M51-HaLRGB

[Rodd]

This is data from the Liverpool Telescope that Goran was kind enough to pass on to me.  For that I am grateful as I am not sure how to obtain it, even though its publicly available.   I do not know how many subs went into each stack--but this consists of a red, green and blue stack, and Ha as well.  I integrated a synthetic luminance and processed the image as an HaLRGB.  Not quite like Hubble data--but a 42 inch (I think) telescope does provide nice data.  This is my first image using publicly available data.  I see why Deep Sky West or other places like that are pretty nice.  Wish they were a bit more economical.

 

 

[Rodd]

Ireprocessed the data--being more careful with the Ha cleaning--still not cleaned very well--a lot of stray Ha speckles I had to remove.  But I think this is a better image--no synthetic luminanmce--I found it dulled the image down noticeably.  Trying to sharpen the image but not having much success.

 

[wimvb]

Very nice. Btw, the Liverpool telescope is a 2 m (80") beastie. Pixels are 15 um, binned to yield 30 um, which at 20 m focal length, still gives an image scale of 0.3 "/pixel.

[DaveS]

And is available for schools to access through the National Schools' Observatory.

That is a mesmerising image, looking at it you could fall right into the core.

[Rodd]

34 minutes ago, wimvb said:

Very nice. Btw, the Liverpool telescope is a 2 m (80") beastie. Pixels are 15 um, binned to yield 30 um, which at 20 m focal length, still gives an image scale of 0.3 "/pixel.

Thanks wim. That probably explains why I could not sharpen that effectively. I will revisit

Rodd

[Rodd]

21 minutes ago, DaveS said:

And is available for schools to access through the National Schools' Observatory.

That is a mesmerising image, looking at it you could fall right into the core.

Thanks Dave. Need to bring down the red a bit

Rodd

[Rodd]

I dropped the red a bit in the non Ha areas--it was getting a bid crimson

 

 

[wimvb]

I found that the LT data responds fairly well to deconvolution. Then, in the non linear phase, you can use HDR compression and enhance the details with MMT.

Most of the data from the LT consists of only a few subs per channel, mostly in the range 30 - 120 seconds. But if you're lucky, it's possible to find several tens of subs per channel.

[Rodd]

2 hours ago, wimvb said:

I found that the LT data responds fairly well to deconvolution. Then, in the non linear phase, you can use HDR compression and enhance the details with MMT.

Most of the data from the LT consists of only a few subs per channel, mostly in the range 30 - 120 seconds. But if you're lucky, it's possible to find several tens of subs per channel.

Yes I used Devonvolution in the subsequent versions and I was surprised how much of an effect it had. Usually I use like .03 dark ring control with a star mask for local seeing if.  I usually find 40-60 iterations works best, and sometimes it is not worth it.  This image only needed 10 iterations and no deringing and the effect was huge

[wornish]

Can anyone share how to access the public data on the Liverpool Telescope?  I have tried the website form but never been able to get any info back.

[wimvb]

16 minutes ago, wornish said:

Can anyone share how to access the public data on the Liverpool Telescope?  I have tried the website form but never been able to get any info back.

In the data archives form, enter a target name (eg M 1)  in the "Resolve name" field. Then press "resolve name now" to get RA and DEC coordinates.

Observational criteria, integration: set a minimum time in order to filter out a lot of rubbish, ">31" will filter out anything up to 30 seconds. Instrument, choose either IO:O or ratcam (older camera). If you want images of planets, you will need the shorter times.

Imaging configurations, choose one filter at a time. Eg, Sloan r' is the name of the red filter. Bessel B is blue, Bessell V is green.

Administrative criteria, choose "select only public data"

Submit your search.

Press  a fits link to show an image on screen. But scroll down the page, and select a "reduced tarball" link to get to a download page. The archives will compress the image files in a tarball archive format to simplify and speed up image download. You will need a program auch as 7zip to unpack the images. The fits files themselves are already compressed and have a .gz file extension.

The above settings returned 43 images, captured between 2013 and 2016, for the proposal id "NSO_Priority_1", which is the LT outreach programme.

Hth.

Good luck

[Rodd]

2 hours ago, wimvb said:

In the data archives form, enter a target name (eg M 1)  in the "Resolve name" field. Then press "resolve name now" to get RA and DEC coordinates.

Observational criteria, integration: set a minimum time in order to filter out a lot of rubbish, ">31" will filter out anything up to 30 seconds. Instrument, choose either IO:O or ratcam (older camera). If you want images of planets, you will need the shorter times.

Imaging configurations, choose one filter at a time. Eg, Sloan r' is the name of the red filter. Bessel B is blue, Bessell V is green.

Administrative criteria, choose "select only public data"

Submit your search.

Press  a fits link to show an image on screen. But scroll down the page, and select a "reduced tarball" link to get to a download page. The archives will compress the image files in a tarball archive format to simplify and speed up image download. You will need a program auch as 7zip to unpack the images. The fits files themselves are already compressed and have a .gz file extension.

