Jump to content

1564402927_Comet2021Banner.jpg.a8d9e102cd65f969b635e8061096d211.jpg

Lodestar Live Multi-spectrum preview


Paul81
 Share

Recommended Posts

Hi All,

Thought I would post an update with the latest progress with LL – multi-spectral EAA. I often use multiple filters to visually observe so makes sense to extend this concept to EAA. Also hardcore imagers use narrow band with great success to tease out faint detail and cut through the orange glow.

At its heart its quite simple, you can apply a colour mask to new images as they are live stacked. This is performed via the Exposure Channel Mask set of checkboxes on the Exposure tab. When a colour channel is checked it means that channel will be updated during live stacking with new data. If the channel is unchecked then it is left untouched.

post-9673-0-33398100-1438899244.png

Selecting All leaves everything as normal – a new image is stacked as we all know and love (i.e. all channels updated with new data). However, you can select one or more separate channels, and based upon this selection the new image will only be stacked into those colour channels.

For example, if using a Ha filter (and going for the Hubble palette), you would take exposures through a Ha filter and just select the Green channel on the Exposure Channel Mask checkboxes. For each new image, only the green channel is stacked, red and blue are untouched.

You then stop the camera, change to a Oii filter and then select the Blue channel (you might also need to switch to the focus mode to adjust the focus). Start the camera and now new images are only stacked into the blue channel leaving the red and green channels untouched.

And you've guessed it, switch to a Sii filter and select red. New exposures are only stacked into the red channel.

You can stack to multiple channels, a popular choice is Ha to red and green and Oii to green and blue (see the Rosette below).

You can use this facility not only with narrow band filters but RGB filters with a mono camera to yield full colour.

You could also use different exposure lengths for different filters!

On a very related (but possibly useful for OSC use), the Display processing tab now has a Display Selected checkbox. Check that and which ever colour channel you have selected on the display processing tab (in the past this allows you to alter the histogram settings of that colour channel) is now exclusively displayed (in greyscale).

post-9673-0-61762100-1438899234.png

So using the Hubble palette, whilst observing in Ha select the green channel on the display processing tab to see only the Ha data. Likewise, once all three filters have been used to observe an object you can switch between them to see the differences in the object between the spectrums.

Here are some examples. The data has been supplied kindly by Dom543 using his OSC Lodestar and NB and RGB filters. I have run LL in its test mode I use for debugging, but these represent what would have been seen at the scope (although I didn't apply any darks).

This is the Rosette nebula with Ha observations stacked to red and green, and Oii stacked to green and blue. Some manual manipulation of the histogram is probably required to suit tonal taste but all adjustments are just as they are in the current version.

post-9673-0-84149500-1438899283.png

This screenshot shows just viewing the nebula in Ha (red).

post-9673-0-90080400-1438899250_thumb.pn

This one in just Oii (blue).

post-9673-0-52140600-1438899259_thumb.pn

As green is Ha and Oii it would basically be a mono version of the false colour composite.

M20 Triffid Nebula RGB.

post-9673-0-13666900-1438899412.png

Eta Carinae Ha and Oii.

post-9673-0-21835800-1438899350.png

I need to make some algorithmic changes to the star processing in the stacking algorithms to cope with the multiple spectrum images but the basics are there.

There are some other minor updates too, in addition to support for the Ultrastar. Look out for the release in the near future – there are some bits to tidy up, optimise and further testing / debugging.

I hope this feature of LL is a worthy addition to EAA and will prove to be useful.

I guess this feature will also be coming soon to the Atik Infinity...

  • Like 7
Link to comment
Share on other sites

Hello Paul,

I am afraid to say most of this post is lost on me as I've not yet got to this stage of the hobby but I have to admit it looks very impressive even to the uneducated like me.  Well done to you as it's clear you are working your socks off and a lot of people will be extremely grateful for your efforts.

Best wishes

  • Like 1
Link to comment
Share on other sites

Hi Paul

Thanks for the preview. No filters here as yet but this is a great incentive to get hold of some! Your channel mapping solution looks to be both simple in use yet extremely flexible. I'm looking forward to seeing what concoctions people come up with over the coming months. 

cheers

Martin

Link to comment
Share on other sites

Brilliant Paul, I was wondering how you were going to make this intuitive to use. I don't yet have RGB filters but was thinking of getting some for doing some quick imaging with the Lodestar, so this new feature is very welcome.

