IC 417 & NGC 1931 - 'The Spider and the Fly' from DSW

[gnomus]

I am processing more of the data from the Deep Sky West Tak 106.  This was collected in January 2017.  Filters are 5nm Astrodons.  Camera is a QSI 683.  The Tak 106 is un-reduced. Data as follows:

• SII: 16 x 1800s
• Ha: 21 x 1800s
• OIII: 17 x 1800s

For a total of 27 hours.

It was put together using the SHO-AIP script in PixInsight.  Then it was back and forth (several times) between PixInsight and Photoshop, trying to get something that looked OK.  I also went back and forth on how much 'processing' I should do on this - I had more garish versions, but eventually settled on this one.  

 

 

[PatrickGilliland]

Done well with this Steve, I actually processed and sent straight to the archives tbh.  Was not impressed with my work!  You have done a nice job though and have some nice colours.

Is that a masked stretch or combo of masked and some AH?  Looking that star near the middle it may be if so you have done a even better job as I struggle to find the mid-range when I use.

Paddy

[gnomus]

Thanks for your comment Paddy.  It is interesting to hear that you struggled with this one.  Given the 27 hours of data, I was expecting processing to be straightforward, but it was anything but.  Most of my efforts ended up looking over-processed and 'cartoon-y'.  I wondered if this effort was understated (it probably is), but I am trying to achieve a 'natural' look, even if that does mean a lack of 'drama'.  I didn't use Masked Stretch.  These are just a few stretches in Histogram Transformation.  I have read that it is best to do HT in a 'oner', but in practice I am finding that this bloats up the stars a bit too much for my taste (probably just me being heavy-handed).  I then went into PS and applied various saturation techniques as described in the 'Lessons from the Masters' book.   I am going to have to read Keller's PI book on Masked Stretch - I am not sure that I ever like the result it gives.

Steve     

[wxsatuser]

I don't know this object but it looks very good.

May be just a bit brighter and a tad more colour, but I'm no expert on narrowband.

[wimvb]

A beautiful image. I am looking at it from my mobile device,so I'm missing all the fine details, but will have a look at the full size on my computer later.

As for masked stretch, try increasing the clipping fraction to 0.01(yes, a whopping 1%) just for fun. I generally find the tools in PI very adaptable, and going outside the norm can give unexpected results. (Sometimes good, sometimes bad, but it's worth experimenting.)

Cheers,

[Barry-Wilson]

Well done Steve, lovely delicate detail and not forced at all.  Certainly not cartoon - quite 'natural' as though peeking through the eyepiece.

When I use HT, I do use it iteratively - the first stretch is often the largest.  I always use HT for Lum and NB data.  More often than not Masked Stretch for RGB (sometimes combine with a HT stretch of the RGB data).  I use ExponentialStretch too from time-to-time after an initial stretch to non-linear.  Masked Curves is also a good technique to refine the stretch.

Interesting to see that you have used the SHO-AIP script.  After early experiments with this I have settled on PixelMath.  perhaps I am getting into a processing 'rut' and maybe should experiment again with this script.  ColourMask script can be useful to adjust colour afterwards, if you remain processing within PI, that is.

Fascinating to see a NB version of this target after studying so hard for my own HaRGB image of this target in December 2016.  Did you produce a classic HST palette in the early stages of processing?  Interesting to see the SII and OIII data.

[swag72]

Nice and subtle processing - I like it.

As an aside Steve, I'm interested that you've gone the DSW way ..... I have to say that I couldn't be processing data that lots of other people also have access to, but that's just me. There's many times when you look around on AB and can tell when some DSW data has come in as there's a load of processes using the same data with, at times. almost negligible differences..... Do you have any thoughts on that? 

[wimvb]

Now from my laptop:

Better (more delicate) than this rendering, imo:

https://apod.nasa.gov/apod/ap100113.html

 

[gnomus]

28 minutes ago, Barry-Wilson said:

Well done Steve, lovely delicate detail and not forced at all.  Certainly not cartoon - quite 'natural' as though peeking through the eyepiece.

When I use HT, I do use it iteratively - the first stretch is often the largest.  I always use HT for Lum and NB data.  More often than not Masked Stretch for RGB (sometimes combine with a HT stretch of the RGB data).  I use ExponentialStretch too from time-to-time after an initial stretch to non-linear.  Masked Curves is also a good technique to refine the stretch.

Interesting to see that you have used the SHO-AIP script.  After early experiments with this I have settled on PixelMath.  perhaps I am getting into a processing 'rut' and maybe should experiment again with this script.  ColourMask script can be useful to adjust colour afterwards, if you remain processing within PI, that is.

Fascinating to see a NB version of this target after studying so hard for my own HaRGB image of this target in December 2016.  Did you produce a classic HST palette in the early stages of processing?  Interesting to see the SII and OIII data.

Barry thanks for tour kind words and thanks especially for your detailed comment.  I found it a fascinating read.  I am sure that I had read (in more than one place) that the HT stretch is best done as one stretch.  I had stuck at this way of doing it largely because I thought this is how it was supposed to be done.  But the stars don't half bloat out when I do it this way.  As such, most of my stretching is done iteratively in PS.  Now that you have given me permission   to do it the same way in PI, I may try again.  One thing I like doing in PS is going back and forth between the 100% (1:1) view and the overall view as I stretch, so that I can keep an eye on the stars.  I haven't found a way of doing this in the PI Preview window.  It would help me a lot if I could.

