Calibration with BIAS Frames. Calibration with DARK Frames. Calibration with Master Frames - MaximDL. Calibration with Flat Frames

Transcription

1 Calibration with BIAS Frames A BIAS frame is an image taken with your camera's shortest possible exposure time with the front of the telescope sealed closed to exclude all light. BIAS frames are used to compensate for pixel-specific offsets in the sensor chip and camera electronics during the readout of the data and its subsequent saving to the memory card or hard drive. It is normal to take at least 20 BIAS frames but because of their short exposure length, these won't take that long and you can store them for use later with images from subsequent sessions. These frames should be averaged together in your stacking software to produce a 'master' image representing your 'average BIAS'. This master BIAS is subtracted from your LIGHTS in a process known rather imaginatively as 'BIAS Frame Subtraction'. In addition to normal readout errors, some sensors can have the occasional faulty column of pixels (sometimes caused by a cosmic ray hit) but a BIAS frame will mask this sort of error very effectively. BIAS data tends not to change much over reasonable periods of time so you can produce a library of BIAS frames for use in subsequent imaging sessions but the sensor characteristics will change with age. I normally refresh mine every 2 months or so but as they are so quick to take, it could be argued that taking them at the end of each session is no real hardship! Calibration with DARK Frames A DARK frame is an image of the same exposure length as the LIGHT frames and like the BIAS frames, they are taken with the front of the telescope sealed closed to exclude all light. DARK frames contain something very useful indeed - a sample of the 'noise' generated purely by the sensor chip and its attendant amplifier. Additionally any 'hot' or 'cold' pixels will also be present in this data. In a similar manner to the BIAS frame, you should take several of these DARK frames and average them together in your stacking software to produce a master DARK frame representing your 'average noise'. Subtracting this data from your LIGHTS will remove a lot of the noise component from them leaving you with a much 'cleaner' image. This process is known as 'Dark Frame Subtraction'. If you have a set point cooled imaging camera, you can produce a library of DARKS at various exposure lengths for use in subsequent imaging sessions. If your camera does not have set point cooling or you have a DSLR which has no cooling at all then it will be necessary to take your DARKS at the operating temperature of your camera. It is normally most convenient to take these at the end of your imaging session and ideally only use the DARKS for this particular session. Many modern cooled CCD cameras produce such an insignificant amount of thermal noise that DARKS just aren't required at all but uncooled CCD cameras and DSLR cameras will benefit greatly from the use of suitable DARKS. Calibration with Flat Frames There are several methods for taking FLAT frames ranging from illuminated LCD panels, electro-luminescent panels and light boxes at the sophisticated end to imaging a cloudy sky through a white 'T' shirt at the low-tech end. In the past, I have used a method known as 'twilight flats' where the sky at twilight is used as a diffuse source of light. Sudden changes in light levels should be avoided, however, the illumination at this time of day can be pretty uniform but care must be taken to ensure that no light from stars is captured. To avoid this eventuality the telescope can be moved a small amount between exposures so that a slightly different area of the sky is imaged in each frame. More recently, I have started using an electro-luminescent panel and have found that I get very repeatable results using this method. Taking FLATS is not a trivial task as the camera must not move in either focus or orientation between the taking of the LIGHTS and the FLATS. Clearly this can be a real inconvenience as it essentially requires the FLATS to be taken immediately at the end of an imaging session. If you are taking twilight FLATS the timing will all be wrong (although you could stay up all night and take 'dawn flats'!!) but you must not allow the focus or camera orientation to change before you have taken them. Because of the additional work required to get FLAT frames right, many imagers (myself included!) often fail to take them. This is a bad mistake as the additional hassle will result in better quality images with much reduced vignetting and no unsightly dust doughnuts and your post-processing will be easier with fewer gradients to worry about. With regard to the problem of dust, you should still do everything you can to exclude any dust from the inside of your equipment in the first place! I was horrified when I examined the first set of FLAT frames that I took with my Electro-luminescent panel as I am fastidious about keeping my optics, filters and sensor face dust free but I had loads of 'dust bunnies' on my FLATS and if they appear in the FLATS, they will be present in your LIGHTS too and they contribute nothing positive to your images. Because, of the requirement of using exactly the same orientation and focus position you cannot produce a library of FLATS for use in subsequent imaging sessions. Calibration with Master Frames - MaximDL Having captured your BIAS, DARK and FLAT calibration frames, you should produce 'masters' of each type. MaximDL automates this process for you so start by placing all your calibration frames in the same directory. Select Process - Set Calibration and browse to your calibration folder using the 'Source Folder' section at the foot of the dialogue box and then click on 'Auto-generate'. This will group all the 'like' frames into the top section of the dialogue box. Now select the BIAS frames and choose 'Average' as the combine mode. Repeat this for the DARK frames. Now select the FLAT frames but this time choose 'Median' as the combine method and if you are using a one shot colour camera, enable the 'Apply Boxcar Filter' box. Finally, click on the 'Replace W / Masters Button' to generate the master calibration files. This process may take a couple of minutes to complete. If you are using a one shot colour camera like a DSLR or one of the many popular colour astro-ccd cameras, you MUST calibrate your LIGHT frames before your convert them to colour. When you are ready to calibrate your LIGHT frames, return to the Set Calibration dialogue box as detailed above and enable the calibration features that you want to implement by selecting them in the top row of the dialogue box. I always choose to BIAS subtract my FLATS as well as applying the BIAS and FLATS to my LIGHTS.

