Shooting Calibration Frames

Sell Your Photos And Videos Online

How to make money selling your photos online

Get Instant Access

Calibration plays an integral role in making top-notch CCD images; but in the excitement of imaging faint nebulae, changing filters, or battling intermittent clouds, it is tempting to skimp on shooting calibration frames. Unfortunately, skimping on calibration steals away the very qualities you want most in your celestial imagery. We discussed the basic idea behind accurate photon counting in Chapter 2, and cover calibration in full detail in Chapter 6. In this section, we focus on the practical details of making high-quality calibration frames.

5.5.1 Basic, Standard, and Advanced Calibration

Calibration comes in three "flavors;" basic, standard, and advanced. Each flavor is a method of performing calibration according to a standard technique, or protocol. Select the technique most appropriate for your imaging program and follow it each night that you make images.

Basic calibration. This technique meets the needs of beginners and observers who are searching for supernovae, comets, or asteroids. It is simple: before or during an observing session, make between five and sixteen dark frames using the same exposure time you use for your raw images. These dark frames are combined into a master dark frame which is subtracted from the raw images.

Standard calibration. The standard calibration technique produces high-quality monochrome or color images suitable for display or publication, or for precise astrometry or photometry. In addition to shooting a set of dark frames with the same exposure as the raw images, you make raw flat frames and flat darks. The total investment of time is ten to fifteen minutes. The darks are combined to make a master dark frame, and the raw flats and flat darks are combined to make a master flat. The master dark frame is subtracted from each raw image, and the result is divided by the master flat frame.

Advanced Calibration. Advanced calibration is used to produce high quality images. It is the most flexible protocol, but also the most involved. In addition to making dark frames (with exposure times longer than the raw images), raw flats, and dark flats, you must make bias frames. The bias frames allow you to adjust the master dark frame so that dark current can be removed from an exposure of any length.

• Tip: AIP4Win automates all three calibration protocols, and provides tools for user-controlled calibration procedures as well. Newcomers to CCD imaging may wish to use the basic calibration protocol until they feel ready for standard or advanced calibrations.

5.5.2 Making Bias Frames

Bias frames are images with zero exposure time and no light. Making them is simple: cap the telescope, close the shutter, and make the shortest integrations your camera allows. Make and save five or more bias frames. During the calibration, you can combine these either by averaging them or by taking the median value to obtain a master bias frame.

Some CCD cameras automatically correct for changes in the bias level by clocking out extra lines after the CCD image has been read, to obtain an average bias. They then correct the image for any deviation from a fixed bias pixel value, such as 100 ADUs. If your camera has this feature, use the fixed bias value instead of a master bias frame.

• Tip: In AIP4Win's advanced calibration, you have the option of creat ing a master bias frame or using a fixed bias value.

Because they have little or no dark current and no photon-induced signal, bias frames are uniform, recording principally the random noise contribution from the sensor's on-chip amplifier. However, because they are so very bland, bias frames reveal noise that might otherwise escape detection. Horizontal or vertical stripes, herringbone patterns, wavy lines, or random bright or dark pixels are signs that noise is creeping into the images.

5.5.3 Making Dark Frames

High-quality dark frames play a key role in calibration not only because they remove dark current, but also because they place the zero-point of the numerical values in your image where it should be located, at zero. This is necessary for accurate photometry, precise astrometry, and true-color imaging. Furthermore, flat-fielding works well only with properly dark-subtracted images.

Dark frames have two components: a zero-point bias and a thermal signal. The thermal signal (also called dark current) accumulates at a rate that depends on the temperature of the image sensor and the duration of the integration. The bias should be constant, but the thermal signal grows linearly with integration time.

Making dark frames is easy: you cap the telescope and (if it has one) you close the shutter of your camera. No light should reach the sensor. You then make an exposure and save the resulting image. If you immediately make another integration of the same length and carefully compare the two, you will find that they differ slightly. The small difference is random noise. By making multiple dark frames and averaging them, you reinforce the constant thermal signal and at the same time average away the random variations.

If you are using the basic and standard calibration protocols, make five or more dark frames using the same integration time you are using for your images. If you are making 60-second and 120-second integrations, you should make two sets of dark frames with matching integration times.

For the advanced calibration protocol, you can make one set of dark frames with an integration time that is longer than your longest image integration. For example, if you are making 30-second, 60-second, and 120-second raw images, you can make five or more 300-second dark frames. In the advanced calibration protocol, you also make a set of bias frames, which can be subtracted from the darks to obtain a direct measurement of the thermal signal. This can then be scaled to match the thermal signal that accumulated in the 30-, 60-, and 120-second integrations. Flexibility in exposure times is a strong incentive to adopt the advanced calibration protocol for all of your imaging.

• Tip: AIP4Win supports basic, standard, and advanced calibration techniques, as well as being able to automatically match a dark-frame of one exposure time to an image with a different exposure time—a significant convenience for observers.

5.5.4 Making Flat-Field Frames

Flat-field frames map the relationship between the intensity of light coming from the sky and the response of each photosite on the sensor. To make a flat frame, all you need to do is shoot an image of a featureless subject. If the light is uniformly distributed across the telescope aperture and spread uniformly in angle, the resulting image records not only how much light came through the optics and struck each photosite, but also how each photosite responded to the light falling on it.

Later, in software, the pixel values in the flat field are used to compensate for light lost in the optics and for the variations in the pixel-to-pixel sensitivity of the sensor. The key to making good flats is simple: the flat must be made using exactly the same setup as used for imaging. Do not remove and replace the camera; do not tighten, loosen, clamp, unclamp, or otherwise change anything about optical system. Ideally, do not even change focus. The flat must be an exact record of how light passed through the telescope when you were making images.

