#1  
Old 28-08-2014, 08:01 PM
lazjen's Avatar
lazjen (Chris)
PI cult member

lazjen is offline
 
Join Date: Dec 2012
Location: Flaxton, Qld
Posts: 1,748
Question Perplexed by Image Processing Result

I need some help understanding what I'm doing wrong, because I just can't see anything reasonable at the moment.

Refer to the attached images for my perplexing problem.

The first is a number of 300 sec Ha stacked with no bias or darks. The only "post processing" was auto tone/contrast in PS.

The second is the same Ha images, but I added my 30 bias frames and 20 300 sec dark frames to it. Again, the only "post processing" was auto tone/contrast in PS. This is not what I was expecting - it's completely destroyed the image.

Click image for larger version

Name:	stacked_ha.jpg
Views:	32
Size:	167.9 KB
ID:	168964 Click image for larger version

Name:	wrecked_ha.jpg
Views:	32
Size:	43.3 KB
ID:	168965

I know the bias frames are ok, because I used them in my recent Lagoon Ha image post (in the beginners section).

Given that I thought the darks were suspect. So I created 3 more tonight as a sample to test - same type of result.

I used the same camera both times, and at the same temperature (-15C).

I was using PI (still learning), so I thought I made a mess there. I get the same results using DSS.

The only thing I can think of that is significantly different (besides exposure time), is that I did the successful 30 sec fits images using linux/ekos, while I did the 300 sec fits images using Windows/APT. I can see some differences in the header, but nothing jumps out at me as being "wrong".

I've included the header info below from one exposure from the 30 and 300 sec. If this is the root cause, what can I do about it?