I don't know how the PI script works, but I assumed it was giving a sort of 'Pixel Math For Dummies' approach.  One thing I like about the script is that you can very quickly try several blends and get a near instant idea of how it will turn out.  The downside is I sometimes forget exactly which recipe I used!  I am pretty sure this was 85% SII and 15% Ha in Red, around 85% Ha and 15% OIII in Green and 100% OIII in blue, but I couldn't swear to it.  I know how to do this this in Pixel Math, but goodness me, we're in the 21st Century!!  

[gnomus]

33 minutes ago, swag72 said:

Nice and subtle processing - I like it.

As an aside Steve, I'm interested that you've gone the DSW way ..... I have to say that I couldn't be processing data that lots of other people also have access to, but that's just me. There's many times when you look around on AB and can tell when some DSW data has come in as there's a load of processes using the same data with, at times. almost negligible differences..... Do you have any thoughts on that? 

Sara, thank you for your kind words.

I know exactly what you mean about DSW.  I was concerned that there are too many folks sharing the same data and this put me off going that route for some time.  But, the last 2-3 months have seen me experience extreme data poverty.  What's a chap to do - I couldn't just spend all my time reprocessing old data.  So I took the plunge.  Ideally, I'd like my own remote setup, or a setup with 2-3 'like-minded' folks somewhere with decent weather.  My main issue is that I am such an odd chap that the 'like-minded' are few and far between.  Anyway, I figured that 12 months with DSW was worth a try.  I might even keep it going if I do get something remote - I'll be retiring soon and I don't especially like Bowls.

[gnomus]

38 minutes ago, wimvb said:

Now from my laptop:

Better (more delicate) than this rendering, imo:

https://apod.nasa.gov/apod/ap100113.html

 

That's very kind of you Wim.  I'm not sure that I am quite in APOD territory, but it's nice to have the compliment.

Steve

[Barry-Wilson]

What I do when using HT is click the 'tick' mark in the bottom control bar so that the HT is tracking the image window I am manipulating.  Then Preview, expand the preview window so it is a goodly size.  Then when you stretch you will see the preview change (I'm sure you know this).  Then when you are satisfied with that particular stretch, click the square apply button; this will apply the preview to the original.  Then click the Refresh symbol, don't close the preview, then re-stretch and repeat.

Not sure if this is what you mean, but the technique above allows you to assess the degree of stretch by getting immediate feedback from the preview.  You could arrange the images such that you could see an STF stretch of a clone of the linear image as a reference.  Maybe move the relevant image windows into Workspace 2 so you de-clutter the desktop and allow the max space for all of the windows.  You can click on/off the preview symbol (top left circle) in the Preview window to help assess the change that stretch has made also.

This also is a convenient excuse to buy a super large monitor to arrange all those windows .

A second processing PC may be required anyway to avoid process-queuing in the Gnomus' household as I'm reading this afternoon. . . . oh no, another convenient excuse to buy more gear and SW .

It's so easy to spend other people's money, don't you find .

Ha, ha!

[gnomus]

8 minutes ago, Barry-Wilson said:

What I do when using HT is click the 'tick' mark in the bottom control bar so that the HT is tracking the image window I am manipulating.  Then Preview, expand the preview window so it is a goodly size.  Then when you stretch you will see the preview change (I'm sure you know this).  Then when you are satisfied with that particular stretch, click the square apply button; this will apply the preview to the original.  Then click the Refresh symbol, don't close the preview, then re-stretch and repeat.

Not sure if this is what you mean, but the technique above allows you to assess the degree of stretch by getting immediate feedback from the preview.  You could arrange the images such that you could see an STF stretch of a clone of the linear image as a reference.  Maybe move the relevant image windows into Workspace 2 so you de-clutter the desktop and allow the max space for all of the windows.  You can click on/off the preview symbol (top left circle) in the Preview window to help assess the change that stretch has made also.

...

Thanks Barry

No.  What I mean is that when you bring up the preview, it is always in 'Zoom To Fit' mode - that is you see the whole image.  In Photoshop, I will zoom in and out during stretches.  Often I will do the stretch at 100% (that means - I think - that a pixel on the image equals a pixel on screen so I am 'zoomed in') - this way I can see the effect of the stretch on individual stars.  The equivalent to 100% in PixInsight would be the ratio 1:1 shown in the top blue bar of the image window.  If I could zoom in and out of the preview so that I can also see this at 1:1 instead of 'Zoom to Fit' I would find things much easier.  You will probably tell me that this is very easy, but if it is I haven't found the key combination yet.   (Incidentally, I do know that I can have a larger preview window - I always make it as large as I can.)

11 minutes ago, Barry-Wilson said:

It's so easy to spend other people's money, don't you find .

Ha, ha!

I use two monitors in my set up.

[swag72]

50 minutes ago, gnomus said:

''''''' Ideally, I'd like my own remote setup, or a setup with 2-3 'like-minded' folks somewhere with decent weather.  My main issue is that I am such an odd chap that the 'like-minded' are few and far between. ......

I'm sure you could find some like minded individuals - It strikes me that there are some on SGL already

[johnrt]

26 minutes ago, swag72 said:

I'm sure you could find some like minded individuals - It strikes me that there are some on SGL already

Yup, I'd love to get involved in a remote set-up, but wouldn't quite know where to begin!

[gnomus]

7 minutes ago, johnrt said:

Yup, I'd love to get involved in a remote set-up, but wouldn't quite know where to begin!

I think the saying goes, 'Be careful what you wish for ....'.  This could start getting expensive....