Good flat fields are not difficult to shoot, but they must be done correctly or they don't work. The key is to image a uniform low-level source of light that fills half the dynamic range of the sensor in a reasonable integration time, between approximately 2 and 20 seconds. The four basic options are twilight flats, sky flats, dome flats, and light-box flats. Twilight flats are images taken of the twilight sky some 30 to 45 minutes after sunset or before sunrise. Sky flats are made from images of the night sky itself. Dome flats are images taken of a large white screen attached to the inside of the observatory dome. Finally, light-box flats are images of a diffusing screen in a lightweight, internally-lit box placed over the end of the telescope. Of the four options, light-box flats are by far the best bet for amateur astronomers.

Light-Box Flats. An illuminated box is the most reliable way for amateurs

Figure 5.4 This simple light box for making flat-field frames slips over the front end of the telescope. When the internal lamps are on, the telescope "sees" a uniformly illuminated field of view. The circle is made of milk plastic, available from plastics suppliers and firms that sell outdoor advertising signs.

Figure 5.4 This simple light box for making flat-field frames slips over the front end of the telescope. When the internal lamps are on, the telescope "sees" a uniformly illuminated field of view. The circle is made of milk plastic, available from plastics suppliers and firms that sell outdoor advertising signs.

to make flats. It provides a uniform source that can be put on the telescope quickly and enables Hats to be made easily with repeatable results.

The key component of a light box is a diffusing sheet of milk plastic that fits right against the front of the telescope. (Milk plastic is pure white material used in internally-lit signs and theater marquees; you can find a local supplier under "Plastic: Rods, Sheets, and Rolls.") Illuminated from behind, a piece of milk plastic diffuser mimics a uniform section of sky.

In a typical home-built light box, the plastic is mounted directly on the end of the telescope. Over the milk plastic is a cube of foam core board or light plywood painted white inside. In the corners of the cube next to the telescope are four lamps. Baffles prevent them from shining directly on the milk plastic, so their light bounces around the interior of the box. By the time the light reaches the back side of the milk plastic, it is extremely uniform.

For a 6- to 8-inch telescope, the box should be about 12 inches on a side and 15 inches high, and proportionately larger for larger telescopes. You can build it from any material, but functor board (available from art-supply and craft stores) works well, because it is both lightweight and white. (If the light box will be handled roughly, use plywood instead.)

To flat-field lunar and planetary images where you're using a long focal ra-

Figure 5.5 Four small halogen lamps send light into the box, where it is reflected by the white interior walls. After several reflections, the distribution of light reaching the milk plastic sheet from the box is almost perfectly uniform. Small shields prevent the lamps from illuminating the milk plastic directly.

Figure 5.5 Four small halogen lamps send light into the box, where it is reflected by the white interior walls. After several reflections, the distribution of light reaching the milk plastic sheet from the box is almost perfectly uniform. Small shields prevent the lamps from illuminating the milk plastic directly.

do, you will need a bright light box. Low-wattage halogen lamps operating on 120-volt house current provide ample light. However, for flat-fielding deep-sky images, you will need a considerably less bright light box. Small tungsten-halogen lamps (such as those made for mini-Maglights) can be run from a lantern battery and work well. Adjust the placement of the lamps and design of the box so that an exposure of 2 to 10 seconds is required to produce a half-full-well signal.

To make raw flats for a master flat-field frame, wait until you have everything focused and working smoothly. Place the box on the front of the telescope, turn on the lamps, and make test exposures to find the integration time that reaches half-full-well. Make enough raw flats to reduce noise to a negligible level when the flats are averaged—ten or more virtually guarantees a good result. Then without changing anything, using the same integration time, turn off the lamps and make the same number of flat dark frames. For example, you might standardize on 16 raw flats with the lamps turned on and 16 flat-darks with them off. For color images, you should (ideally) make a separate set of flats for each color filter—but in practice, you may find that "white light" flats are adequate.

Master flats made using the techniques described above easily attain a signal-to-noise ratio of 1000 or better, so you can be sure the flat field will add negligible noise to the final image.

• Tip: AIP4Win's automated calibration functions do the work of con verting your flat-field data into a master flat frame. You need only to select the raw flats and the flat darks, and the rest is automatic.

5.5.5 Defect Mapping

Although modern CCDs boast remarkably high quality, manufacturers offer them in several grades. Grading is usually based on the number of point defects (singlepixel defects), cluster defects (several contiguous defective pixels), and column defects (all or part of a column is dead). Regular "hot" pixels are not considered defects, because they can be corrected by dark subtraction and flat-fielding. In large arrays, even the highest grades may have a few point defects.

However, defects can be corrected by mapping them and replacing those pixels known to be defective. The defect map can be stored in a specially coded image in which pixel values act as instructions to apply a specific action, such as replacing a point defect with the median of the surrounding pixels. Defect mapping and correction need to be done only once for a given CCD; it can then be applied after calibration to all images taken with the camera.

• Tip: Defect mapping is built into AIP4Win's calibration function. To make a defect map, make a uniform flat field and threshold it to reveal the non-responding pixels, clusters, and columns. AIP4Win can classify defects by type. After dark subtraction and flat-fielding, the defective pixels, rows, and columns are replaced with the median of nearby pixels.

Was this article helpful?

0 0
Digital Camera and Digital Photography

Digital Camera and Digital Photography

Compared to film cameras, digital cameras are easy to use, fun and extremely versatile. Every day there’s more features being designed. Whether you have the cheapest model or a high end model, digital cameras can do an endless number of things. Let’s look at how to get the most out of your digital camera.

Get My Free Ebook


  • arvi
    How important is shooting calibration frames?
    2 years ago

Post a comment