2 RGB Colour Conversion - MaximDL If you are using a one shot colour (OSC) camera like a DSLR or one of the many popular colour astro-ccd cameras, to release the colour in your images, you must first 'de-bayer' the data. Colour CCD cameras have mono sensors but there is a layer of red, green and blue filters sublimated on top of the sensor. When the data is read from the sensor, a special algorithm encodes this data into what is known as a 'Bayer Matrix'. De-Bayering decodes this data into the Red, Green and Blue channels. There are several parameters that need to be set to allow this to be carried out successfully. Using an image of a deep sky object makes it difficult to determine the correct settings but there is a simple way to do this. Set up your telescope and camera during the day and if you are intending to use a light pollution filter or an Infra Red filter, insert this in the light path. Now take a set of images using 'focus' mode by selecting View - CCD Control Window - Focus. With 'Screen Stretch' set to Moon, start the sequence by selecting a fast exposure setting (less than 0.1), binning of 1 x 1 and enable 'Continuous'. Now adjust focus until you have a colourful daytime object in focus (although it will be displayed in mono at this stage) and then stop the imaging sequence. Select Colour - Convert RGB and set 'Scaling' to 100% in each of the three (Red, Green and Blue) boxes as an obvious start point. Now enable the X and Y offset boxes in each permutation (both Off, X On - Y Off, Y On - X Off and both On) until you get sensible colours. Finally adjust the scaling percentages to remove any colour cast to the image. This daytime image taken through your normal imaging filter will ensure that your night sky images are at least sensibly colour balanced as a start point and correctly de-bayered before you proceed to process your images any further. If you find that your images generally have a colour BIAS towards one colour, reduce this colour's percentage value until you get the desired result. Stacking Colour Images - MaximDL Once you have calibrated and colour converted your images, you can proceed to stack them. Although you can carry out the stacking process in 'memory' by loading all the images first, this is not the recommended method and can result in quickly running out of memory space! Select File - Combine Files and select the images that you want to stack using the Combine Files dialogue box. Use standard Windows procedures to select as many images as you want and then click on the 'Combine' button. There are all sorts of methods for combining files but I much prefer the 'Manual 1 Star' method as this allows you to examine each image and either include it or reject it from the stack. Ensure that 'Use Centroid' and 'Auto Next' are enabled then right click on the image and select 'Crosshair'. This will place a crosshair bisecting the centre of the frame. Choose a suitable star near the crosshair to use as your alignment star and click on its centre. You will immediately be presented with the next frame so select the same star again. Continue this process, rejecting any poorly exposed or star trailed frames until all the frames have been selected and the system 'pings' to advise that it has completed the selection process. You must now choose the actual combining method and my preference here would be 'SD Mask'. Select SD Mask and then click on the 'Settings' button and choose Number of Passes = 1, Sigma Factor = 1.5, Normalization = Linear, Normalization Area = 60%, disable Ignore Pixels over and enable Ignore Black Pixels and click on OK. Change FITS format to IEEE Float and finally click on OK to start the stacking process. This could take quite some time if you have a large sensor and lots of light frames! When the stacking process has completed, save the new FITS file so that you can return to it at any time in the future if necessary. To prepare the stacked image for export to PhotoShop, adjust the left hand slider in the screen stretch dialogue box fully to the left, click on the + magnifying glass and adjust the left hand slider until it is just to the left of the start of the histogram curve. Now adjust the right hand slider to the left until a reasonable image appears. Finally save the image as a 16 bit TIFF file after selecting Stretch and tick the following boxes - Linear Only + Screen Stretch + 16 bit. Stacking/Aligning Mono Images - MaximDL Once you have calibrated your images, you can proceed to stack them in their individual filter groups. Select File - Combine Files and select the images that you want to stack using the Combine Files dialogue box. Use standard Windows procedures to select as many images as you want and then click