Photometric filter for imaging nebulae - Forums

Topic starter

Jul 21, 2025

Is there any reason why I shouldn’t use a photometric V filter is a replacement for a Lum filter in an LRGB image?? Well it would be VRGB in this case …..

With an ever changing work schedule and the outlook this week is pointing to finishing just on sunset and the forecast is for a string of clear nights ….. I want to maximise my imaging time so
if I don’t need to swap out my photometric V filter for a Lum filter then I won’t but is there any reason why I can’t or shouldn’t use one for general imaging use??

Walter Leonhard Schramböck

Jul 21, 2025

The main goal of a luminance filter is to represent the whole spectrum (more or less) and is being used primarily to enhance details in your image. The v filter represents mostly the blue/green spectrum, why would you like to cut out the rest of the colours from luminance? If you combine this with RGB you would overemphasize blue and green.
Here is an interesting blog entry on baader's website how you could make use of photometric filters in combination with LRGB for enhancing dark nebulae:

https://www.baader-planetarium.com/en/blog/baader-photometric-filters-dark-nebulae-appear-in-a-new-light/

Helpful Insightful Respectful

andrea tasselli

Frequent contributor

Jul 21, 2025

Sam Badcock:
Is there any reason why I shouldn’t use a photometric V filter is a replacement for a Lum filter in an LRGB image?? Well it would be VRGB in this case …..

With an ever changing work schedule and the outlook this week is pointing to finishing just on sunset and the forecast is for a string of clear nights ….. I want to maximise my imaging time so
if I don’t need to swap out my photometric V filter for a Lum filter then I won’t but is there any reason why I can’t or shouldn’t use one for general imaging use??

V isn't even close to L, in fact is somewhat short of bandwidth w.r.t. G.

Sam Badcock

Topic starter

Jul 21, 2025

andrea tasselli:
Sam Badcock:
Is there any reason why I shouldn’t use a photometric V filter is a replacement for a Lum filter in an LRGB image?? Well it would be VRGB in this case …..

With an ever changing work schedule and the outlook this week is pointing to finishing just on sunset and the forecast is for a string of clear nights ….. I want to maximise my imaging time so
if I don’t need to swap out my photometric V filter for a Lum filter then I won’t but is there any reason why I can’t or shouldn’t use one for general imaging use??

V isn't even close to L, in fact is somewhat short of bandwidth w.r.t. G.

I think I was getting my transmission images mixed up …… thinking they were similar!!

Sam Badcock

Topic starter

Jul 21, 2025

Walter Leonhard Schramböck:
The main goal of a luminance filter is to represent the whole spectrum (more or less) and is being used primarily to enhance details in your image. The v filter represents mostly the blue/green spectrum, why would you like to cut out the rest of the colours from luminance? If you combine this with RGB you would overemphasize blue and green.
Here is an interesting blog entry on baader's website how you could make use of photometric filters in combination with LRGB for enhancing dark nebulae:

https://www.baader-planetarium.com/en/blog/baader-photometric-filters-dark-nebulae-appear-in-a-new-light/

Yeah I think I got my transmission images mixed up thinking they were very similar!!

I’ll give that article a read over soon, but for now we’ll swap them out

Tony Gondola

Frequent contributor

Jul 21, 2025

A V filter from zero to zero passes the same frequencies as a classic uv/ir cut filter does. The difference is that for a uv/ir, the ends of the passband are very steep. For a V filter the pass band on the right side is very gently sloping. That means that overall, you'll get less light to the sensor using a V filter verses an L. Since the purpose of a luminance filter to to gather as much light as possible within the RGB range, a standard L filter will get you more signal. You can use a V filter for that role but it doesn't make any sense in terms of classic LRGB imaging. If you want to maximize your imaging time go with L. Not sure what the deal is because it only takes a few seconds to change filters and focus from offsets anyway.

Helpful Insightful

andrea tasselli

Frequent contributor

Jul 21, 2025

This is the band-pass info for V (for Visual) Johnson-Cousin filter:

So it does not cover the 420-680 nm pass band of typical L filters. Let alone the FWHM which is around 90 NM, against around 260 nm for a typical L filter. Peak transmission is also quite low compared with an L filter (~74% vs., 95%). All this because the filter tries to replicate the human eye response to light since it was used to cross-correlated the sensor/film photometric response to earlier visual observations (or even current).

