of expected halo intensity and explain why? I have never really found consitent explanations on what properties determine halo intensity.
- Variant 1: S+H+O: ANTLIA 3 nm 2”
- Variant 2: S+H: ANTLIA 3 nm 2” + O: Chroma 3 nm 2”
- Variant 3: S+H+O: Astronomik 4 nm 2” (MFR)
- Variant 4: S+H+O: Astronomik 4 nm 2” (MaxFR)

Well written Respectful