on the 'Combine' button. There are all sorts of methods for combining files but I much prefer the 'Manual 1 Star' method as this allows you to examine each image and either include it or reject it from the stack. Ensure that 'Use Centroid' and 'Auto Next' are enabled then right click on the image and select 'Crosshair'. This will place a crosshair bisecting the centre of the frame. Choose a suitable star near the crosshair to use as your alignment star and click on its centre. You will immediately be presented with the next frame so select the same star again. Continue this process, rejecting any poorly exposed or star trailed frames until all the frames have been selected and the system 'pings' to advise that it has completed the selection process. You must now choose the actual combining method and my preference here would be 'SD Mask'. Select SD Mask and then click on the 'Settings' button and choose Number of Passes = 1, Sigma Factor = 1.5, Normalization = Linear, Normalization Area = 60%, disable Ignore Pixels over and enable Ignore Black Pixels and click on OK. Change FITS format to IEEE Float and finally click on OK to start the stacking process. This could take quite some time if you have a large sensor and lots of light frames! When the stacking process has completed and you have carried out the procedure for each of the filter types, save each of the new FITS files so that you can return to them at any time in the future if necessary. To prepare the stacked images for export to PhotoShop, load in the individual stacked images and carry out an alignment on them by selecting Process - Align - Add All - OK and use a Manual 2 Star alignment choice. In the same manner as described above, select a star near the centre and then for the second star select one in a far corner. When the alignment process has completed, save the images as 16 bit TIFF files after selecting Stretch and tick the following boxes - Linear Only + Screen Stretch + 16 bit. Selective Sharpening with Layer Mask - PS Sometimes it can be very useful to just sharpen certain parts of an image and using a 'layer mask' is great way to do this. Start by producing a duplicate layer by selecting Layer - Duplicate Layer and name this layer 'Sharp'. Select this new layer from the Layers palette and set the blending mode to 'Overlay'. The image will take on a high contrast appearance but this is normal. Select Filter - Other - High Pass and adjust the pixel radius to achieve the amount of sharpening that gives the result you want and click 'OK' (somewhere between 2 and 4 pixels normally works for me). This will result in the whole image being sharpened which may be desirable but the next stage will allow you to be very selective over which parts of the imaging will have the sharpening applied to them. The next part of the process is to produce a layer mask by selecting Layer - Layer Mask - Hide Selection. It will appear as though the sharpening has gone away but it is still there hidden in the mask. Double click on the 'set foreground' colour square in the Tools palette and set all the CMYK settings to 0% to select white as the foreground colour. Finally choose a suitable brush size by clicking on the brush tool in the Tools palette and set mode to normal, opacity to 100% and flow to 100%. Now gently brush over the areas that you want to show through as sharp.

3 Sharpening using 'overlay' mode - PS Sharpening of deep sky objects is something that should be tackled with great care. Whereas lunar and planetary images often respond very well to sharpening, deep sky objects don't always fare so well. Simply using the unsharp mask filter in PhotoShop can result in unwanted image artefacts like star halos. However, the following process does work well with some deep sky objects, especially if the sharpened version is overlaid on top of the original image and the opacity adjusted until the best balance between sharpness and the original image is achieved. The process is as follows:- Open the image, use 'save as' to save the image under a different name incorporating the word 'sharp' and then select the original image again. This procedure will give you two versions of the same image on screen at the same time. With the original image selected, produce a duplicate layer and name this layer 'Sharp'. Select this new layer from the Layers palette and set the blending mode to 'Overlay'. The image will take on a high contrast appearance but this is normal. Select Filter - Other - High Pass and adjust the pixel radius to achieve the amount of sharpening that gives the result you want and click 'OK'. Flatten this image using Layer - Flatten Image, select the flattened image, copy it to the clipboard and paste it on top of