Is any LP filter better than none?

[smr]

I'm really thinking I ought to add a LP filter to my optical train as I image in bortle 6 skies from my back garden. 

I live about a 10 minute drive from the city centre in a village, we have a lot of LED street lamps now, most of the street lighting has been replaced with these, but there are some of the old lamps on a few streets around the village. Therefore the IDAS D2 filter would seem to be the logical choice, but it's very expensive at around £200, and no one in the UK has had any stock for a couple of months now. 

So I am wondering if any light pollution suppression filter is better than none. There are targets I really want to get some imaging time on this month. I was wondering if something like the Skytech CLS filter would at least filter out the airglow and help.

These are a couple of what my recent stacked images look like if it helps.

 

m42

rosette

 

m42 with small linear stretch and rough rgb alignment, with gradient.

 

 

with gradient xterminator applied

 

 

So do you think I should get an LP filter, will it be a waste of time buying any LP filter, or is it pointless unless I go for the IDAS D2?

I wouldn't say my light pollution is severe, from the back garden I cannot directly see any LED lights at all. I just see a wash of light pollution airglow really.

 

Thanks for any advice.

[knobby]

Great question, I've read that lots of integration time can do the same as an LP filter but would love to hear opinions... either way LP filter means longer exp needed as does ... longer exposure for longer integration time !

[smr]

2 minutes ago, knobby said:

Great question, I've read that lots of integration time can do the same as an LP filter but would love to hear opinions... either way LP filter means longer exp needed as does ... longer exposure for longer integration time !

At the moment I really can't go past 60 second exposures with 430mm on my HEQ5 unguided, 2 minutes and I'm throwing away half my subs. Is there a mathematical calculation to determine how much longer my sub frames would need to be if I did add an LP filter? ie. if the exposure length is then 90 seconds I *may* be able to do that, but if it means having to expose for 2 minutes then I am will need to auto guide.

But I am digressing somewhat, the main point is do I really need an IDAS D2, or would I see a good enough benefit in using anyother LP filter.

[vlaiv]

Not easy question to answer really.

Like anything else, quality of image depends on signal/noise ratio.

Filters block light. This reduces signal. One part of noise in above signal/noise ratio comes from light pollution. This is the part that hurts images. It is not background light level - this can be easily processed out with setting black point, and doing background removal if there are gradients (and there usually are since LP depends on elevation, and FOV rotates with respect to horizon).

If LP filter blocks enough LP signal (thus reducing associated shot noise), while at the same time blocking small part of target signal, so that overall signal to noise ratio increases - it's worth using.

If LP filter blocks more signal from target than it reduces LP signal (and associated shot noise) - then it will actually hurt your image.

Why NB filters work so good in high LP? Because they don't block any of target signal (in specific wavelength) while cutting most of LP signal (which is broad band).

So in order to say if particular LP filter is worth using, one should have spectral response from that filter, camera (and any other filters used) and spectrum of target and that of LP. We lack most of this information so there is no real way to say if filter is good other then trying it out. Even trying it out is going to be based on "subjective measurement". If you have enough experience without filter and know what sort of quality of image you can expect under your LP and given imaging time for most targets, and you find that by using filter you get better results in same time - then it's worth using.

Just one more note - effectiveness of LP filter depends on selected target too. For example emission nebulae rich in Ha, OIII, SII and such will benefit from UHC or CLS type of filter. Broadband target like galaxies might not benefit from these filters as they are very aggressive in amount of light they block.

[knobby]

Hope fully someone local to you may have an idea

[smr]

52 minutes ago, vlaiv said:

Not easy question to answer really.

Like anything else, quality of image depends on signal/noise ratio.

Filters block light. This reduces signal. One part of noise in above signal/noise ratio comes from light pollution. This is the part that hurts images. It is not background light level - this can be easily processed out with setting black point, and doing background removal if there are gradients (and there usually are since LP depends on elevation, and FOV rotates with respect to horizon).

If LP filter blocks enough LP signal (thus reducing associated shot noise), while at the same time blocking small part of target signal, so that overall signal to noise ratio increases - it's worth using.

If LP filter blocks more signal from target than it reduces LP signal (and associated shot noise) - then it will actually hurt your image.

Why NB filters work so good in high LP? Because they don't block any of target signal (in specific wavelength) while cutting most of LP signal (which is broad band).

So in order to say if particular LP filter is worth using, one should have spectral response from that filter, camera (and any other filters used) and spectrum of target and that of LP. We lack most of this information so there is no real way to say if filter is good other then trying it out. Even trying it out is going to be based on "subjective measurement". If you have enough experience without filter and know what sort of quality of image you can expect under your LP and given imaging time for most targets, and you find that by using filter you get better results in same time - then it's worth using.

Just one more note - effectiveness of LP filter depends on selected target too. For example emission nebulae rich in Ha, OIII, SII and such will benefit from UHC or CLS type of filter. Broadband target like galaxies might not benefit from these filters as they are very aggressive in amount of light they block.

Thanks for that, very good. So LP filters block the light coming from the target, but it's all about SNR ultimately and if the LP filter blocks enough LP which improves the SNR that's what makes them worth it. Thanks for the explanation. I think I have moderate LP, certainly not what I would call rural, class 3/4 etc. it's class 6, so I guess I'll probably have to just go for one and see what it does, with a good return policy from whoever I buy it from. Wex are selling the Skytech CLS and have a good returns policy, but I am still dithering a bit.