Helpful Insightful

Walter Leonhard Schramböck

Jul 21, 2025

How could your image benefit from bamdpass-restriction if you do not want to show a specific band?

Sam Badcock

Topic starter

Jul 21, 2025

Walter Leonhard Schramböck:
How could your image benefit from bamdpass-restriction if you do not want to show a specific band?

It wouldn’t ….. this is definitely a case of “can I get away with being lazy” but in the same sense, curiosity

Walter Leonhard Schramböck

Jul 21, 2025

Sam Badcock:
Walter Leonhard Schramböck:
How could your image benefit from bamdpass-restriction if you do not want to show a specific band?

It wouldn’t ….. this is definitely a case of “can I get away with being lazy” but in the same sense, curiosity

Maybe trying it would bring clarity.

Freestar8n

Jul 22, 2025

The purpose of a Luminance signal in LRGB imaging is to create a perceptually pleasing result with good perceived detail by taking advantage of the way perceptual luminance and perceptual color are processed by the human visual system. But it is all based on perceptual signals. In the case of perceived luminance, that is a signal that is strongly peaked in the green and very weak in the blue. So if you want to apply this LRGB trick to capture more signal in L, you need to use perceived luminance.

In amateur astro imaging, the photometric V filter is a good match to perceived luminance. A typical green filter is also a good match, except it doesn't capture much red. But the weak presence of blue in the signal is critical to having blue rendered with the correct perceived luminance in the final result - or else it will be desaturated and washed out.

People seem to have a hard time accepting this since it is so ingrained that the "L" filter captures a signal called "luminance" that applies in the LRGB process - but it simply doesn't. And it does bad things to colors as a result. I prefer to call the "L" filter a "T" filter representing total flux over the visual spectrum - and people are effectively doing TRGB imaging - not LRGB imaging. It isn't just a matter of semantics or purist requirement - it's using a signal that may be strong but is not well suited to the goal of capturing good color and detail. It will fundamentally wash out blue regions and, to a lesser extent, red ones.

So - using a V filter instead of L should work well for capturing more detail while at the same time preventing the saturation of faint blue structures. But it probably wouldn't have any benefit over pure RGB imaging in a given total imaging time. Instead, I have recommended for a long time that people at least try RGGB imaging - since that is guaranteed not to distort the color, but it will enhance the capture of perceived luminance via the extra green. And it's no coincidence that is how OSC is designed to work - for precisely the same goals.

The basic concept behind LRGB imaging is used all over the place, including in image compression. But in those applications, it is the actual perceived luminance signal that is being used. I'm not aware of TRGB being used for anything except in the amateur astro world. And a lot of people are switching over to RGB and dropping the T ("L").

Using a T filter to capture a monochrome gray scale image of a scene would be great for capturing high SNR and detail - but only if it isn't combined with color information. As a proxy for luminance it will fight with the actual luminance signal carried within the RGB. The RGB luminance is correct and the T luminance is wrong - in terms of the perceptual result.

If you want more signal in red you could simply label the "L" filter with an "R" instead - and you would get much more red signal. Labeling total visual flux as "Perceived Luminance" is no different… It's not the signal it needs to be in order to end up with good color and detail.

In the end, "LRGB" is just another example of no free lunch. But I think RGGB could be an example of a good lunch at a discount…

Frank

Well Written Helpful

Sam Badcock

Topic starter

Jul 22, 2025

The purpose of a Luminance signal in LRGB imaging is to create a perceptually pleasing result with good perceived detail by taking advantage of the way perceptual luminance and perceptual color are processed by the human visual system. But it is all based on perceptual signals. In the case of perceived luminance, that is a signal that is strongly peaked in the green and very weak in the blue. So if you want to apply this LRGB trick to capture more signal in L, you need to use perceived luminance.

In amateur astro imaging, the photometric V filter is a good match to perceived luminance. A typical green filter is also a good match, except it doesn't capture much red. But the weak presence of blue in the signal is critical to having blue rendered with the correct perceived luminance in the final result - or else it will be desaturated and washed out.

People seem to have a hard time accepting this since it is so ingrained that the "L" filter captures a signal called "luminance" that applies in the LRGB process - but it simply doesn't. And it does bad things to colors as a result. I prefer to call the "L" filter a "T" filter representing total flux over the visual spectrum - and people are effectively doing TRGB imaging - not LRGB imaging. It isn't just a matter of semantics or purist requirement - it's using a signal that may be strong but is not well suited to the goal of capturing good color and detail. It will fundamentally wash out blue regions and, to a lesser extent, red ones.

