Jump to content

Banner.jpg.b89429c566825f6ab32bcafbada449c9.jpg

Recommended Posts

NGC 2014 and Dragon's Head nebula in the Magellanic Cloud ( LMC ) not far from the Tarantula Nebula by Mike O'Day ( 500px.com/MikeODay ).

This image shows multiple bright nebula and star clusters in an area adjacent to the The Tarantula Nebula ( NGC 2070 ) in the nearby irregular galaxy, the Large Magellanic Cloud ( LMC ). The largest of these are the bright pink nebula in the mid-right part of the image ( NGC 2014 ) and the blue nebula in the lower middle ( NGC 2030 ).

.....

Updated image - reprocessed to impove colour balance ( April 15th )

58f156f1b8cb5_NGC2014intheLargeMagellanicCloud-170222Dcompressed.thumb.jpg.c7338e58904d730f0d523a4d9a96f8f9.jpg

( please click / tap on image to see it larger and sharper )

....

Original:

58bb7b8ce4917_NGC2014intheLargeMagellanicCloud-170222Ccompressed.thumb.jpg.7585ece4ec93aed950b0dc434b74a8ce.jpg

 

( click on image to see larger and clearer ( grrr... image compression in version above :( ))

----------

This is the first image captured as part of a new image capture and processing workflow I am trying out...

Roger Clark ( http://www.clarkvision.com/articles/index.html ) has a number of articles addressing colour processing and the performance of modern DSLR sensors.  The "take homes" for me have been:

1. With a modern sensor ( one with on-sensor dark current suppression technology ) one may not need to capture dark frames ( in order to remove the now non-significant pattern noise ).

2. "Correct" white balance processing should start by using "daylight" RBG channel multipliers ( to get the star colour 'right') and any histogram adjustment to improve white balance of darker parts of the image should involve aligning the left side of histogram curves ( ie. not the peaks )

So, the workflow to produce the image above involves calibraiton with Superbias & Master Flat but no dark frame subtraction (neither post nor in-camera).  Roger Clark speaks of using a "bad pixel map" as the basis of reducing hot pixels.  I have not figured out how to produce one yet.  However, with a little bit of dithering during guiding ( and the very busy image :) ) the hot pixels that are in the image below are not too overwhelming.

With regard to colour balance; I tried using the "daylight" factors reported by the camera but these resulted in an image and stars that were quite blue.  This image was based on the factors reported by DXOMark for the Nikon D5300 ( R x 2.12, G x 1, B x 1.49 ( D50 standard )).  This was better but I still felt the need for a final tweak in Photoshop ( colourBalance Highlights +15 Cyan/Red, -5 Magenta/Green ) to improve the colour in the stars and mid-tones.

{ DXOMark "white balance scales" for D5300 found at: https://www.dxomark.com/Cameras/Nikon/D5300---Measurements on the "color response" tab }

---------

Details:

Bright Nebulae:
NGC 2014 ( upper right, pink) size 30 x 20 arcmin Mag +8
NGC 2020 size 2.0 arcmin ( small blue-green oval nebula )
NGC 2030 
NGC 2032 ( Dragon's Head nebula - blue, central bottom of image )
NGC 2035 size 3.0 x 3.0 arcmin 
NGC 2040 size 3.0 x 3.0 arcmin 

Open clusters:
NGC 2004 size 2.7 arcmin Mag +9.6
NGC 2011 size 1 arcmin Mag +10.6
NGC 2021 size 0.9 arcmin Mag +12.1

Annotated :

IMG_0596.JPG.c20bf9f94b721def2d7428f5011cb4cc.JPG

Image centre RA 05h 33m 32.362s, Dec -67° 32' 18.145" (nova.astrometry.net) 
Orientation: up is west, right is South 
Field of view (arcmin): 58.8 x 39.2
Scale (full size image) 0.585 arcsec/pixel