With the advent of these new higher resolution cameras for EAA and your evolving software I wonder whether we are seeing an interesting convergence between Video and Imaging, ie: the ability to produce high quality 'finished' images within minutes whilst still at the scope. Some may not consider it 'live' viewing, but it's not traditional imaging either because you can see the results very quickly on the night without hours of capturing and then post processing at home. It's certainly a concept that appeals to me!

Rob

Edited by RobertI
  • Like 1
Link to comment
Share on other sites

  • 2 weeks later...

I have had to pleasure to beta test the new multi-spectral feature that Paul is building into LodestarLive.

Here are two images made with monochrome cameras and the current test build of LL.

First is the Dumbell Nebula M27 from yesterday night taken with a Lodestar x2 monochrome camera.

post-26379-0-39809400-1440430254.jpg

The second one was taken a couple of nights ago with the SX-825 mono. This is a 75% crop of the original capture.

post-26379-0-17474200-1440430359_thumb.j

Both images are 5x30sec mean stacks.

The first three frames were captured with an O-III filter and assigned to the green and blue channels. Then two more H-alpha frames added to the stack assigned to the red channel. The minimum would be two frames stacked, one O-III and one H-alpha. I stack more because I like the resulting nice smooth texture. Exposure times are the same as would be with the same filters without the multispectral feature.

Please note that since we are using monochrome cameras, exposure times are shorter than would be with color cameras for two reasons. First, we don't lose light in the micro-filters on the pixels of the color sensor. Second because all pixels of the mono sensor are detecting the light equally. In case of a mono sensor only the yellow and magenta pixels are sensitive to H-alpha light and only the cyan and green pixels to O-III. (Right now Seattle is engulfed in the smoke of the massive forest fires further East. At the time the image was taken yesterday nigh, only the three main stars of the Summer triangle were visible with bare eyes on the entire sky. At 11:15pm even those were gone and I had to call off astronomy for the night. Moisture condences on the smoke particles as seeds and the whole sky becomes uniform opaque white illuminated by the moon. Due to the smoke, exposure times were probably the double of what they would be under normal conditions.)

An added benefit is that narrowband imaging is not affected by light pollution or moon glare. The moon was up and past first quarter yesterday night. This new feature of LodestarLive makes it possible to produce genuine multi-color images with NB filters.

As far as I know, this multi-spectral, multi-color capability is an absolute first ever in real-live camera assisted amateur astronomy. The new features built into LodestarLive make the multi-spectral approach much more efficient and much easier to use than my first attempts to do similar but inferior things with a color camera this past spring http://stargazerslounge.com/topic/239385-multi-colour-narrowband-in-lodestarlive/.

We are grateful to Paul for his continued efforts and devotion to keep us, users of LodestarLive, ahead of everyone else on the world and at the cutting edge of new real-time viewing technology.

Clear Skies!

--Dom

  • Like 2
Link to comment
Share on other sites

Don,

You are right. The 825 shows advantages only on objects with file detail. I will put up images of the Veil, where you can see this. As far as the compact M27 goes, the Lodestar wins.

I use Baader filters and they are parfocal. Refocusing between frames would be possible in LL without losing the live stack. But would be a chore and fun killer. Parfocal filters are very strongly recommended for live filter swapping.

If I were not in this time crunch, I would make a Lodestar vs. SX-825/Ultrastar pros and cons post. It would have items in both columns for both cameras. Disregarding prices, it would probably result in a tie for real-time EAA use.

Good to see you back!

Clear Skies!

--Dom

Link to comment
Share on other sites

Western Veil with SX-825 mono and LodestarLive Multispectral

post-26379-0-69974000-1440443260_thumb.j

Please click on the image to see it in full resolution. It makes a difference in the detail visible.

3x60sec H-alpha assigned to the red channel + 3x60sec O-III frames assigned to the green and blue channels.

All live mean stacked together using LodestarLive's multispectral capability.

