Multi-spectral help needed

[RobertI]

Hi all,

I’m looking to use StarlightLive and my Lodestar mono do some multi-spectral EAA in the near future but before I splash out on RGB filters and wheel I thought I would trying it out using my Ha filter. What I was envisaging was mono captures of suitable objects overlaid with Ha, for example brighter galaxies such as M82. I think Martin has demonstrated that this is possible. My questions are:

• I believe I do this by selecting all three channels (RGB) in Image Acquisition mode and capturing the unfiltered captures as normal (stacked), then pausing the capturing, fitting the filter, selecting only the Red Channel and starting the capturing again. Is this how it’s done?
• I guess I would be using the same exposure for both the unfiltered and Ha filtered (say 30 seconds) at F4.5 – would this bring out enough Ha?
• Can I can manipulate the Ha curve separately?
• I’ll be using the RC6 - am I likely to need to refocus between the mono and the Ha?

Regarding the RGB filters which I plan to eventually get:

• There seem to be a range of filters available ranging from absorptive at £75, the dichroic at £129 to premium filters at £££hundreds. Will the cheaper absorptive range be adequate?
• All filter sets include an IR/UV cut filter – would I need to use this?
• What sort of loss of sensitivity can I expect using RGB filters? Eg: will 30 seconds now require 45 seconds?
• Will refocus between filters be required?

Any help much appreciated.

Rob

[Astrojedi]

Hi all,

I’m looking to use StarlightLive and my Lodestar mono do some multi-spectral EAA in the near future but before I splash out on RGB filters and wheel I thought I would trying it out using my Ha filter. What I was envisaging was mono captures of suitable objects overlaid with Ha, for example brighter galaxies such as M82. I think Martin has demonstrated that this is possible. My questions are:

• I believe I do this by selecting all three channels (RGB) in Image Acquisition mode and capturing the unfiltered captures as normal (stacked), then pausing the capturing, fitting the filter, selecting only the Red Channel and starting the capturing again. Is this how it’s done?
• I guess I would be using the same exposure for both the unfiltered and Ha filtered (say 30 seconds) at F4.5 – would this bring out enough Ha?
• Can I can manipulate the Ha curve separately?
• I’ll be using the RC6 - am I likely to need to refocus between the mono and the Ha?

Regarding the RGB filters which I plan to eventually get:

• There seem to be a range of filters available ranging from absorptive at £75, the dichroic at £129 to premium filters at £££hundreds. Will the cheaper absorptive range be adequate?
• All filter sets include an IR/UV cut filter – would I need to use this?
• What sort of loss of sensitivity can I expect using RGB filters? Eg: will 30 seconds now require 45 seconds?
• Will refocus between filters be required?

Any help much appreciated.

Rob

Rob,

Here are some thoughts...

One of the most important things to always keep in mind with mono imaging with filters is balancing the contribution of each filter. This is where most folks starting out with mono imaging trip up. I will give specific tips below.

1. Using just a HA filter: With Paul's SW you have a lot of flexibility here in exploring different approaches. You can just do HA and assign to all channels. Or you can assign HA to the Red channel and then do a non HA exposure assigned to all channels. If doing the latter given that the unfiltered sub will have a much higher signal you will need to find out the right ratio in the number of filtered vs. unfiltered subs you need to stack. I would start with 5 HA subs for each unfiltered sub.

2. Using RGB filters: MANY words of caution here:

DON't (sorry for the caps but this is very important) buy cheap LRGB filters. Cannot emphasize this enough. Here is why:

a. Make sure you get filters that are parfocal otherwise you are in for a world of pain (refocusing after each filter change starts to get painful fast)

b. Good LRGB filters have gaps between the response of each filter to cut out LP. Otherwise you will need to use a LP filter along with the RGB filters which causes 2 problems: first you lose even more light due to transmission losses. Second the optical train becomes more complicated to get right.

c. Getting the color balance right is very difficult when doing LRGB if you don't have good quality filters. Baader LRGB for example are specifically designed so that you can do subs in 1:1:1 ratio and get good color balance right off the bat. Otherwise be ready for a lot of histogram tweaking.

d. In general good quality (more expensive) filters have higher transmission and more accurate easier color.

