Elephants Trunk In Hubble Pallet

[Rodd]

Here's the 9.3 hours of exposure I mentioned last month processed in PixInsight.

TVnp101is at FL 540mm, SBIG STT 8300 mono camera with self guiding filter wheel and Baader NB filters.  Astro Physics Mach 1 GTO. Focused using Bhatinov mask.  Created synthetic luminosity channel using the SHO-AIP script (60%Ha, 20% SII).  Channel weights were Red: 80% SII and 20% Ha, Green %60 Ha, Blue 100% OIII.  The only processing done after integration was automatic background extraction (subtraction and division), and Histogram/color saturation tweeks.

Ha: 10 x 20min

OIII: 9 x 20 min

SII: 9 x 20 min

Only had 30 flats (10 per filter) better than none I suppose

300 bias frames

33 dark frames

I'm really looking for a way to increase clarity and eliminate noise.  I tried TGV denoise and deconvolution, but I gad to reverse them--they made things worse.  How do you folks get the amazing clarity you get?  Any criticism will be like rain in a desert.  Which version do you like the best?

[pyrasanth]

It is very easy to be critical of others people's work but much harder than to be objective & kind at the same time.

The Hubble palette which you aspire to produce in your image is not reflected in the image you have presented. The colours are wrong at nearly every level. Have a look at this link which will give you some idea if you have Photoshop at reworking the image. http://bf-astro.com/hubblep.htm

You need many more subs. I would suggest at least 30 in each of your narrow band filters. Try for 8-10 minutes per sub initially & get the guiding spot on. Focus is also critical & I suggest focusing between filters. Noise increases rapidly at the same rate of initial exposure then falls off rapidly as signal increases and over takes the noise which stays fairly constant. So if you want to increase S/N then take more subs. Don't take OIII with the moon out & remember that SII often needs nearly 2-3 times as much as HA or OIII as it is normally very weak in most objects.

You need a copy of the book deep sky imaging primer (Bracken). That will be a great help.

Remember narrow band filters are not a substitute for dark skies or to be used as light pollution filters. The blocking of undesired light pollution is just a very fortunate side effect of these filters.

[toxic]

nice image but just a tad over saturated and a use of HLVG would do wonders to this thumbs up to you and keep it up Rodd

[Rodd]

It is very easy to be critical of others people's work but much harder than to be objective & kind at the same time.

The Hubble palette which you aspire to produce in your image is not reflected in the image you have presented. The colours are wrong at nearly every level. Have a look at this link which will give you some idea if you have Photoshop at reworking the image. http://bf-astro.com/hubblep.htm

You need many more subs. I would suggest at least 30 in each of your narrow band filters. Try for 8-10 minutes per sub initially & get the guiding spot on. Focus is also critical & I suggest focusing between filters. Noise increases rapidly at the same rate of initial exposure then falls off rapidly as signal increases and over takes the noise which stays fairly constant. So if you want to increase S/N then take more subs. Don't take OIII with the moon out & remember that SII often needs nearly 2-3 times as much as HA or OIII as it is normally very weak in most objects.

You need a copy of the book deep sky imaging primer (Bracken). That will be a great help.

Remember narrow band filters are not a substitute for dark skies or to be used as light pollution filters. The blocking of undesired light pollution is just a very fortunate side effect of these filters.

Thanks-but alas--I have PixInsight--not photoshop.  There is a SH-Aip script that obviously is beyond me.  SN increases with exposure length and # of subs.  9-10 20min subs should produce a pretty clean image.  Its me--not the number of subs or integration time I fear.

[toxic]

hope you don't mind Rodd had a little play with your image 10 mins just desaturated it and a dose of HLVG and a bit of denoise then added a high pass filter to it in photoshop.

[Rodd]

hope you don't mind Rodd had a little play with your image 10 mins just desaturated it and a dose of HLVG and a bit of denoise then added a high pass filter to it in photoshop.

post---new.jpg

Have at it--but I wouldn't waste my time if I were you--I have been over this set of subs many, many times, using everything I can think of and nothing i do really makes it respectable.  I started out with an unsaturated pic like yours, but it was wanting, so i saturated it.  That was obviously not the answer.  I find denoising creates artifacts.  I am at a complete loss.  My equipment is good and the raw subs look good, so its my processing.    I appreciate the critique.  

[johnrt]

Looking at the data you have stretched much too hard and have too much saturation, a more subtle hand is needed. There is also too much green, SCNR green in Pixinsight will give you a much better colour balance.

Don't be too hard n yourself, it is a good image, it just needs some refinement.

[pyrasanth]

This is an earlier narrow band image of this object that I took last year. Notice that the colours are much less saturated & the noise is reasonably controlled.

