Back to forums

A broadband filter question about NB filters

Filters Acquisition Broadband H-alpha Monochrome Narrowband RGB Deep Sky
6 replies•182 views
Tim Ray avatar
Tim Ray avatar

Speaking on the phone with another member of the community, we got into a discussion about filters and signal strength. Our scenario is uses a mono camera, a 2” Red filter used in RGB imaging and a 5-7 nm Ha filter.

If the following statements are true:

The Ha notch filter falls within the bandwidth of the Red filter.

A photon that is emitted or a photon that is reflected is still a photon.

Then:

The Red filter should show the same signal level as the Ha filter on a target, emitted or reflected. I know that is not the case, if it was, we would all see same Ha levels in Red and Ha filters….

My question is why?

Engaging
Arun H avatar
Arun H avatar
For most objects, including emission nebulae, the emitted or reflected light in red is not just emission line H-alpha but a broad range of wavelengths from reflected broadband starlight as well as other line emissions in the red range (such as nitrogen and SII). That’s why the signal strengths are different.
Well Written Helpful Insightful Concise
Tim Ray avatar
Tim Ray avatar

Thanks for the reply. But if the broadband red filter has a broader range of frequencies which includes the Ha, would not then the red filter have a the Ha + everything else = at least the same as a Ha filter if not higher or more signal. Or are you suggesting that the additional frequencies the red filter passes somehow opaques the Ha. I understand that a NB filter blocks out everything but the Ha, Oiii etc. I just don’t understand why when I use a broad band filters (LRGB) why they do not collect all of the NB + Broadband within the range of the broadband filter at the same level a NB filter delivers on its own…

We spent a long time on the phone having this thought experiment…

Médéric Hébert avatar
Médéric Hébert avatar

Tim Ray · Nov 20, 2025, 05:43 AM

I just don’t understand why when I use a broad band filters (LRGB) why they do not collect all of the NB + Broadband within the range of the broadband filter at the same level a NB filter delivers on its own…

It does, look at the mean pixel value, R will always be higher than Ha assuming the same condition and same camera setting. The reason R doesn’t show nearly as much detail is because the background level is raised more than the signal.

Charles Hagen avatar
Charles Hagen avatar

Tim Ray · Nov 20, 2025, 05:43 AM

But if the broadband red filter has a broader range of frequencies which includes the Ha, would not then the red filter have a the Ha + everything else = at least the same as a Ha filter if not higher or more signal.

In short, it does. If you use the same gain and exposure and assume both to be ideal filters (100% transmission) the Red image will always contain more absolute signal than the Ha. Here’s the thing though, that does not mean it will have better Signal to Noise Ratio (SNR). Much of the signal that is passed by the broadband red filter will be light pollution, while much less light pollution passes through the Ha filter. (This could be artificial, natural, celestial or terrestrial) A key part of this is understanding that light follows a Poisson distribution. This is a relationship that describes the expected distribution of light values for a mean photon flux - it has a special property that it’s RMS error is equal to the square root of the mean value. So if we look at two situations, one where we have 60 photons or “Real signal”, ie. the stuff in space we are trying to image. 50 photons are Ha emission and 10 are some broadband reflection. Now let’s assume that only 5 photons of light pollution make it through the Ha filter but 100 make it through the broadband filter. Now we have two different SNRs to look at:

Red: 60 photons of signal / sqrt(60 photons of signal + 100 photons of LP) = ~4.74
Ha: 50 photons of signal / sqrt(50 photons of signal + 5 photons of LP) = ~6.74

Despite the Red channel having more signal, 20% more even, it will look markedly worse than the Ha which now has a higher signal to noise ratio. This can make it look like the Red has less signal, but it is in fact still greater

Well Written Helpful Insightful Engaging
Arun H avatar
Arun H avatar
Tim Ray:
We spent a long time on the phone having this thought experiment…


What Charles Hagen said above is correct.

You are conflating Signal with Signal to Noise.

The Red Filter will always have more Signal than the H- alpha, but lower signal to noise from a light polluted site because the noise from the broad range of wavelengths overwhelms the relatively weak DSO signal from those same wavelengths.

Using the narrow band filter cuts out most of the noise. It also cuts out the signal from most wavelengths except specific line emissions that its bandpass allows. Therefore, the SNR is increased - at the cost of also rejecting broadband signal. So lower signal, but higher SNR. Note that this only works when the DSO emits strongly in the specific emission wavelength(s).

You can see this for yourself if you use a red filter and H-alpha on a predominantly reflection nebula (like the blue dust of IC 405 or on M 45). The narrowband filter will have a very hard time capturing much of anything, because the signal is so weak. But narrowband filters will work very well on the emission parts of IC 405, as an example.
Well Written Helpful Insightful Engaging
Tim Ray avatar
Tim Ray avatar

Thank you to everyone that responded. All of you have given great info and I now feel I have good idea what is actually happening… It was a great thought experiment that JH and I were discussing. I am glad all of you helped to shed light on our filter question…

Peace and Clear Skies!

Tim