Telescope: Orion Optics CT12 Newtonian ( mirror 300mm, fl 120mm, f4 ).
Corrector: ASA 2" Coma Corrector Quattro 1.175x.
Effective Focal Length / Aperture : 1410mm f4.7

Mount: Skywatcher AZ Eq6 GT
Guiding: TSOAG9 Off-Axis-Guider, Starlight Xpress Lodestar X2, PHD2 

Camera: Nikon D5300 (unmodified) (sensor 23.5 x 15.6mm, 6016x4016 3.9um pixels)
Format: 14bit NEF
Noise reduction: off
Filter: none

HDR combination of seven sets of exposures (20 & 22 Feb 2017):
58 x 240 sec ISO 800
8 x 120 sec ISO 800
8 x 60 sec ISO 800
8 x 30 sec ISO 800
8 x 14 sec ISO 800
8 x 7 sec ISO 800
8 x 3 sec ISO 800

Pixinsight: 26 Feb 2017

Links:

500px.com/MikeODay
photo.net/photos/MikeODay

 

Link to comment
Share on other sites

  • Replies 40
  • Created
  • Last Reply

Looks really good. The background looks kind if red, but that may very well be caused by weak Ha.

2 hours ago, mike005 said:

 

Roger Clark ( http://www.clarkvision.com/articles/index.html ) has a number of articles addressing colour processing and the performance of modern DSLR sensors. :

1. With a modern sensor ( one with on-sensor dark current suppression technology ) one may not need to capture dark frames ( in order to remove the now non-significant pattern noise ).


Noise reduction: off

 

Isn't that contradictory to the workflow? (I.e., you turn off noise reduction that will reduce fuxed pattern noise) Or are they two different things?

I made hot pixel maps in PixInsight by binarising a master dark. Clipped the low signal such that only bright hot pixels were left. It needed some experimenting to get it right.

Link to comment
Share on other sites

7 hours ago, mikeyscope said:

Great image and a very busy area of sky ...that NGC 2020 looks an interesting object in itself, another 'bubble' type nebula...

Thanks Mikey.  Yes, NGC 2020 is interesting; sort of like a bursting bubble - I'll have to look it up and see if I can find a detailed image.

Link to comment
Share on other sites

7 hours ago, wimvb said:

Looks really good. The background looks kind if red, but that may very well be caused by weak Ha.

Isn't that contradictory to the workflow? (I.e., you turn off noise reduction that will reduce fuxed pattern noise) Or are they two different things?

I made hot pixel maps in PixInsight by binarising a master dark. Clipped the low signal such that only bright hot pixels were left. It needed some experimenting to get it right.

Thanks Wim.

I'm not totally sure about the red.  I think you are right and that it is Ha - there seems to be a pattern to it.  

Sorry, my sentence was not very clear...

I think what Roger Clark is saying is that modern DSLR cameras have in sensor dark-current suppression.  This means that they should have negligible fixed pattern thermal noise and since darks are used to remove/reduce fixed pattern thermal noilse ( as oppossed to random thermal noise ) then, so long as hot pixels are not a problem ( either reduced during capture via dithering or using a bad pixel map ), it is not beneficial to subtract darks from the lights ( in fact, without pattern noise, subtracting darks adds more noise than it reduces ).  

Up until now I have done dark subraction in-camera via Long Exposure Noise Reduction.  For this image I turned it off.

Having said all that, I'm not sure that the D5300 does in fact have adequate ( or any ) dark current suppression - I am still experimenting.

Cheers

Mike

Link to comment
Share on other sites

19 hours ago, wimvb said:

 The background looks kind if red, but that may very well be caused by weak Ha.

 

I think it is Ha.

I went looking for images and this excellent annotated image of the LMC seems to show ( low centre left ) an extensive Ha region across the relevant parts of my image 

IMG_0562.JPG

"This image shows the entire Large Magellanic Cloud, with some of the brightest objects marked..."

Credit:

Rober Gendler/ESO

https://www.spacetelescope.org/images/heic1011d/

 

Link to comment
Share on other sites

Nice image, good to see these DSOs from southern skies, but does the backgound perhaps have a bit of a purplish cast?