3. Using narrowband filters along with LRGB: This brings its own challenges:

a. Again ensure that the narrowband filters are parfocal with the LRGB filters (most are not unless the manufacturer specifically says so). Will make your life a lot easier down the road.

b. Balancing the channels: E.g.:Many people prefer using HA as luminance. I have found that results is a very washed out color as the HA overwhelms the other channels. A better approach is to assign HA to the Red channel and figure out a good ratio of red vs. HA frames for the Red channel (will vary based on your optics, camera and the filters you use).

Again none of this is meant to discourage you but just to make you aware of the complexities of imaging with filters. I actually enjoy this complexity as it leads to experimentation and scientific learning.

Hiten

[Martin Meredith]

Hi Rob

I think Hiten has nailed a lot of the points I would make. Especially important is parfocality. If you are planning to mix mono (i.e. unfiltered) with filtered then you need to get the clear 'C' filter (or use the 'L') to avoid refocusing. Nowadays, I mainly keep the filter wheel on the scope and when I'm shooting mono I'm actually going thru the C filter. I guess I lose a few photons but the ability to turn the filter wheel and take darks outweighs that.

I found a good deal for the Baaders on Amazon which included the 3 normal narrowband as well as RGBLC - all parfocal. Since you've already got Ha then this might not be of interest, but the Ha you have may not be parfocal with what you end up buying. 

In terms of the LL experience, what you've described is one way to do it, but there is huge scope for experimentation. The image below which I posted in the middle of some other thread was actually captured the other way round: first, narrowband, then some mono (in heavy moonlight). As it was at the end of a session I was just playing around so I didn't document it as well as I ought to have done, but I mapped quite a lot of Ha to R, then mapped OIII to both G and B, then added luminance (L filter) just to G and B. I'm not sure why, but I thought I'd leave the Ha intact on the R channel. I animated it simply to show the effect of clicking the DisplaySelected checkbox.

Your other questions:

The IR/UV cut filter is what I think is called L in the Baader set. Its up to you whether you use it or not. I guess it could have a beneficial effect on an achromat, less so on a reflector.

Loss of sensitivity is not something I've measured formally yet, but I've noticed a drop, but not much. Having said that, I mainly bought the colour filters for open clusters rather than dim objects. I did find a galaxy example that is roughly comparable for mono vs RGB. Left is 28 x 15s while right is 24 x 15s (8 x 15 in each of RGB). It isn't a great comparison as the colour image is really badly focused (probably due to a stiff manual filter wheel at the time), so don't read too much into it...

Looking forward to seeing your results with Ha/mono.

Martin

[Paul81]

Hi Rob,

These are some good questions! I am not sure I have answers for you on all points.

I think you might need to re-focus between filters, so I'd stop capture and switch to focus mode, switch filter and check that focus is still optimum and switch back to acquisition mode. I think filters try to be parfocal but I am sure I have seen the long exposure folks mentioning they always check. You can switch mode without stopping - it automatically stops when you switch mode. Good news, is that V3 will remember different exposure times for focus mode vs dark and acquisition so this will be a little nicer.

You may also need longer exposures for narrowband, but given this needs a change of dark i'd just observe for longer in narrowband filters to get the same level of signal up. The other idea would be to pick an 'in the middle' exposure length. So if you sued 10s for LRGB and 20s for Ha/OIII then perhaps use 15s so that one exposure length is used to avoid switching darks etc.

As for how to combine the data, I think this is something open to the community to discover as more people use the feature. With more experimentation I am sure the ideal methods will come apparent. I suspect RGB will require some manual histogram tweaking but hopefully the controls will help to make it simple. 

I am hoping to get a filter wheel and filters at Astrofest so I can start to play with multispectral too! So far I have developed and tested using donated data only so will be nice to see how it works in anger!

[RobertI]

Thanks for the fantastically detailed responses Hiten and Martin. Some very valuable insights.

Hiten - i shall heed your advice on not going cheap for filters for the reasons you describe, sounds like it will save a huge amount of time in refocssing and adjusting colour balance. So regarding method 1, where the Ha is mapped to red and the unfiltered is mapped to RG&B, any attempt to increase the level of the Ha in the final image will make the image look redder? So the only way to increase the brightness of the Ha is more subs relative to the unfiltered - is that correct?