the image now named with the word 'sharp'. Adjust the opacity until you are happy with the blend between the two layers. Removal of a 'noisy' background - PS The secret to smooth images is taking lots of subframes but often this is not possible so image smoothness will be missing. There is a partial solution for this and the key is to blend in a smoothed duplicate of your image. There are several ways of producing a smoothed duplicate; slightly blur a copy of the original image or use an external noise reduction program specifically designed to carry out the task of noise reduction. I use two different programs for this purpose, one called 'Neat Image' and the other called 'Imagenomic Noiseware'. There is also an excellent PhotoShop plugin called Noise Ninja produced by Picturecode that does a similar job. Start by saving a copy of the image with a different file name in TIF format. Load the copy image into the noise reduction software of your choice and let it automatically process your image using the default settings. Once the process has been carried out, you can click on the image using either program to see a 'before and after' comparison. However, if the reduction has been overdone using the default settings, try adjusting the 'Luminance Channel' slider control in the 'Noise Reduction Amounts' section of Neat Image or the 'Luminance' slider control in the 'Noise Suppression' section of Noiseware until you get the right amount of smoothing. You will know if the noise reduction has been overdone as the stars will tend to run into one another and the image will lose clarity. Once you are happy with the result, save the image with a new file name. Open this image in your image processing software, select it, copy it and then paste it on top of your original image, producing a new layer in the process. If you are using a PhotoShop plugin, start my creating a duplicate layer and run the plugin filter on this layer. Finally, from the Layers palette, select the new layer and adjust the opacity until the original background noise just becomes apparent and you will have the best of both worlds, the good detail of the main object but a much 'cleaner' background. This process can also be used to good effect to alleviate the 'noise' that appears in images that have been shot through dampness in the upper atmosphere. LRGB Combine - PS You must first align your three (four if you have Luminance data) filter channels using MaximDL to allow for any changes in image orientation due to the filter changing, especially if using a manual filter wheel. Save each file as a 16 bit TIFF file. Open the three mono image TIF files (Red, Green and Blue) and convert them to Greyscale by selecting Image - Mode - Greyscale for each image. Now load the Luminance channel and convert it to RGB by selecting Image - Mode - RGB. Open the Channels palette by selecting Window - Channels. Click on the Channels palette menu and select 'Merge Channels'. Select 'RGB' from the drop down list box, set 'Channels' to '3' and click on 'OK'. Using the list box, specify which file applies to each of the three colours and click on OK. Your RGB colour image will immediately appear and should be saved as a PSD file under a new name, along the lines of 'm45_2_210309_rgb.psd'. This ensures that if any later processing proves to be faulty, you can always return to the original combined image. The Luminance channel can now be combined with the RGB channels in the following way:- Initial Stretch - PS From PhotoShop, select Layer - New Adjustment Layer - Levels and you will note the default histogram is that of 'RGB'. The left hand slider adjusts the black point and maps the pixels value to a level of 0, the right hand slider adjusts the white point and maps the pixel value to a level of 255. The remaining levels are redistributed by the software between levels 0 and 255 and this redistribution of levels increases the tonal range of the image, effectively making an increase to the overall image contrast. The centre slider is the most effective of the three as this adjusts the 'gamma' level of the image experiment by sliding this control to the left and note the amazing effect is has on your image but don't overdo it. Select the Luminance image as the active image, select the contents using Select - All and then copy the contents to the clipboard using Edit - Copy. Select the RGB image and paste the Luminance channel on top of it using Edit - Paste. Open the Layers palette using Windows - Layers and set the blending mode to 'Luminosity' using the drop down list box (the default is 'Normal'). The resultant image is the LRGB image which you should save as a PSD file under a new name, along the lines of 'm45_2_210309_lrgb.psd'.