23 hours ago, mike005 said:

This is the first image captured as part of a new image capture and processing workflow I am trying out...

Roger Clark ( http://www.clarkvision.com/articles/index.html ) has a number of articles addressing colour processing and the performance of modern DSLR sensors.  The "take homes" for me have been:

 

Please bear in mind, I've heard various people warn against the advice on Mr Clark's site - here's an example thread.

Link to comment
Share on other sites

4 hours ago, Knight of Clear Skies said:

Nice image, good to see these DSOs from southern skies, but does the backgound perhaps have a bit of a purplish cast?

Please bear in mind, I've heard various people warn against the advice on Mr Clark's site - here's an example thread.

Yes, that is a nice image!

That thread about ClarkVision was interesting and taking a look at Mr. Clark's own gallery kind of talks against using his methods.......

Link to comment
Share on other sites

17 hours ago, Knight of Clear Skies said:

Nice image, good to see these DSOs from southern skies, but does the backgound perhaps have a bit of a purplish cast?

Please bear in mind, I've heard various people warn against the advice on Mr Clark's site - here's an example thread.

Thanks - I really love this region of the sky; it is a shame that Summer has been so sub-tropical in Sydney - with wet and/or humid nights resulting in only a few good "astrophoto" nights since Christmas :(

I'm still not sure about the background.

This area seems to have a large diffuse Ha emision nebula that covers most if not all of the frame of my image ( see link above to ESO image ).  Having said that, I am still experimenting with my workflow and it is quite possible that I have messed up :)

I did look at the link - clearly a number of people who passionately believe that they are right and Roger Clark is wrong.  On his website Dr Clark also has the odd not-so-subtle dig at "internet experts".  Me, I'm am just looking for good ideas to improve my workflow ...

If you look through the images I have posted on this site over time you will see a huge range in the colours of  stars - with "yellow" stars ranging from green-yellow through to magenta.  I am attracted to Dr Clark's "star colour" ideas because they have the potential to provide consistency in the representation of star colours in my images.  That is, previously, I effectively adusted the individual RGB factors to force the histogram peaks to align (ie. assumes that the background is grey ) and, with the actual backgrounds having very wide colour ranges, this results in images with neutral backgrounds but inaccurate star colours.  I want to start with good star colour and then tweak the background to get as close to "real" as I can.

I no doubt have many rainy nights of experimenting to go :)

Link to comment
Share on other sites

I think the star colour might be getting close to right ...

In the image above HD 269771 should have a spectrum of M0 and colour index of +1.38  orange ( ref: SkySafari ) and from a link on Mitchel Charitiy's site ( http://www.vendian.org/mncharity/dir3/starcolor ) the star should have an RGB pixel value of near to #ffd39d or 255 211 157

IMG_0570.JPG   IMG_0567.JPG 

And HD 269777 should be B -0.18 blue-white - ~ #b9cdff or 86 205 255

IMG_0574.JPG IMG_0576.JPG

Mmmm, still need to think about the background ...

Link to comment
Share on other sites

I think you are correct about the background. While you can make it more neutral with tools like dbe or background neutralization in PixInsight, the question is if you should. Imo, the background can stay a little red, just to indicate the presence of weak Ha. This is a decision one should make for each image, and not just apply neutralisation as a standard procedure in a fixed workflow.

See the last image here, for an example.

http://deepskycolors.com/deepspace.html

Link to comment
Share on other sites

22 hours ago, wimvb said:

I think you are correct about the background. While you can make it more neutral with tools like dbe or background neutralization in PixInsight, the question is if you should. Imo, the background can stay a little red, just to indicate the presence of weak Ha. This is a decision one should make for each image, and not just apply neutralisation as a standard procedure in a fixed workflow.

See the last image here, for an example.

http://deepskycolors.com/deepspace.html

Thank you Wim.