[smr]

Here's 120 seconds on the HH, without an LP filter, and the histogram is not even half way over, so I'm not losing any dynamic range ?

 

[vlaiv]

Don't worry about dynamic range.

What you should worry about is clipping/saturation. And even that is not something you should overly worry about because it can be easily fixed with few short subs. You should be doing that anyway if you want to keep the color on brightest stars that easily saturate.

Dynamic range is important in regular photography where one shoots a single shot of something.

With AP - stacking is adding dynamic range, so even if single exposure lacks dynamic range - stacking many of them will extend dynamic range past anything single shot can do on best cameras under ideal conditions. This is important in AP since ratio of very bright signal to very faint signal surpasses anything in daytime photography (think stars - mag5 star and mag20 star - that is 15 magnitudes of difference or 100x100x100 = 1,000,000 times more light - that is dynamic range over 1:1,000,000 in single photo).

 

[david_taurus83]

FYI I have the Skytech CLS CCD in my 600D. Below is a comparison with and without the filter. 50mm 1.8 and 10s exposures. There was one for sale on here not long ago.

 

Edit: Still in the FS section.

https://stargazerslounge.com/topic/328558-skytech-cls-ccd-filter/

[smr]

1 hour ago, david_taurus83 said:

FYI I have the Skytech CLS CCD in my 600D. Below is a comparison with and without the filter. 50mm 1.8 and 10s exposures. There was one for sale on here not long ago.

 

Edit: Still in the FS section.

https://stargazerslounge.com/topic/328558-skytech-cls-ccd-filter/

Thanks, although as I have an unmodified 80D the ccd variant wouldn't be for my Camera. How does it compare to the D2 ?

 

Presumably your 600D is modded.. you can really see the Ha in the HH nebula with the filter on. Do you live in a city/village? Sodium lamps? 

[Scooot]

I’ve recently stopped using a light pollution filter (CLS) with my canon 450d. I always found it very difficult to process star colours with the filter and much easier without. Although the single subs with the filter look better. Once I’ve applied dynamic background extraction to the unfiltered stack I found my star colours were much better, even before colour calibrating. My “light bulb” moment came with my last two Hyades images that are in the constellation challenge. The first with a bright moon and filter was very difficult. Since then all my images have been without the filter and I’m much happier with the star colours. I might however use the filter again if I’m targeting nebulae, I’ll have to see.

I do see others with great results using a filter so my experiences might be unusual, or just down to inexperienced processing. 

[vlaiv]

Examining spectral response curve of both filters can say a lot about them - if you understand what they represent.

https://www.sciencecenter.net/hutech/idas/lps-wideplots.htm

 

Compare that to CLS for example:

https://www.astronomik.com/en/photographic-filters/cls-ccd-filter.html

And here is a graph that can help with LP spectrum:

And for leds :

These are not to scale (led being dimmer in above graph - but that will depend on strength of lighting and LP - this is just distribution of spectra - note that HPS looks the same in both examples).

This graph I find very interesting:

You can spot dominant lighting type depending on which part of night we are looking at - it also shows that structure of light varies with time of night - as people go to bed and turn off their lights.

 

 

[vlaiv]

4 minutes ago, Scooot said:

I’ve recently stopped using a light pollution filter (CLS) with my canon 450d. I always found it very difficult to process star colours with the filter and much easier without. Although the single subs with the filter look better. Once I’ve applied dynamic background extraction to the unfiltered stack I found my star colours were much better, even before colour calibrating. My “light bulb” moment came with my last two Hyades images that are in the constellation challenge. The first with a bright moon and filter was very difficult. Since then all my images have been without the filter and I’m much happier with the star colours. I might however use the filter again if I’m targeting nebulae, I’ll have to see.

I do see others with great results using a filter so my experiences might be unusual, or just down to inexperienced processing. 

Nothing unusual - CLS just kills off one third of spectrum and makes proper color balance impossible. If you want to use filter and keep any chance of color balancing afterwards - you need LPS filter that does not just cut a part of spectrum out - but has some dominant LP spectral lines removed (see above post and notice difference between P2 and D2 or V4 - latter two are closer to CLS/UHC type of filter, while P2 really balances things out - there is no single large gap between 520nm and 640nm as most CLS filters have).

[smr]

I suppose that until I buy an autoguiding setup I can do without an LP filter but when I get into the realms of 3-5 minute exposures I will have to have one as the data would be well over to the right on the histogram.

[wxsatuser]

In my experience a decent LP filter will mean longer subs.

Here I have quite bad LP and with an LP filter, Astronomik CLS clip filter, subs went from 2minutes tops to 5 and 10minutes.

Here is a single 600sec sub of the California nebula with the above filter and a Canon 60Da, EF 200mm L lens at f4 ISO 800.

There is no real problem getting a reasonable colour balance.

[david_taurus83]

15 hours ago, smr said:

Thanks, although as I have an unmodified 80D the ccd variant wouldn't be for my Camera. How does it compare to the D2 ?

 

Presumably your 600D is modded.. you can really see the Ha in the HH nebula with the filter on. Do you live in a city/village? Sodium lamps? 

I can't compare directly to the D2 as I've only used the IDAS with the ASI1600 mono. With the mono however I found it very easy to remove gradients and the colour balance was quite good. But using LRGB filters as well prob helped this.

 

I image from right under 3 orange streetlamps!