Image with range 0 1

Questions, support and discussions
rbarbera
Posts: 6
Joined: 18 Dec 2019, 09:01

Image with range 0 1

Post by rbarbera » 18 Dec 2019, 09:11

After a successful align and average stack the image is shown with value 0 and 1. I can see the individual images on the preview (very noisy, but I'm able to apply stretching to it). But the final image is unusable. Any clue on what can be happening here?

one of the images fits header:

SIMPLE = T / file does conform to FITS standard
BITPIX = 8 / number of bits per data pixel
NAXIS = 3 / number of data axes
NAXIS1 = 4928 / length of data axis 1
NAXIS2 = 3264 / length of data axis 2
NAXIS3 = 3 / length of data axis 3
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
INSTRUME= 'Nikon DSLR DSC D7000 (PTP mode)' / CCD Name
TELESCOP= 'SynScan ' / Telescope name
OBSERVER= 'Unknown ' / Observer name
OBJECT = 'm27 ' / Object name
EXPTIME = 3.000000E+01 / Total Exposure Time (s)
PIXSIZE1= 4.780000E+00 / Pixel Size 1 (microns)
PIXSIZE2= 4.780000E+00 / Pixel Size 2 (microns)
XBINNING= 1 / Binning factor in width
YBINNING= 1 / Binning factor in height
XPIXSZ = 4.780000E+00 / X binned pixel size in microns
YPIXSZ = 4.780000E+00 / Y binned pixel size in microns
FRAME = 'Light ' / Frame Type
FOCALLEN= 6.00E+02 / Focal Length (mm)
SCALE = 1.643523E+00 / arcsecs per pixel
SITELAT = 3.947530E+01 / Latitude of the imaging site in degrees
SITELONG= -3.760000E-01 / Longitude of the imaging site in degrees
AIRMASS = 1.593141E+00 / Airmass
OBJCTRA = '19 59 06.18' / Object J2000 RA in Hours
OBJCTDEC= '22 43 43.69' / Object J2000 DEC in Degrees
RA = 2.997758E+02 / Object J2000 RA in Degrees
DEC = 2.272880E+01 / Object J2000 DEC in Degrees
EQUINOX = 2000 / Equinox
CRVAL1 = 2.9977576208E+02 / CRVAL1
CRVAL2 = 2.2728803080E+01 / CRVAL1
RADECSYS= 'FK5 ' / RADECSYS
CTYPE1 = 'RA---TAN' / CTYPE1
CTYPE2 = 'DEC--TAN' / CTYPE2
CRPIX1 = 2.4640000000E+03 / CRPIX1
CRPIX2 = 1.6320000000E+03 / CRPIX2
SECPIX1 = 1.6435234055E+00 / SECPIX1
SECPIX2 = 1.6435234055E+00 / SECPIX2
CDELT1 = 4.5653427930E-04 / CDELT1
CDELT2 = 4.5653427930E-04 / CDELT2
CROTA1 = 8.9411300000E+01 / CROTA1
CROTA2 = 8.9411300000E+01 / CROTA2
DATE-OBS= '2019-12-15T18:15:35.345' / UTC start date of observation
COMMENT Generated by INDI
ISOSPEED= 3200 / ISO Speed
END


Resuts headers:
SIMPLE = T / file does conform to FITS standard
BITPIX = -32 / number of bits per data pixel
NAXIS = 3 / number of data axes
NAXIS1 = 4928 / length of data axis 1
NAXIS2 = 3264 / length of data axis 2
NAXIS3 = 3 / length of data axis 3
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
INSTRUME= 'Nikon DSLR DSC D7000 (PTP mode)' / CCD Name
TELESCOP= 'SynScan ' / Telescope name
OBSERVER= 'Unknown ' / Observer name
OBJECT = 'm27 ' / Object name
PIXSIZE1= 4.780000E+00 / Pixel Size 1 (microns)
PIXSIZE2= 4.780000E+00 / Pixel Size 2 (microns)
XBINNING= 1 / Binning factor in width
YBINNING= 1 / Binning factor in height
XPIXSZ = 4.780000E+00 / X binned pixel size in microns
YPIXSZ = 4.780000E+00 / Y binned pixel size in microns
FRAME = 'Light ' / Frame Type
FOCALLEN= 6.00E+02 / Focal Length (mm)
SCALE = 1.643523E+00 / arcsecs per pixel
SITELAT = 3.947530E+01 / Latitude of the imaging site in degrees
SITELONG= -3.760000E-01 / Longitude of the imaging site in degrees
AIRMASS = 1.593141E+00 / Airmass
OBJCTRA = '19 59 06.18' / Object J2000 RA in Hours
OBJCTDEC= '22 43 43.69' / Object J2000 DEC in Degrees
RA = 2.997758E+02 / Object J2000 RA in Degrees
DEC = 2.272880E+01 / Object J2000 DEC in Degrees
EQUINOX = 2000 / Equinox
CRVAL1 = 2.9977576208E+02 / CRVAL1
CRVAL2 = 2.2728803080E+01 / CRVAL1
RADECSYS= 'FK5 ' / RADECSYS
CTYPE1 = 'RA---TAN' / CTYPE1
CTYPE2 = 'DEC--TAN' / CTYPE2
CRPIX1 = 2.4640000000E+03 / CRPIX1
CRPIX2 = 1.6320000000E+03 / CRPIX2
SECPIX1 = 1.6435234055E+00 / SECPIX1
SECPIX2 = 1.6435234055E+00 / SECPIX2
DATE-OBS= '2019-12-15T18:15:35'
COMMENT Generated by INDI
ISOSPEED= 3200 / ISO Speed
COMMENT 1 Written by Astrometric Stacking Program. www.hnsky.org
CALSTAT = 'S'
HISTORY 1 Stacking method AVERAGE
EXPTIME = 240 / Total luminance exposure time in seconds.
LUM_EXP = 30 / Average luminance exposure time.
LUM_CNT = 8 / Luminance images combined.
LUM_DARK= 0 / Darks used for luminance.
LUM_FLAT= 0 / Flats used for luminance.
LUM_BIAS= 0 / Flat-darks used for luminance.
LUM_TEMP= 999 / Set temperature used for luminance.
BZERO = 0 / Scaling applied to data
DATAMIN = 0 / Minimum data value
DATAMAX = 554 / Maximum data value
CBLACK = 0 / Indicates the black point used when displaying the image.
CWHITE = 1 / indicates the white point used when displaying the image.
END