FITS header 30 sec (ekos)
Code:
================================================================================
FITS header number 1 at block number 1.
--------------------------------------------------------------------------------
SIMPLE  =                    T / file does conform to FITS standard
BITPIX  =                   16 / number of bits per data pixel
NAXIS   =                    2 / number of data axes
NAXIS1  =                 1530 / length of data axis 1
NAXIS2  =                 1020 / length of data axis 2
EXTEND  =                    T / FITS dataset may contain extensions
COMMENT   FITS (Flexible Image Transport System) format is defined in 'Astronomy
COMMENT   and Astrophysics', volume 376, page 359; bibcode: 2001A&A...376..359H
BZERO   =                32768 / offset data range to that of unsigned short
BSCALE  =                    1 / default scaling factor
EXPTIME =                  30. / Total Exposure Time (s)
PIXSIZE1=                   9. / Pixel Size 1 (microns)
PIXSIZE2=                   9. / Pixel Size 2 (microns)
XBINNING=                    1 / Binning factor in width
YBINNING=                    1 / Binning factor in height
FRAME   = 'Light   '           / Frame Type
DATAMIN =                1038. / Minimum value
DATAMAX =               56709. / Maximum value
INSTRUME= 'SBIG ST-8 Dual CCD Camera' / CCD Name
DATE-OBS= '2014-08-01T11:39:47' / UTC start date of observation
CCD-TEMP=    -14.8095167806398 / CCD Temperature (Celcius)
END
--------------------------------------------------------------------------------
Data section number 1 beginning at block number 2.
Skipped 1084 blocks of data of size 2880 bytes (3121920 bytes). 
================================================================================
End-of-file after 1 HDU in 1085 x 2880-byte blocks (3124800 bytes).
--------------------------------------------------------------------------------
FITS Header 300sec (APT)
Code:
================================================================================
FITS header number 1 at block number 1.
--------------------------------------------------------------------------------
SIMPLE  =                    T / file does conform to FITS standard
BITPIX  =                   16 / number of bits per data pixel
NAXIS   =                    2 / number of data axes
NAXIS1  =                 1530 / length of data axis 1
NAXIS2  =                 1020 / length of data axis 2
EXTEND  =                    T / FITS dataset may contain extensions
COMMENT = 'SBIG FITS header format per:'
COMMENT = ' http://www.sbig.com/pdffiles/SBFITSEXT_1r0.pdf'
BZERO   =                32768 / offset data range to that of unsigned short
BSCALE  =                    1 / default scaling factor
OBJECT  = '        '
TELESCOP= '        '
INSTRUME= 'SBIG ST-8 Dual CCD Camera' / Camera Model
OBSERVER= '        '
DATE-OBS= '2014-08-25T10.42.01.000' / GMT START OF EXPOSURE
EXPTIME =                 300. / EXPOSURE IN SECONDS
CCD-TEMP=    -15.2331909548563 / CCD TEMP IN DEGREES C
XPIXSZ  =                   9. / PIXEL WIDTH IN MICRONS
YPIXSZ  =                   9. / PIXEL HEIGHT IN MICRONS
XBINNING=                    1 / HORIZONTAL BINNING FACTOR
YBINNING=                    1 / VERTICAL BINNING FACTOR
XORGSUBF=                    0 / SUB_FRAME ORIGIN X_POS
YORGSUBF=                    0 / SUB_FRAME ORIGIN Y_POS
EGAIN   =                 2.59 / ELECTRONS PER ADU
FOCALLEN=                   0. / FOCAL LENGTH IN MM
APTDIA  =                   0. / APERTURE DIAMETER IN MM
APTAREA =                   0. / APERTURE AREA IN SQ-MM
CBLACK  =                    0 / BLACK ADU FOR DISPLAY
CWHITE  =                65535 / WHITE ADU FOR DISPLAY
PEDESTAL=                 -100 / ADD TO ADU FOR 0-BASE
DATAMAX =                65535 / SATURATION LEVEL
SBSTDVER= 'SBFITSEXT Version 1.0' / SBIG FITS EXTENSIONS VER
SWACQUIR= 'Astro Photography Tool - APT v.2.50' / DATA ACQ SOFTWARE
SWCREATE= 'Astro Photography Tool - APT v.2.50'
FILTER  = 'Ha      '           / OPTICAL FILTER NAME
SNAPSHOT=                    1 / NUMBER IMAGES COADDED
DATE    = '2014-08-25'         / GMT DATE WHEN THIS FILE CREATED
RESMODE =                    0 / RESOLUTION MODE
EXPSTATE= '125     '           / EXPOSURE STATE (HEX)
RESPONSE=                2000. / CCD RESPONSE FACTOR
NOTE    = 'Local time:8/25/2014 at 20:42:01'
TRAKTIME=                   0. / TRACKING EXPOSURE
END
--------------------------------------------------------------------------------
Data section number 1 beginning at block number 3.
Skipped 1084 blocks of data of size 2880 bytes (3121920 bytes). 
================================================================================
End-of-file after 1 HDU in 1086 x 2880-byte blocks (3127680 bytes).
--------------------------------------------------------------------------------
Any ideas?
Reply With Quote
  #2  
Old 28-08-2014, 08:26 PM
RickS's Avatar
RickS (Rick)
PI cult recruiter

RickS is offline
 
Join Date: Apr 2010
Location: Brisbane
Posts: 10,584
Hi Chris,

Looks like your files are 16-bit signed integer from the BSCALE/BZERO values. Try the settings I suggested here and see if they help with your existing data: http://pixinsight.com/forum/index.php?topic=7445.0

If not, I'd be happy to take a look at some of your files early next week (I hope to be imaging this weekend!) and see if I can figure out what is going wrong.

Cheers,
Rick.
Reply With Quote
  #3  
Old 28-08-2014, 09:12 PM
lazjen's Avatar
lazjen (Chris)
PI cult member

lazjen is offline
 
Join Date: Dec 2012
Location: Flaxton, Qld
Posts: 1,748
Thanks for the reply Rick. Unfortunately that didn't seem to make any difference at all to my images, although I'll retry on the weekend sometime.

I'm hoping to be out collecting data too for some of this weekend.
Reply With Quote
  #4  
Old 29-08-2014, 03:45 AM
Shiraz's Avatar
Shiraz (Ray)
Registered User