I just re-did a part of the processing and if anything the result was lttle more red ( perhaps a little less magenta ).  I believe there is quite a bit of Ha here but there is a possibility that I may be introducing some red during the HDR processing as the short exposures have quite a bit of red read noise (?) that may be creaping in.  I will have a go at fully processing just the 240sec  subs and see what the background looks like.

Anyway, a blend of the original and the new HDR version can be seen below.

 

Link to comment
Share on other sites

Updated version - it is subtle but the stars are a tad smaller with slightly less saturated peak values.  Also, the background is a tiny bit less magenta / more red and there is a very small incrrease in contrast and vibrancy...

NGC 2014 in the Large Magellanic Cloud - 170222 B1 compressed.jpg

Please click on image to see without compression softening 

 

 

Link to comment
Share on other sites

23 hours ago, wimvb said:

I think you are correct about the background. While you can make it more neutral with tools like dbe or background neutralization in PixInsight, the question is if you should. Imo, the background can stay a little red, just to indicate the presence of weak Ha. This is a decision one should make for each image, and not just apply neutralisation as a standard procedure in a fixed workflow.

See the last image here, for an example.

http://deepskycolors.com/deepspace.html

Ok, this image I think confirms that the red is "real": http://deography.com/dragons-head-nebula-ngc2035/ 

So I should stop worrying about it and move on to th next one :)

 

Link to comment
Share on other sites

Hmmm :(

I think I have found the source of the additional  red in the short subs ...

As part of preparing to use the new method I had re-integrated my set of bias frames ( 600 odd ) in order to produce an RGB CFA format bias frame.  I then used the Superbias tool to create a a Superbias file that I used to calibrate all my subs.

It turns out that the Superbias file I created, whilst resetting the black point as it should, is not removing the glow that progressively builds towards the bottom of the frame.  However, i just tested the integrated bias frame ( ie pre Superbias tool file ) and it does completely remove the pattern in the noise ( albeit with a slight increase in overall random noise in the calibrated subs ).

I will now go ahead and reprocess all the subs and see if it has any impact on the final image - it may not because to produce the images above I clipped the shadows of all of the short subs in order to reduce/eliminate the red.

 

Link to comment
Share on other sites

46 minutes ago, mike005 said:

Hmmm :(

I think I have found the source of the additional  red in the short subs ...

As part of preparing to use the new method I had re-integrated my set of bias frames ( 600 odd ) in order to produce an RGB CFA format bias frame.  I then used the Superbias tool to create a a Superbias file that I used to calibrate all my subs.

...

 

???

Not debayered, I hope. Calibration frames should not be debayered. Debayering is done after calibration. I would think that PixInsight takes care of this. Unless you hadn't checked the tic box CFA in the batch script.

But I guess you already know this.

Link to comment
Share on other sites

10 hours ago, wimvb said:

???

Not debayered, I hope. Calibration frames should not be debayered. Debayering is done after calibration. I would think that PixInsight takes care of this. Unless you hadn't checked the tic box CFA in the batch script.

But I guess you already know this.

Thanks Wim.

 

My process for bias creation;

A. set up DSLR conversion ( in Format Explorer ) to:

 - RGB CFA ( not debayered )

- no black point correction

- no colour balance 

- all weights = 1

B. Integrate with no normalisation

 

My process for light subs:

A.  Same DSLR converions as bias frames

B. Calibration - with master bias and master flat

C. White balance with PixelMath:  R x 2.12, G x 1, B x 1.49 ( from DXOMark sensor measurements for D5300 )

D. Use Pixel Math $T x 4 to scale 14 bit images to full 16 bit range ( ie. I discovered that my 14 bit files are converted to the range 0 to 0.25 by Pixinsight and not 0 to 1 )

E. Batch Debayer

F. Integrate each set of subs with same exposure

G. HDR processing 

 

I don't know why the Superbias file I produced does not work properly.  I will have to look into it further.