Meade 10" f6.3 SCT used with 3.3 focal reducer. This is poor man's hyperstar!

Some uneven illumination is noticeable near the center of the image due to the extreme focal reduction combined with the larger sensor..

This image is dim due to the forest fire smoke through which it was taken.

Atmospheric conditions are very bad here right now and my time is running out.

Here the same object in "Swedish" palette false colors.

post-26379-0-32127600-1440443427_thumb.j

Again, please click on the image to see it in full resolution. It makes a difference in the detail visible.

To get Swedish palette colors, H-alpha is mapped to the red and green channels and O-III is assigned to the blue channel.

Clear Skies!  

--Dom

  • Like 1
Link to comment
Share on other sites

Veil Close-up with SX-825 mono and LodestarLive Multispectral

post-26379-0-47991600-1440444245_thumb.j

This is a substantially reduced image. Please click on it to see it in 50% reduced size. 

The full size and full resolution capture can be seen here http://stargazerslounge.com/uploads/gallery/album_3893/gallery_26379_3893_663204.jpg.

This image is noisy and the exposure times are extreme due to the smoke through which it was taken. All stars were gone from the sky by the time I got to this.

As said before, atmospheric conditions are very bad here right now and my time is running out.

3x120sec H-alpha + 3x120sec O-III frames mean stacked using LodestarLive's new multispectral capability.

Meade 10" SCT at F4.0 and  Baader 7nm and 8.5nm H-alpha and O-III parfocal filters.

Under normal conditions 60 sec exposures should be sufficient for this object at f4.0.

Clear Skies!

--Dom

  • Like 3
Link to comment
Share on other sites

Should add a mode for the SX spectragraph - if you set to dither on the mount - you can do a 3D spectral image. Then you can remove the sky glow by subtracting the skyglow spectra from the image spectra.

Also this technique means you also can select or adjust which bandwidths you want in the image, then render accordingly.

:)
Link to comment
Share on other sites

Like Dom, I've been lucky enough to beta test an early version of Paul's latest LL, with multi-spectral capabilities. So here are 3 objects I looked at during a short session a few nights ago under poor skies in a very bright place the day before full moon (SQM 15.3-16, which equates to naked eye limit of just over mag 1.7-2 :eek: ). All shots taken with 10s subs and live dark subtraction (at 25C this was necessary).


First, NGC 6683, a rather elongated open cluster in Ophiuchus, chosen simply because it was visible in my restricted field of view at a decent altitude of 58 degrees. My goal was to see how well star colours could be picked up via RGB imaging with the mono camera. This is a RGB composite, with no refocusing between filters. The process is as Paul described: select red filter; select red channel on the Exposure Channel Mask panel of the Exposure Control tab; expose (I collected 4 x 10s subs using mean stacking); pause; repeat for green and blue.


post-11492-0-33803400-1441052949.png


During or following each exposure (or at any point) it is possible to adjust the display as normal, with separate histogram adjustments for the three channels. The magic comes when seeing the 3 channels composited into a colour image (which is done whenever the Display Selected checkbox (on the Channel Selection panel of the Display Processing tab) is unchecked.


Next, I looked at M16 as a narrowband target. In this case I followed the above procedure but selected the red channel for Ha, OIII for green and SII for blue, a mapping known as HOS (Other mappings are possible e.g. Hubble palette: SHO i.e. SII to red, etc). Here I extended the overall exposure slightly (6 x 10s for each channel). 


post-11492-0-39038400-1441053025.png


Here are the individual H, O and S contributions separately. Although the caption says 18x10, these are actually 6x10 in each case. In this way we see that M16 mainly emits in Ha, with some OIII contribution and very little SII. 


post-11492-0-79178300-1441053069.png


post-11492-0-98127100-1441053096.png


post-11492-0-36216200-1441053126.png


The contributions from each channel can vary (as Dom has shown to good effect) so I left the SII filter in for a while and continued stacking, in order to boost its meagre contribution. This is the result of adding 34 x 10s of SII to the 6 x 10 each of the other two. The 34x is just the arbitrary point at which I stopped. The possibilities for blending are endless and I got the feeling that this innovation in LL will add a new dimension to near real-time observing.


post-11492-0-96458200-1441053182.png


A final object I looked at before the threat of automatic sprinklers realised itself was M11, the Wild Duck cluster. For this the parameters are the same as the earlier open cluster (i.e. 4x10s per channel of RGB). Here I wanted to see how well the blue and yellow members were represented.


post-11492-0-18404900-1441053283.png


I half expected this new style of near-real-time observing to be more frenetic due to changing filters, pausing etc, but it wasn't the case as the controls are very intuitive. I do recommend something like a filter wheel (or a sufficiently large filter drawer) though, as changing filters attached via a nosepiece and refocusing would be a pain. I didn't refocus at all during the session.


