Question: 3nm Narrowband filters vs 7nm Narrowband filters

[MonsterMagnet]

Hi All.

I would like some advise with regards to switching from my existing mono filters (Ha = 7nm, Sii = 6nm, Oiii = 6nm, all from Optolong) to some 3nm filters probably also from Optolong.

I live in central London with the following sky conditions from Clear Outside:

• 17.74 Magnitude
• Bortle 9

In these conditions I think I would expect a decent increase in data capture between 7nm and 3nm filters - is that likely?  I can image fine in mono with these conditions but aren't we all always seeking improvements?

Will I need to do longer exposure times with 3nm filters vs 7nm filters?  Currently I tend to do 120 or 180 seconds.

Thanks.

MM

[Elp]

As far as I know it'll likely be similar, the difference being how much background the narrower bandpass blocks out. When I tried a 3nm filter I saw little difference.

[MonsterMagnet]

19 minutes ago, Elp said:

As far as I know it'll likely be similar, the difference being how much background the narrower bandpass blocks out. When I tried a 3nm filter I saw little difference.

Thanks for the reply.

My understanding was that with my heavy light polluting the narrower bandwidth would be better and perhaps there would be less halos on very bright stars.  I was looking at the Optolong set.

MM

[Elp]

The benefit needs to also be weighed up against any potential added cost. I don't have an SHO Optolong set but the LP Optolong filters I've used all halo. You'd have to go up the cost tree to get ones which don't but they're literally astronomical in price (Chroma, Astrodon).

For me the filters I have (my SHO are Baader) work acceptably well for me. For more signal and less overall noise, time is your friend (also your enemy).

Edited July 8 by Elp

[MonsterMagnet]

As an extension to this topic, if I did decide to go for 3nm, would people recommend the Optolong SHO set or the Antlia SHO set?

The Antlia are about 200 GBP more expensive.

MM

[symmetal]

13 hours ago, MonsterMagnet said:

In these conditions I think I would expect a decent increase in data capture between 7nm and 3nm filters - is that likely?

You won't get an increase in NB data using a narrower passband filter but you will get a reduction in unwanted sky background 'data' and its associated noise. So your overall image S/N will increase.  

13 hours ago, MonsterMagnet said:

Will I need to do longer exposure times with 3nm filters vs 7nm filters?

No, not necessarily. Optimum exposure time is to expose long enough such that the camera read noise contribution becomes negligible, usually when the sky background noise is at least 5x greater than the camera read noise. Exposing longer than this gives no benefit and you should start another exposure.

These optimum exposure times are useful for LRGB or OSC broadband imaging.

With narrow band imaging however, you would need to expose for a very long time to get the sky background noise 5x the read noise, possibly an hour or more, depending on your scope speed and light pollution, so for narrow band it's best to expose for as long as possible before other factors start spoiling your image, like tracking errors or the likelihood of planes flying through. With my f2.2 RASA 11 I'd like to do 10 min narrowband exposures, but I occasionally get slightly oval stars at 10 mins so have opted for 5 min narrowband exposures instead.

Your 3nm filters would need longer exposures before your image sky background level equals that from your 7nm filters, but just use as long an exposure as you're comfortable with, the longer the better. 🙂

Alan

[MonsterMagnet]

6 hours ago, symmetal said:

You won't get an increase in NB data using a narrower passband filter but you will get a reduction in unwanted sky background 'data' and its associated noise. So your overall image S/N will increase.  

No, not necessarily. Optimum exposure time is to expose long enough such that the camera read noise contribution becomes negligible, usually when the sky background noise is at least 5x greater than the camera read noise. Exposing longer than this gives no benefit and you should start another exposure.

These optimum exposure times are useful for LRGB or OSC broadband imaging.

With narrow band imaging however, you would need to expose for a very long time to get the sky background noise 5x the read noise, possibly an hour or more, depending on your scope speed and light pollution, so for narrow band it's best to expose for as long as possible before other factors start spoiling your image, like tracking errors or the likelihood of planes flying through. With my f2.2 RASA 11 I'd like to do 10 min narrowband exposures, but I occasionally get slightly oval stars at 10 mins so have opted for 5 min narrowband exposures instead.

Your 3nm filters would need longer exposures before your image sky background level equals that from your 7nm filters, but just use as long an exposure as you're comfortable with, the longer the better. 🙂

Alan

 

Thanks for your answers.

I used to expose for 5 minutes when I started out with OSC but with the move to mono I've actually reduced that quite a lot to 120 or 180 seconds as there were numerous threads where people argued that overall integration time was more important than individual exposure time.

Seems I might need to readdress this assumption.