 

Cheers

Mike

Link to comment
Share on other sites

That's an elaborate process. I just use batch preprocessing script to calibrate my light frames, then image registration and integration. Due to light pollution, colour balance is off anyway, so I let DBE, background neutralization and colour calibration take care of white balance.

Btw, multiplying the pixelvalues by 4 doesn't change anything, other than pushing the data in the higher bits of 16. If before you had a pixelvalue 00x...x, you now have x...x00. But it still is the same data.

 

Link to comment
Share on other sites

8 hours ago, wimvb said:

That's an elaborate process. I just use batch preprocessing script to calibrate my light frames, then image registration and integration. Due to light pollution, colour balance is off anyway, so I let DBE, background neutralization and colour calibration take care of white balance.

Btw, multiplying the pixelvalues by 4 doesn't change anything, other than pushing the data in the higher bits of 16. If before you had a pixelvalue 00x...x, you now have x...x00. But it still is the same data.

 

Yes that is what I used to do  but my star colour was always off.  I don't think DBE can fix star colour.

The plan is that the RGB ratios restore the colour balance of the stars.

Multiplying by 4 is intended to equalise the maximim/clipped values across all channels.   The reason for this is again related to white balance of the stars.   What I found is that as the exposure increases, green is the first to clip.  So suppose one has an image before scaling of R = 0.5, G=0.25 (clipped), B=0.2.  After scaling this becomes, R=0.138, G=0.25 (clipped), B= 2.98.  This means that there are no green values abbove 0.25 and the colour balance of the highlights is wrong.  Multiplying by 4 makes all three chanel clip at 1 and should ensure that the colour balance is right for the highlights (albeit that the high values are clipped for all channels - as of course they would be if the sensitivity of the sensor was the same for all sensor pixels.)    I then use HDR processing to replace the clipped pixels with lower exposure values.

Anyway, that is the thought ... 

Link to comment
Share on other sites

Another complication discovered; scaling by x4 results still results in the clipped values for R, G, B pixels having values less than 1 ...

 

1. Pixinsight converts the 14bit NEF file to a real number in the range 0..1 such that 

0 -> 0.0

16383 -> 0.25

 

2. The bias master frame has a median value of ~ the camera bias 

For the D5300 this is ~ 600 or 600/65535 = 0.00916 on the 0 .. 1 scale

 

3. For any light frame saturated channel, the range after ImageCalibration (with the master bias only) then becomes

0 .. (0.25-0.00916)

so the scaling factor to make the clipped pixels have a value of 1 becomes

1/(0.25-0.00916) = 4.125  (ie. not 4 as I have been using)

 

4. The scale and white balance step then becomes ( in my case )

R:  $T[0]*2.12*4.125

G: $T[1]*4.125

B: $T[2]*1.49*4.125

 

-----------

The upshot of this...

My current draft workflow becomes :

a. Calibrate with master bias

b. Scale and white balance using pixelMath as above

c. Calibrate with master flat

d. Batch Debayer

e. Align

f. Integrate

g. HDR

 

I suppose it will be worth it if I can get finally get consistency in my star and backbground colours ?

 

Link to comment
Share on other sites

To repair stars that become saturated/clipped, I use the HSV repair script (scripts -> utilities).

http://wimvberlo.blogspot.se/2017/01/star-repair-in-pixinsight.html?m=1

http://wimvberlo.blogspot.se/2017/02/star-repair-in-pixinsight-part-2.html?m=1

If you do the repair before colour calibration, it prevents other stars from becoming too strongly coloured.

Link to comment
Share on other sites

41 minutes ago, wimvb said:

To repair stars that become saturated/clipped, I use the HSV repair script (scripts -> utilities).

http://wimvberlo.blogspot.se/2017/01/star-repair-in-pixinsight.html?m=1

http://wimvberlo.blogspot.se/2017/02/star-repair-in-pixinsight-part-2.html?m=1

If you do the repair before colour calibration, it prevents other stars from becoming too strongly coloured.

Thanks for the heads-up Wim, I'll have a look at it.

Cheers

Mike

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

×
×
  • 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.