CALIBRATION OF CCD IMAGES

Calibration (or PRE-PROCESSING) of an astronomical CCD deep-sky image consists of removing the bias and thermal contribution (DARK FRAME) and dividing the resultant image by the FLAT-FIELD in order to standardize the response of each image pixel.

Calibrated = (Raw - Bias - Thermal) / Flat = (Raw - Dark) / Flat

The DARK frame should be removed firstly from the RAW image then the result should be divided by the FLAT-FIELD.

DARK FRAME

A dark frame is an integration in which no light strikes the CCD. It records the BIAS noise and THERMAL noise for a specific CCD temperature and integration time:

DARK FRAME = BIAS FRAME + THERMAL FRAME

To take a dark frame you can simply cap your telescope and integrate using the same time that you are using for acquiring the raw image. If you are not able to control the chip temperature (+/- 0.1 °C), dark frames should be taken immediately before or after taking a raw image.

For better results shoot lots of dark frames and MEDIAN SUM them.

BIAS FRAME

The bias frame is an image made with an integration of zero seconds and shutter closed. It contains the amplifier zero-point offset, the random readout noise from the amplifier and the noise from camera electronics.

CCD cameras made for scientific imaging usually include the ability to read out a bias frame but for the simple purpose of subtracting THERMAL FRAME and BIAS, a DARK FRAME serves equally well (DARK FRAME = BIAS + THERMAL FRAME). The examination of a bias frame tells you if your camera is working properly: if you see wavy lines or patterns, your camera may not be functioning well.

FLAT-FIELD

The FLAT-FIELD frame is a photosite-by-photosite map of a CCD's sensitivity to light. It is an image of a uniform object such as twilight sky or a sheet of opal glass attached to the inside of the observatory dome. Chip sensitivity, vignetting and dust all appear as variations in the sensitivity of the CCD itself: division by FLAT-FIELD will remove these defects.

When you make a raw FLAT-FIELD, then you must subtract the THERMAL and the BIAS frame from it (or the DARK). For the lowest possible noise, MEDIAN SUM more then one FLAT-FIELD to obtain a Master Flat-Field.

Three different methods of calibration can be used:

  • Basic calibration - requires a master dark with the same integration time as the raw image;
  • Standard calibration – requires a master dark that has the same integration time as the raw image and a master flat;
  • Advanced calibration – requires a master flat, a master dark (same or longer integration time) and a master bias.

Basic calibration is adequate for simple observing tasks. Standard calibration corrects for vignetting and CCD nonuniformity, but constrains you to the same integration timr for your images and darks. Advanced calibration gives you the freedom to use different integration times for your images and provides superior dark subtraction via dark-frame matching.

Basic Image Processing (Noise)
Image Calibration/Reduction
Image Calibration - Flat-Field

H-ALPHA AND LUMINANCE PROCESSING FLOW

Calibrate, align and combine CCD images (Maxim DL)

  • Calibrate raw images

  • Do any de-blooming repair prior to alignment and combination (e.g. New Astro DeBloomer for Maxim DL)

  • Use manual or automatic alignment operations

  • Use SDMask for all combination operations

  • Perform a DDP-style stretch on the H-alpha combined image (Don't do any sharpening as part of DDP. This is accomplished by selecting the "user" kernel filter, and setting the filter coefficients to 1.0 in the center, and zeros elsewhere)

  • Save the image (FITS format 16-bit)

    • Final Processing (Photoshop CS2, CS3)

  • Load the FITS DDP-stretched image into Photoshop (using Fits Liberator)

  • Use Shadow & Highlight (50:50:10, 50:50:10) or Levelizer (Starizona Photoshop Plugin Package)

  • Use Curves to ajust contrast and brightness

  • Use (if necessary) Neat Image Photoshop plug-in to remove grain

  • Save the final 16-bit TIF file

  • Convert to 8-bit and save JPG file

    • LLRGB PROCESSING FLOW

    Calibrate, align and combine CCD images (Maxim DL)

  • Calibrate raw images

  • Do any de-blooming repair prior to alignment and combination

  • Combine images for each channel (L, R, G, and B) (use SDMask for all combination operations, use manual or automatic alignment operations)

  • Perform color balancing for each color channel (e.g. SBIG ST-10XE RGB= 1.0, 1.1, 1.9)

  • Combine the images into an LRGB image in Maxim DL

  • Perform a DDP stretch on the Luminance image

  • Perform a DDP stretch on the LRGB image

  • Boost saturation after DDP if needed (save in 16-bit TIFF format)

    • Final Processing (Photoshop CS2, CS3)

  • Load the Luminance FITS DDP-stretched image into Photoshop (using Fits Liberator)

  • Use Shadow & Highlight (50:50:10, 50:50:10) or Levelizer (Starizona Photoshop Plugin Package)

  • Use Curves to adjust contrast and brightness

  • Use (if necessary) Neat Image Photoshop plug-in to remove grain

  • Load the LRGB 16-bit TIFF file

  • Align L and LRGB images

  • Combine L and LRGB images (LLRGB) (Change the blend mode to "luminosity", change the opacity to ~50%, Flatten image)

  • Save in TIFF or JPG format

    		•