So - using a V filter instead of L should work well for capturing more detail while at the same time preventing the saturation of faint blue structures. But it probably wouldn't have any benefit over pure RGB imaging in a given total imaging time. Instead, I have recommended for a long time that people at least try RGGB imaging - since that is guaranteed not to distort the color, but it will enhance the capture of perceived luminance via the extra green. And it's no coincidence that is how OSC is designed to work - for precisely the same goals.

The basic concept behind LRGB imaging is used all over the place, including in image compression. But in those applications, it is the actual perceived luminance signal that is being used. I'm not aware of TRGB being used for anything except in the amateur astro world. And a lot of people are switching over to RGB and dropping the T ("L").

Using a T filter to capture a monochrome gray scale image of a scene would be great for capturing high SNR and detail - but only if it isn't combined with color information. As a proxy for luminance it will fight with the actual luminance signal carried within the RGB. The RGB luminance is correct and the T luminance is wrong - in terms of the perceptual result.

If you want more signal in red you could simply label the "L" filter with an "R" instead - and you would get much more red signal. Labeling total visual flux as "Perceived Luminance" is no different... It's not the signal it needs to be in order to end up with good color and detail.

In the end, "LRGB" is just another example of no free lunch. But I think RGGB could be an example of a good lunch at a discount...

Frank

A lot of this went over my head but I half followed along!!

I was going to be using the L or V filter for getting more signal for dusty regions, dark nebulae, IFN etc to enhance these features

Tony Gondola

Frequent contributor

Jul 22, 2025

IFN is a broadband target so the way to get the maximum flux would be to shoot with no filter at all. I've found this to be very effective with galaxies when compared against using an L filter. Of course, for this to be effective your camera should have good extended near IR sensitivity and you need to use all reflecting optics to get a sharp result.

Well Written Helpful Insightful Concise

Freestar8n

Jul 23, 2025

If the IFN is slightly bluish and you are trying to capture that color, then using an "L" filter or "C" for luminance will wash out the color and steer it towards gray - and it will be particularly bad if there is any IR contributed when using a C filter.

If you are just doing many, many hours of RGB then the average color of a blue object might be (0, 0, 95) after one hour, then (0, 0, 103) after two hours - and over time the average would settle down to the true average value of (0, 0, 100). That is the long term average you are aiming for - and it is a pure, deep blue.

But since blue contributes so little to the perceived luminance, it registers as a very dark color to the eye. But that's fine because that's what it is.

The perceived luminance for a given rgb triplet is approximately L = 0.21R + 0.71G + 0.072B, so green constitutes 70% of perceived luminance, while blue is only 7% and ten times weaker than green.

So the perceived luminance of (0, 0, 100) would be only about 7, while the T signal (from an "L" filter) would be 0 + 0 + 100 = 100.

If you then use photoshop or PI and say, "I want the color here to come from (0, 0, 100) but replace the luminance with 100 instead of 7" - the only way it can try to keep the same hue but boost the perceived luminance is to boost the values in R and G - making it brighter, but also desaturating it and making it gray.

You can get a feel for how this works with tools such as https://colordesigner.io/convert/labtorgb

If you enter an RGB value of 0, 0, 100 you will see it is a deep blue and you can calculate the Lab values for it. If you then take those Lab values and convert back to RGB you will get 0, 0, 100 as expected. But if you increase the L value a bit you will see the R and G values rising to meet the demand - and also see the color being washed out.

This is something I noticed many years ago when I first was trying LRGB, and since I was trying for good color and good detail - the loss of color was frustrating. I wish I had known then what I know now - that the whole idea is fundamentally flawed since the T signal does not work well as a proxy for luminance.

Using a V filter as a proxy for luminance should be ok, and the fact that it only captures blue very weakly is a good thing - because the luminance of a deep blue object really is small. But it won't gain you anything by doing VRGB compared to just full time on RGB.

LRGB makes some sense if you are mainly after detail and don't care much about color - particularly blue. But otherwise if you are after faint blue parts of a galaxy or wisps of nebulosity - pure RGB would be preferable.

Frank

Helpful Insightful