MM

[ollypenrice]

I can compare experiences with the 7nm Baader and the 3nm Astrodon.

The Astrodon is, if anything, faster and gives signifcantly more striking contrasts and much smaller stars. It is also significantly more 'moon proof' so I would expect it to be better in other LP as well.  Was it worth the considerable expense? Yes.

Olly

[Pompey Monkey]

See below

Edited July 9 by Pompey Monkey
Forgot to include Olly's quote

[Pompey Monkey]

5 hours ago, ollypenrice said:

I can compare experiences with the 7nm Baader and the 3nm Astrodon.

The Astrodon is, if anything, faster and gives signifcantly more striking contrasts and much smaller stars. It is also significantly more 'moon proof' so I would expect it to be better in other LP as well.  Was it worth the considerable expense? Yes.

Olly

I will second this, although I have a Chroma 3 nm set rather than the Astrodons. 

Before installation at our remote location, I tried them at my Bortle 6-7* home obsy. The way that they lifted the nebulosity from the background was waaay better than the 7 nm set I was using before. They did cost a small fortune though 

* lightpollution.info says that I'm in a Bortle 5 area, but it's plain out lying: I've not seen the Milky Way for a few years now. Sardonic lol.

[MonsterMagnet]

On 09/07/2024 at 09:23, ollypenrice said:

don is, if anything, faster and gives s

Thanks all.

Can't stretch to the Chroma or Astrodon at the moment but have managed to source 3nm Antlia filters second hand which seem to have better reviews and specifications than the Optolongs.

MM

[Vroobel]

I have an Optolong 6.5-7nm SHO and their LRGB. I can see that particularly OIII subs are very bright, so I consider replacing the wider OIII with 3nm one. Does it make sense? I know that the oxygen contains two wavelengths. 

 

[Clarkey]

FWIW I replaced my 7nm filters with 3nm Antlia filters. However, until it gets dark again, I cannot tell you whether it was worth it! I believe the Optolongs are a bit prone to halo's on the OIII filter, but they are good value.

[900SL]

On 09/07/2024 at 09:45, MonsterMagnet said:

 

Thanks for your answers.

I used to expose for 5 minutes when I started out with OSC but with the move to mono I've actually reduced that quite a lot to 120 or 180 seconds as there were numerous threads where people argued that overall integration time was more important than individual exposure time.

Seems I might need to readdress this assumption.

MM

Individual subs need to be long enough to swamp read noise. This can be difficult with narrowband filters. I go for 5 -6 minutes, bortle 5 with 4.5 nm Antlia Edge filters at unity gain on a 533MM

[MonsterMagnet]

12 hours ago, 900SL said:

Individual subs need to be long enough to swamp read noise. This can be difficult with narrowband filters. I go for 5 -6 minutes, bortle 5 with 4.5 nm Antlia Edge filters at unity gain on a 533MM

Isn't there barely any read noise on an ASI2600MM?

Understand that less signal is getting through the filter but I still thought overall integration time was the important aspect here?  I used to shoot at 5 minutes but reduced that recently after reading that.

MM

[Elp]

Overall is still important, but by shooting longer per sub you reduce the instances (based on number of subs) of read noise in your stack.

[MonsterMagnet]

55 minutes ago, Elp said:

Overall is still important, but by shooting longer per sub you reduce the instances (based on number of subs) of read noise in your stack.

Thanks.

Seems to be differing information with regards to this, lots of sources saying with new cameras only total integration time is what matters.  I'll go back to 5 minutes then but that does increase the chance of throwing away data due to guiding errors etc.

MM

[Elp]

12 minutes ago, MonsterMagnet said:

go back to 5 minutes then but that does increase the chance of throwing away data due to guiding errors etc.

You have to adjust accordingly. I started with an azgti and even though could do 3 minutes autoguided at around 400mm imaging 120mm guiding, any slight breezes and stars would sometimes elongate, after dithering the next sub would also usually be bad, so I dialled it back to around 60s. I still usually do 60s due to the amount of satellites because you can sometimes see their trails in the stack if they're particularly bright (or planes) and the kappa sigma range doesn't get rid of the trail, 60s even for narrowband. But on beefier mounts can certainly do longer, and for me I usually need to as I'm using a small pixel camera, for faint signal depending on target you really need to to get the pixels to respond. I think narrowband, a region of between 1-5 minutes is fine, fainter signal do the longer end, and I'm imaging from bortle 7, but I am using fast optics, slower optics you'd definitely see more of a benefit doing the longer end.

[MonsterMagnet]

Thanks.