My feeling is that the multi-spectral capability will be great for outreach. 


Thanks to Dom for alerting us to the possibilities of multi-spectral observing and to Paul for implementing it in an easy-to-use fashion, and to Don for demonstrating just how impressive narrowband Ha can be even under heavily light-polluted skies.


Martin
  • Like 3
Link to comment
Share on other sites

Very nice images Martin!

I am still hunting for an object with sufficient S-II signal to show up and make a noticeable difference in real-time.

The Southern targets in the Sagittarius-Scutum area are too low from my current location and directly above the bright city lights of downtown.

I found that it is not absolutely necessary to suspend stacking for the filter change. The frame during which the change takes place will be out of focus and hence automatically rejected by the stacking algorithm. But with some practice changing filters takes less time than an integration cycle, hence pausing and resuming is faster.

I use a TS filter drawer system with several extra drawers. It works for me as I am sitting next to the telescope and it only requires 15mm back focus. A filter wheel is probably a more elegant solution but may require more back focus. When using camera lenses, there is only about 30mm or less space available between the lens and the camera for all adapters and filter stuff.

Clear Skies!

--Dom

Link to comment
Share on other sites

Thanks Dom. S-II is subtle but with time it can make a contribution (so leaving it as the last filter and stacking away makes sense).

I see what you mean about LL rejecting frames during filter change and I'll try this. Personally, I'm not too bothered about pausing or indeed waiting for a single missed sub. I need to practice turning the wheel the right way in the dark though ...

Since the Lodestar can slide right up almost touching the filter, the main thing to be concerned with re focus is the lens side of the wheel. Actually, I've just discovered that the filter wheel is essentially symmetrical and can be used facing either way, since the tube on the lens side can be removed and the c-mount or whatever presumably attached directly to that face. You might be able to get away with 30mm. I'll give it a go with one of my camera lenses at some point.

In the next post I'll attach a few more images from last night.

Martin

Link to comment
Share on other sites

Here are more multi-spectra images from last night, taken both before and during moonrise, with a decent SQM of 19.9 falling to around 19 at the end of a 2 hour session. I used 20s subs throughout. I didn't change the colour hue or saturation settings at all, having tuned them roughly on a previous night. In spite of numerous (20+) filter changes I only refocused once. (BTW I've just started to use a Bhatinov mask and I wonder why I didn't do so earlier -- it is such a time-saver and focusing now only takes 10s, including adding and removing the mask. Can't recommend it enough.)

First, the wonderful object M7, too far south to be widely appreciated for the absolute beauty that it is. This was captured at 17 degrees above the horizon and is composed of a single 20s sub through each of the R, G and B filters. 

post-11492-0-33522900-1441187050.png

I couldn't resist a look at Albireo. This is 1s in each of the three channels.

post-11492-0-88692200-1441187204.png

Here's M20, again with a single 20s though each filter. 

post-11492-0-57736300-1441187405.png

Here's the same thing with an extra double shot of Halpha (i.e., mapping Halpha to all 3 channels):

post-11492-0-49733900-1441187493.png

This hasn't, to my eyes, really added much nebulosity, and instead has had the effect of toning down the hues (since what is being added is essentially greyscale because it is mapped to all three channels), but it does demonstrate the possibilities of the flexible channel-mapping approach in LL.