I hope you don't mind me posting the image. It is not my intention to derail your thread. Could you put your subs into a dropbox folder & the forum might be able to take a look for you as to what we could do.

[Rodd]

This is an earlier narrow band image of this object that I took last year. Notice that the colours are much less saturated & the noise is reasonably controlled.

I hope you don't mind me posting the image. It is not my intention to derail your thread. Could you put your subs into a dropbox folder & the forum might be able to take a look for you as to what we could do.

Eletrunk_NB.png

Great image--don't mind at all.  I recently found the PixInsight Color Mask Script and will give that a try.  The image you posted is similar to the best I've seen of this object.  I believe I am missing a key ingredient that must be present before a good pic can be made.  I will put the 3 stacks (Ha, OIII, SII)_ on drop box and put the link on this forum page.  I will quote you so you will be notified.  Probably over the next couple of days.    Each stack has about 3.5 hours.  I understand dithering helps--which I did not do.  Will start using that with the np101is.

Thanks

[Uranium235]

Might be better if you can post the raw master Ha stack (unprocessed) - fits or tif will do. That will give us some insight into whether you have pushed it too hard or not.

With this target its possible to use that Ha stack as a luminance master, then layer in the colour (using Ps) which gives you a good degree of control over any noise. You will probably need to surrender star colour, but thats usually the case when doing non-rgb processes.

[gnomus]

Rodd

Like you I welcome constructive criticism, but I wonder if you are being a bit hard on yourself.  Your image has a lot going for it.  It still astounds me that we can get images like yours from our backyards.

I've only tried one Hubble Palette myself (also of the Elephant) and found it magnitudes harder than LRGB.  Colour balance and noise were my main problems.  I had about the same total exposure time as you. 

I have PI and find that it is very easy to be too aggressive with it, especially with things like OIII and SII where I had very little signal (compared with my Ha).  PI is fantastic for many things, but I found I got the best result when I processed the individual channels in PI but then combined and did further tweaking in Photoshop.  The final colour balance was all down to Photoshop.  I followed this tutorial:  http://bf-astro.com/hubblep.htm

I know that you don't have PS but I wonder if you should consider getting hold of a copy - .  

If you want to see my final attempt click on this thumbnail:  

I think I have overstretched and certainly "over-brightened" things.  There were also some technical issues (the data was captured using two different telescopes).  So it's another one that I've put on the "must do again" list.

[Rodd]

Rodd

Like you I welcome constructive criticism, but I wonder if you are being a bit hard on yourself.  Your image has a lot going for it.  It still astounds me that we can get images like yours from our backyards.

I've only tried one Hubble Palette myself (also of the Elephant) and found it magnitudes harder than LRGB.  Colour balance and noise were my main problems.  I had about the same total exposure time as you. 

I have PI and find that it is very easy to be too aggressive with it, especially with things like OIII and SII where I had very little signal (compared with my Ha).  PI is fantastic for many things, but I found I got the best result when I processed the individual channels in PI but then combined and did further tweaking in Photoshop.  The final colour balance was all down to Photoshop.  I followed this tutorial:  http://bf-astro.com/hubblep.htm

I know that you don't have PS but I wonder if you should consider getting hold of a copy - .  

If you want to see my final attempt click on this thumbnail:  

I think I have overstretched and certainly "over-brightened" things.  There were also some technical issues (the data was captured using two different telescopes).  So it's another one that I've put on the "must do again" list.

Nice image--I can see how one might see a similarity between yours and mine--I do at first glance.  And I am amazed that I can take pics from my backyard that rival what Palomar could do in the 1940s.  I fing RGB harder to capture.  I did an RGB (HaRGB actually) of the Elephants Trunk and it was nowhere as good as the NB--I think there were 3.5 hours only, but the nebulosity was very faint, the stars were multitudinous, and it was not very satisfying.   I am starting  a new discussion in which I upload the Master stacks--calibrated and aligned but not processed (they look near black).  I will put them in FITS as apposed to the XISF format PixinSight has gone to.  Have fun.

[Rodd]

Nice image--I can see how one might see a similarity between yours and mine--I do at first glance.  And I am amazed that I can take pics from my backyard that rival what Palomar could do in the 1940s.  I fing RGB harder to capture.  I did an RGB (HaRGB actually) of the Elephants Trunk and it was nowhere as good as the NB--I think there were 3.5 hours only, but the nebulosity was very faint, the stars were multitudinous, and it was not very satisfying.   I am starting  a new discussion in which I upload the Master stacks--calibrated and aligned but not processed (they look near black).  I will put them in FITS as apposed to the XISF format PixinSight has gone to.  Have fun.

OK--I'll get PS-0-but what version  lightbox, CC--which one should I get?