I used to just about get 5 minutes with my az-gti but it was very variable.  I've got an AM3 now and the guiding performance is excellent so I don't think 5 minutes would be an issue, just thought the same overall integration time with 120 second or 180 second subs would be just as good if not better due to more subs for stacking to remove noise and trails.

I am bortle 9 though, hence the move to 3nm filters.

MM

[licho52]

On 11/07/2024 at 12:01, MonsterMagnet said:

Isn't there barely any read noise on an ASI2600MM?

Understand that less signal is getting through the filter but I still thought overall integration time was the important aspect here?  I used to shoot at 5 minutes but reduced that recently after reading that.

MM

 

With broadband it's mostly true, you can use short exposures.  With narroband it's still a factor however and it depends on focal ratio - with 5 or faster it's ok to do 5 min subs, but at f/6 it's 10 and at f/8 it's 15 minutes with 3nm filters.  Otherwise the dark areas of the sensor basically get nothing and all you get is sensor pattern there (no, darks don't remove it).

[Clarkey]

10 minutes ago, licho52 said:

With narroband it's still a factor however and it depends on focal ratio - with 5 or faster it's ok to do 5 min subs, but at f/6 it's 10 and at f/8 it's 15 minutes with 3nm filters

I don't think it is really as simple as this. You need to consider other things such as pixel scale, QE, light pollution etc, etc. I have never used more than 5 minute subs, even at f8. Using longer subs just means more chances for a poor sub and more data loss. If you live in B1 or B2 you might need to use longer subs just to swamp the read noise. However, the OP lives in B9 so short subs is probably more sensible.

[sn2006gy]

3nm filters will need longer exposures than 7nm filters "in general" as they're letting in 2.3x less light.

I shoot in very widefield, and I find the 3nm to be extremely well suited since my aperture is small and my speed isn't fast.  This lends itself to very detailed images that I couldn't get with wider filters and I'm in a B2 sky so I have to take much longer exposures than 5nm or 7nm filters.

I use the Antlia Pros and I absolutely love them. This is one of my images using the 2" set on a 6200mm pro.

I lost some resolution on the attached image trying to do a crazy LRGB SHO combination, but the pure SHO was beautiful and sharp (trying to bring in the dark clouds that are abundant in this area)

[MonsterMagnet]

23 hours ago, licho52 said:

 

With broadband it's mostly true, you can use short exposures.  With narroband it's still a factor however and it depends on focal ratio - with 5 or faster it's ok to do 5 min subs, but at f/6 it's 10 and at f/8 it's 15 minutes with 3nm filters.  Otherwise the dark areas of the sensor basically get nothing and all you get is sensor pattern there (no, darks don't remove it).

 

I did some experimentation with my purchased Antlia 3nm filters and it has made a huge difference.  It's actually having issues plate solving with the new filters at the exposures I was using to move the mount so they will need to be increased and I barely see stars until I use a 15 second exposure.

I tried upping my exposure time as suggested and tested both 5 minutes and 10 minutes and both worked very well with excellent Ha exposures.  The 10 minute one is better but hard to tell how much more information is then until I have a proper chance to stack some frames.  My only concern with 10 minute exposures is the increased chance of bad frames and also less frames at the end to stack meaning some noise removal from dithering and trails might be harder to remove in the stacking process.

I tested the Oiii on Sadr and I did see a halo but to be honest it's a lot better than what I was getting before.

I've attached a single 10 minute frame (STF applied in PI only) of lobster, bubble and NGC7538.

MM

 

[Stuart1971]

On 08/07/2024 at 11:57, MonsterMagnet said:

Thanks for the reply.

My understanding was that with my heavy light polluting the narrower bandwidth would be better and perhaps there would be less halos on very bright stars.  I was looking at the Optolong set.

MM

I too used the same 6.5nm Optolong  filters, and also swapped to the 3nm version of them from a Optolong, about 9 months ago, and can say they are superb with no halos, but then I didn’t really get any on the 6.5nm version either, but you will get more signal in a shorter exposure, but that because there is less light pollution in the image, so I use the same length exposure images and get a better image in the end and easier to process…

[Clarkey]

On 17/07/2024 at 21:32, sn2006gy said:

3nm filters will need longer exposures than 7nm filters "in general" as they're letting in 2.3x less light.

Yes, but 2.3 times less of the frequencies you don't want. In terms of signal from Ha it will be about the same.

On 17/07/2024 at 21:32, sn2006gy said:

I'm in a B2 sky so I have to take much longer exposures than 5nm or 7nm filters.

Just consider yourself lucky! B2 is most of the UK is just a dream.... But I accept in your case longer subs will be needed to swamp read noise.