Martin - I had not thought about the advantages of a filter wheel (espcially an motorised one) for taking darks, another potential advantage. I had also not appreciated the significance of a C or L filter to retain the correct focus. I had seen your mono/Ha image before, partly what inspired me to have a go, I think mono+Ha could be fun to start with (also means I don't have to buy anything else!), but without the L or C filter i would need to refocus so not quite so simple as ai thought. Your comparison between mono and colour is interesting, and despite the focus issue I think its clear there is some light loss, but I guess I would be looking at the brighter galaxies so hopefully not a massive issue.

At least I am now aware of the pitfalls, so I shall keep it simple to start with and see how it goes.

Thanks again.

Rob

[RobertI]

Hi Rob,

These are some good questions! I am not sure I have answers for you on all points.

I think you might need to re-focus between filters, so I'd stop capture and switch to focus mode, switch filter and check that focus is still optimum and switch back to acquisition mode. I think filters try to be parfocal but I am sure I have seen the long exposure folks mentioning they always check. You can switch mode without stopping - it automatically stops when you switch mode. Good news, is that V3 will remember different exposure times for focus mode vs dark and acquisition so this will be a little nicer.

You may also need longer exposures for narrowband, but given this needs a change of dark i'd just observe for longer in narrowband filters to get the same level of signal up. The other idea would be to pick an 'in the middle' exposure length. So if you sued 10s for LRGB and 20s for Ha/OIII then perhaps use 15s so that one exposure length is used to avoid switching darks etc.

As for how to combine the data, I think this is something open to the community to discover as more people use the feature. With more experimentation I am sure the ideal methods will come apparent. I suspect RGB will require some manual histogram tweaking but hopefully the controls will help to make it simple. 

I am hoping to get a filter wheel and filters at Astrofest so I can start to play with multispectral too! So far I have developed and tested using donated data only so will be nice to see how it works in anger!

Thanks for the reply Paul. Its great to hear that stacking auto pauses when swtiching to focus mode - will make life much easier should the need arise, once again your software make things very simple! I am assuming that it is not possible to mix stacking modes, ie: to sum stack the Ha and to Median stack the Unfiltered into kne image? i seem to recall that changing the stacking method resets the stack? Looking forward to trying this out.

Rob

[Astrojedi]

Rob,

You will have to experiment and find out how many HA subs you need vs. unfiltered. The unfiltered subs will obviously contribute much more signal so to bring out the HA detail (assuming equal sub exposure time) you will need more HA subs. Exactly how many will depend on the object and your preferences (how much you want to emphasize the HA signal). It is definitely going to look red & white and I expect get redder as you stack more HA.

Hiten

[aparker]

Rob,

 

I have started playing with narrowband just a bit, and my main learning is that I need to do significantly longer exposures to get enough bright stars to enable successful stacking by SLL.  Looking at M1 the other night, I was easily able to do 20 sec exposures through the LP filter and have them stack.  In fact 10 sec surely would have been fine as well.  But I had to go up to 60 sec in Ha, OIII, SII to get even the majority of frames to stack.  Looking at the stacking stats it was clear with shorter subs that very few stars were being found.  A minor aside is that it is also apparently the case that even filters from the same manufacturer are not always perfectly parfocal. My Astronomik OIII and SII, for example.  

 

Good luck, and please let us know what workflow you end up using for this!

[Paul81]

Hi Alex,

Did you happen to be exporting your narrowband exposures to FITS? One of the things I worked on for V2.0 was modifications to the star detection algorithm to take into account the differences between non-narrowband and also between the narrowband filters themselves. This is actually quite tricky, and is part of the overall stacking algorithm that evolves the more it gets used under different circumstances and set-ups.

 

If you are able to send me any narrowband FITS data it would hugely help in this area. Although by the nature of the reduced amount of photons hitting the sensor there will be a need to increase exposure times, it was a design goal to minimise this as much as possible.

 

Oh V2.1 should be around hopefully by the weekend - its all done I just need to build it for release and get it sent to SX. Its a busy week though...

 

Paul

[Astrojedi]

I have never had any problems stacking HA subs in v1.1 using 825 and sub exposures as low as 20s (but best results are achieved at 60s). I typically run the setup at f3.4 - F4 when doing narrow band.

Alex, what f ratio and HA filter are you using?