han.k
Posts: 316
Joined: 23 Nov 2017, 15:04
Location: The Netherlands
Contact:

Re: Image with range 0 1

Post by han.k » 18 Dec 2019, 09:24

The 8 bit image is strange. Normally it should be a 16 bit image. A DSLR is at least 12 bit so 8 bit is compromising the quality. I can see it is written by INDI (Ekos?) Could you provide me with one (or more) images created by Ekos for testing. Upload it somewhere and give me the link.

Clear skies, Han

han.k
Posts: 316
Joined: 23 Nov 2017, 15:04
Location: The Netherlands
Contact:

Re: Image with range 0 1

Post by han.k » 18 Dec 2019, 09:33

Any idea if you can set in INDI 16 bit instead of 8 bit? Camera driver?

Clear skies, Han

han.k
Posts: 316
Joined: 23 Nov 2017, 15:04
Location: The Netherlands
Contact:

Re: Image with range 0 1

Post by han.k » 18 Dec 2019, 09:45

Below a quote from the INDI forum describing how to switch to 16 bit. I haven't tested 8 bit stacking for at least 18 months since it is very very rare. It is very possible it goes wrong due to the 8 bit. So the advice is to create 16 bit images. I will do some testing with artificial 8 bit images.

Han

However, as soon as you start Ekos you get an INDI Control Panel popup. Go to the tab of your camera and then select the last tab with config parameters of your camera. You may need to expand the popup or click the little arrow button to the right of the tabs to scroll that far to the right. In that tab you should be able to select 16 Bit. Then go to the tab called Options and click the Save button in the Configuration row. That way the setting gets persisted in an XML file and Ekos will set the camera to 16 Bit in the future as well. No need to restart Ekos or KStars by the way, the setting is picked up immediately.

rbarbera
Posts: 6
Joined: 18 Dec 2019, 09:01

Re: Image with range 0 1

Post by rbarbera » 18 Dec 2019, 10:30

Thanks for the clue!. I'll change the INDI driver configuration to use 16bits. There is a "bug/feature" on INDI that rounds down the bits per pixel info, so I've entered 12 and it has been round to 8 :?

I will upload a pair of frames latter for your reference.

rbarbera
Posts: 6
Joined: 18 Dec 2019, 09:01

Re: Image with range 0 1

Post by rbarbera » 18 Dec 2019, 10:37

han.k on this link you can find a zip file with some 8 bits FITS

https://www.dropbox.com/s/nyvsjal0jrz8u ... d.zip?dl=0

han.k
Posts: 316
Joined: 23 Nov 2017, 15:04
Location: The Netherlands
Contact:

Re: Image with range 0 1

Post by han.k » 18 Dec 2019, 15:55

I have the images and can reproduce the problem. Will try to fix it.

I noted that images are already in colour= demosaiced. For a better star colour you better leave them raw next time.

12 bit FITS files don't exist. It is either 8, 16, 32 or 64 bit. So for 12 bit DSLR image use 16 bit file format.

han.k
Posts: 316
Joined: 23 Nov 2017, 15:04
Location: The Netherlands
Contact:

Re: Image with range 0 1

Post by han.k » 18 Dec 2019, 17:37

It is fixed in the Windows and Linux version. Are you running MacOS?


After stacking go to pixelmath1 and use auto correct colour. Press Auto, Press apply. Next time try let the de-mosaic be done in ASTAP. There this colour correction can be set automatically in tab STACK METHOD, section Raw one shot colour images. Select there the de-mosaic method "Astro-C, colour for saturated stars. Range 0-4095 (12 bit)" (corrected 19-12-2019)

Han

rbarbera
Posts: 6
Joined: 18 Dec 2019, 09:01

Re: Image with range 0 1

Post by rbarbera » 18 Dec 2019, 20:21

Thank you Han for for quick response and support. Don’t worry about those images, I’m just testing and leaning to use the equipment.

Any way, I’m very glad to know that you are so supportive with us. Thank you very much.

King regards.

han.k
Posts: 316
Joined: 23 Nov 2017, 15:04
Location: The Netherlands
Contact:

Re: Image with range 0 1

Post by han.k » 19 Dec 2019, 07:21

I was a good exercise and allowed me to detect a bug in version 0.9.306 (wrong file format average stacking) due to a deleted code line. So get version 0.9.397.

For de-mosaic algorithm, the advice is method "Astro-C, colour for saturated stars. Range 0-4095 (12 bit)". I was incorrect with 0 -16383 (14 bit)

Han

Post Reply