The above settings returned 43 images, captured between 2013 and 2016, for the proposal id "NSO_Priority_1", which is the LT outreach programme.

Hth.

Good luck

Another learning curve!😳

[wimvb]

2 hours ago, Rodd said:

Another learning curve!😳

There ain't no such thing as a free lunch. 😉 

[Rodd]

8 minutes ago, wimvb said:

There ain't no such thing as a free lunch. 😉 

I know, but a 7 course meal? A discount would be nice!

Rodd

[gorann]

Sorry for the late response but I have been busy building an obsy, and my body is aching.

Looking very good Rodd!

I was more careful in adding Ha in my version of it since Ha is the weak point with regard to data. I just checked and this is the data I gave you in numbers of subs (they range from 30 to 120 s, with 90 s as an average):

Bessel B (blue): 58

Bessel V (greenn): 42

sdss-r (red): 74

Ha: 3

So totally about 4 hours, but that is a lot with a 2 metre reflector on a montain top with Bortle 1 sky.

Stll, I really like your Ha effect on the image - makes the M51 glow! Maybe I was too careful.

As Wim said, there is no free lunch so you have to fiddle a bit to find out how to download the data.

By the way, here is my version of the data from a year ago with a bit less Ha:

[wornish]

Wow,  just wow!

The level of detail and the whole feel of the image is superb.

[wimvb]

8 minutes ago, wornish said:

Wow,  just wow!

The level of detail and the whole feel of the image is superb.

As Göran said, 2 meter aperture on a high mountain top above the clouds will do it. With an imaging scale of 0.3 "/pixel, the data generally responds well to deconvolution. You can sharpen the images a lot. But this target is somewhat of an exception as far as integration time is concerned. Many of the images that Göran and I processed only had a few subs per channel.

For details, you can have a look at our astrobin pages.

https://www.astrobin.com/users/wimvb/

https://www.astrobin.com/users/gorann/

[Rodd]

3 hours ago, gorann said:

Sorry for the late response but I have been busy building an obsy, and my body is aching.

Looking very good Rodd!

I was more careful in adding Ha in my version of it since Ha is the weak point with regard to data. I just checked and this is the data I gave you in numbers of subs (they range from 30 to 120 s, with 90 s as an average):

Bessel B (blue): 58

Bessel V (greenn): 42

sdss-r (red): 74

Ha: 3

So totally about 4 hours, but that is a lot with a 2 metre reflector on a montain top with Bortle 1 sky.

Stll, I really like your Ha effect on the image - makes the M51 glow! Maybe I was too careful.

As Wim said, there is no free lunch so you have to fiddle a bit to find out how to download the data.

By the way, here is my version of the data from a year ago with a bit less Ha:

I like my palette but your image is sharper I think. I am on an iPhone so hard to say. I didn’t realize there was so much data in the image. Loping forward to more data!

[Rodd]

1 hour ago, wornish said:

Wow,  just wow!

The level of detail and the whole feel of the image is superb.

Not sure if this was meant fir me or Goran.  Thanks either way!

Rodd

[wornish]

8 hours ago, Rodd said:

Not sure if this was meant fir me or Goran.  Thanks either way!

Rodd

Both of you.  Your skill in post processing is amazing. 

[gorann]

 

10 hours ago, Rodd said:

I like my palette but your image is sharper I think. I am on an iPhone so hard to say. I didn’t realize there was so much data in the image. Loping forward to more data!

Yes, it seems like my version show more details and gives a sharper feeling but I can also see that you managed to save details in the brightest parts of the cores a bit better than me.

[Rodd]

2 hours ago, wornish said:

Both of you.  Your skill in post processing is amazing. 

Thanks!

Rodd

[Rodd]

16 minutes ago, gorann said:

 

Yes, it seems like my version show more details and gives a sharper feeling but I can also see that you managed to save details in the brightest parts of the cores a bit better than me.

All you have to do is drop the brightness in the core and you should see details. How did you achieve the sharpness?  I deconvolved in linear state and sharpened. Maybe I did not use the right settings (I don’t realize that the pixel scald was .3).

Rodd

[gorann]

As always I did PI deconvolution at the very end, so after stretching the image. Before that I had used HiPass filtering in PS which is also very good for increasing local contrast. I did this image about a year ago so I cannot remember the settings (and did not note them down). It could be that the big dose of Ha that you used made the image more noisy and difficult to sharpen. I should have told you that there were only 3 subs of Ha. Maybe you could make a Ha free version, sharpen it as much as possible, and then you use as lum on your HaRGB image?

I will have a go att saving my cores - did not know there were that much details in there until your image showed it.

[gorann]

Looking at my image again I could see some deconvolution artefacts so here is a new version where I backed off a bit on the deonvolution (mixed it 50:50 with the version before decon) and also brought back some details in the cores.

Sorry for hijacking your thread with my version of the data Rodd!

PS. your image should be flipped. It is now a mirror image. Liverpool telescope sometimes does that - not sure why.