Finally M27, first in RGB (here a stack of 3 x 20s for each channel):

post-11492-0-33941700-1441187744.png

and a few minutes later the same object using HOS mapped to RGB -- i.e. Ha to R, O-III to G and S-II to B -- with the same exposures (who says narrowband exposure have to be long?)

post-11492-0-89376700-1441187892.png

Its fascinating to look at the separate contributions, which are quite different (and not just in intensity):

post-11492-0-17511100-1441188138_thumb.p

Near live multi-spectra is going to be a lot of fun!

cheers

Martin

  • Like 4
Link to comment
Share on other sites

I agree with Martin that multi-spectrum offers more than just pretty colorful images with even mere mono cameras.

One of its main benefits is that it allows deeper insight into the structure and composition of nebulas.

E.g. the Rosette has  a fair amount of O-III and H-beta content, they are just overpowered by H-alpha. On the other hand the Jellyfish, which is a supernova remnant, consists only of H-alpha.

The blue reflection nebula of the Trifid is much larger than commonly seen. It extends to the entire area that is usually seen as red. It again is only overpowered by the stronger H-alpha radiation. I think that I have posted the pure blue filtered capture of M20 earlier somewhere on this forum. The same applies to the Cocoon.

Clear Skies!

--Dom

Link to comment
Share on other sites

Here's 'Fish three ways' (M17) from last night

RGB (5 x 15s of each)

post-11492-0-00939400-1441361690.png

HOS -> RGB (4 x 15s each)

post-11492-0-36745700-1441361837.png

SHO -> RGB (2 Ha and 4x15 of each of O-III and SII to avoid overwhelming with Ha, although the green of the Ha is still too dominant)

post-11492-0-60025500-1441362044.png

Seeing was poor but I checked out some other objects, including NGC 6726 in Corona Australis, just over the border from Sagittarius, which feels like visiting the southern hemisphere, but it was crazily low and facing a light dome so I didn't catch much of the nebulosity. I'm sure this must be a superb field for those further south. I like the well-matched double (mags 6.4 and 6.7) too. For those suitably placed, with a slightly larger FOV it would be possible to also capture the glob NGC 6723.

post-11492-0-80453300-1441362303.png

Martin

  • Like 4
Link to comment
Share on other sites

Dom and Martin

Fantastic - this is exactly what I had in mind for multi spectrum [emoji3] I like seeing the different facets the objects have in the different emission bands.

I am really happy with the initial results you guys are getting (just need my kit fixed and get some filters). There is still some work to do but it's definitely coming into shape!

Paul

Sent from my iPhone using Tapatalk

Link to comment
Share on other sites

  • 4 weeks later...

Cocoon Nebula with Lodestar mono and LodestarLive Multispectral

The blue reflection components are usually the most challenging to capture.

Real time captures of the Cocoon often go for the easy H-alpha only and ignore the blue entirely.

This capture synthesizes blue and H-alpha plus just a touch of all-spectrum luminance to make the stars look white.

post-26379-0-91689900-1443786757.jpg

500mm f2.0, Meade 10" SCT, Blue and H-alpha filters,
3x30secBlue+3x30secHa+1x5secLum exposures mean stacked,
Lodestar x2m with LodestarLive v.0.11,
Live image captured from Seattle in-city (light pollution + moisture filled air).

BTW, the Cocoon is much more challenging than the Trifid.

Edited by Dom543
  • Like 1
Link to comment
Share on other sites

  • 2 weeks later...

Helix Nebula with SX-825 mono and LodestarLive Multispectral

post-26379-0-65606300-1444660292_thumb.j

2x60sec H-alpha + 2x60sec O-III frames live mean stacked in LodestarLive.
Meade 8" SCT at f4.0 and Baader 7nm and 8.5nm H-alpha and O-III filters.
Image has been captured from under the Boston light dome and at about 20 degrees above the horizon.
The file size of this image has been reduced to 50%. See the full size, full resolution capture here http://stargazerslounge.com/uploads/gallery/album_3893/gallery_26379_3893_63743.jpg.

This is a real-time capture of a live image with no post processing applied.

I am taking advantage of the nice New England autumn nights, The foliage of the trees is getting as colorful as this image.

Around midnight I had to wait about 30 minutes, while the Helix went behind some of those colorful leaves and then came out again.

Clear Skies!

--Dom

Edited by Dom543
  • Like 2
Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
 Share

×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue. By using this site, you agree to our Terms of Use.