Shiraz is offline
 
Join Date: Apr 2010
Location: ardrossan south australia
Posts: 4,796
Just a possibility, but it might also depend on whether your darks have had bias removed or not. If they still have bias, you need to check the "calibrate" box in the master dark block (when doing the calibration of the lights), or you could end up subtracting the bias twice and destroying the images.
Reply With Quote
  #5  
Old 29-08-2014, 05:46 AM
troypiggo's Avatar
troypiggo (Troy)
Bust Duster

troypiggo is offline
 
Join Date: Nov 2008
Location: Brisbane, Australia
Posts: 4,837
If you've used different image capture software for the lights, darks, and biases, perhaps check the pedestal settings in each of the different software used. See the APT ones mentions a pedestal of -100, while Ekos doesn't mention at all.
Reply With Quote
  #6  
Old 29-08-2014, 05:48 AM
lazjen's Avatar
lazjen (Chris)
PI cult member

lazjen is offline
 
Join Date: Dec 2012
Location: Flaxton, Qld
Posts: 1,748
Thanks for that thought Ray, but I forgot to mention I think I did what you said, but I've also done attempts not using bias at all - just the lights and darks.

In fact, until I sort out getting the flat frames again, the bias frames probably aren't helping at all.

I was thinking this morning about the problem and Rick's idea. Another thing I forgot to mention is that all the bias and darks were taken using Linux/ekos. I think I'll try creating a few darks in Windows/APT and see if that makes a difference. If it does, then that will be quite annoying but something I can at least handle.
Reply With Quote
  #7  
Old 29-08-2014, 05:50 AM
lazjen's Avatar
lazjen (Chris)
PI cult member

lazjen is offline
 
Join Date: Dec 2012
Location: Flaxton, Qld
Posts: 1,748
Quote:
Originally Posted by troypiggo View Post
If you've used different image capture software for the lights, darks, and biases, perhaps check the pedestal settings in each of the different software used. See the APT ones mentions a pedestal of -100, while Ekos doesn't mention at all.
Ok, that sounds good. I don't recall seeing any pedestal settings in APT. Is there some way to "normalise" these?
Reply With Quote
  #8  
Old 29-08-2014, 06:01 AM
RickS's Avatar
RickS (Rick)
PI cult recruiter

RickS is offline
 
Join Date: Apr 2010
Location: Brisbane
Posts: 10,584
Quote:
Originally Posted by lazjen View Post
Ok, that sounds good. I don't recall seeing any pedestal settings in APT. Is there some way to "normalise" these?
Assuming a pedestal of 100, the Pixelmath expression $T-100 would do it.

Cheers,
Rick.
Reply With Quote
  #9  
Old 29-08-2014, 06:04 AM
lazjen's Avatar
lazjen (Chris)
PI cult member

lazjen is offline
 
Join Date: Dec 2012
Location: Flaxton, Qld
Posts: 1,748
Quote:
Originally Posted by RickS View Post
Assuming a pedestal of 100, the Pixelmath expression $T-100 would do it.
Woot, into the deep end with PI.

So, I'd do the Pixelmath on my darks/bias and leave the lights alone (since they've got a pedestal setting -100)? Or have I got that around the wrong way?
Reply With Quote
  #10  
Old 29-08-2014, 07:50 AM
RickS's Avatar
RickS (Rick)
PI cult recruiter

RickS is offline
 
Join Date: Apr 2010
Location: Brisbane
Posts: 10,584
Quote:
PEDESTAL= -100 / ADD TO ADU FOR 0-BASE
This means that you need to subtract 100 to remove the pedestal.

I'd take a look at the average (mean and median) and minimum values in each type of frame just to confirm what is going on.
Reply With Quote
  #11  
Old 30-08-2014, 09:08 AM
lazjen's Avatar
lazjen (Chris)
PI cult member

lazjen is offline
 
Join Date: Dec 2012
Location: Flaxton, Qld
Posts: 1,748
I'll be trying out this stuff during the next lot of bad weather nights. I did get some darks using APT and they work fine with the lights, so it's definitely the mixed applications that causes the grief.

Thanks for helping me get there!
Reply With Quote
Reply

Bookmarks

Thread Tools
Rate This Thread
Rate This Thread:

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT +10. The time is now 04:01 AM.

Powered by vBulletin Version 3.8.7 | Copyright ©2000 - 2020, Jelsoft Enterprises Ltd.
Advertisement
Astromechanics
Advertisement
Lunatico Astronomical
Advertisement
Bintel
Advertisement
Testar
Advertisement
EQ8-R
Advertisement
Celestron RASA
Advertisement
Star Adventurer
Advertisement
Astronomy and Electronics Centre
Advertisement