4 Basic Light Pollution Removal - PS Load the calibrated and stacked image into PhotoShop, select Layer - New Adjustment Layer - Levels and you will note the default histogram is that of 'RGB'. Using the drop down list box, choose the RED channel. You will immediately note a change to the histogram display. Ensure that the 'preview' option is selected so that you can see the changes that you are making to the appearance of the image in real-time. Slide the left hand 'slider' to the right until it is lined up with the start of the steep 'slope' of the histogram - it is important not to go past this point or you will be clipping your data. You will note that the image has changed its hue to a greenish tinge. Now, repeat the process for the green channel, noting that the hue has now got a blue tinge to it. Finally, repeat the process for the blue channel and you will be rewarded with a much better background sky colour and a far better colour balanced image. Noel Carboni's Light Pollution Removal Tool Although the above method will do a reasonable job of removing light pollution, Noel Carboni has a tool specifically designed for this purpose and I believe that the purchase price of his actions is justified in this single action. Using the 'Curves' Tool - PS The Curves tool is a very powerful tool that takes adjustment of the 'levels' to a new height although it can be used just to set the black and white points by adjusting the top and bottom of the 'curve' However, its real power comes from the flexibility it empowers you with in adjustment of the tonal range of an image. This flexibility comes from its ability to add additional control points to the 'curve'. You can add these additional control points by simply clicking at selected points on the curve. Once these have been added they can each be repositioned anywhere within the grid producing a range of different effects by altering the contrast and colour saturation at various points in the curve. From PhotoShop, select Layer - NewAdjustment Layer - Curves. One of the most popular and useful curve shapes is what is typically referred to as an 'S' curve. This shape is produced by adding two control points to the original 'curve' and pulling the top point up and to the left and the bottom point down and to the right forming an approximately 'S' shaped curve. The upper control point lightens the lighter parts of the image while the lower control point darkens the more shadowy areas. It should be noted that the 'steeper' the curve, the greater will be the contrast increase. A by-product of the increase in contrast is an increase in saturation levels. Now this may well be a desirable effect in your particular image but if it doesn't achieve the effect you require then it will have to be corrected, perhaps by using the saturation control tools. However, there is another way in which this saturation change can be controlled and I would recommend that you experiment with changing the 'blending mode' of the adjustment layer from 'Normal' to 'Luminosity' in PhotoShop. With the blending mode set like this, the effects will be limited to the luminosity of the image which can produce a very desirable result. Adjusting Colour Balance - PS Depending on your sky conditions, the choice of object, the individual exposures used for each filter (if you used a mono astro-ccd camera) and the settings used during de-bayering (if you used a DSLR camera or one shot astro-ccd camera) you could end up with an image that has an unwanted colour cast to it. This can be adjusted quite easily by producing a second adjustment layer, and working on that. From PhotoShop, select Layer - New Adjustment Layer - Colour Balance. 'Preserve luminosity' should be selected. With this tool, the changes you make to one colour will have an effect on another and you will note that the three sliders each pertain to two colours - Cyan/Red, Magenta/Green and Yellow/Blue. Adjusting the slider away from the Red will reduce the Red component and increase the Cyan component and so on with the other two pairs. Make careful, small adjustments to the required colour pair(s) experimenting with the shadows, midtones and highlights selections until you have got the colour balance right. Once the overall colour balance is to your liking you can move on to adjusting the saturation levels. Saturation Control and Boost - PS It is not uncommon to find that your images lack the colour depth that you want or that previous processing procedures have boosted the colour unnecessarily but there are simple tools to assist you with correction of these problems. From PhotoShop, select Layer - New Adjustment Layer - Hue/Saturation (the default setting is 'Master'). Adjusting the 'Saturation' slider to the right will increase the saturation whereas moving it to the left will decrease it. There is an important issue to remember here though - don't make all your adjustments in one go, rather, make a lot of much smaller adjustments. This will ensure that you have plenty of control and that you get a smooth, natural appearance to the finished image. Adjusting the saturation in this manner is the most basic method and there is the risk of producing unwanted artefacts especially if a lot of adjustment is required to release the colour. Luckily, there are other, more sophisticated methods that can be used to good effect. See 'False Luminance Channel Colour Boost', 'Match Colour Saturation Boost', 'Soft Light and Luminosity Colour Boost' and the 'LAB Colour Boost'.

5 False Luminance Colour Boost - PS If you have a DSLR or one shot colour astro-ccd camera it would be useful to take advantage of a Luminance channel to enhance your RGB data. With a suitable Hydrogen Alpha object like an emission nebula, you could take a series of subframes using an Ha filter to use as your Luminance channel. However, if your object is not suitable for this sort of filtering then producing a false Luminance channel can give you some additional flexibility in post processing, especially if you are aiming for an increase in saturation levels. A false Luminance channel and its combination with the RGB channels can be carried out as follows:- 'Match Colour' Saturation Boost - PS This process is particularly effective on the red colours of emission nebulae but be careful not to overdo it! Start by creating a duplicate layer. From PhotoShop, select Layer - Duplicate Layer, set the name to 'match' and change the blend mode to 'Colour'. Now select Image - Adjustments - Match Colour and increase the Colour Intensity to suit your taste. Later, you can adjust the intensity of the new layer by reducing its opacity. From PhotoShop, open your RGB image and Select Layer - Duplicate Layer Type in a new name like 'Desat' and click 'OK'. Desaturate the duplicate layer by selecting the layer in the Layers palette then select Image - Adjustments - Desaturate. Finally, set the blending mode to Luminosity using the drop down list box in the Layers palette. You can now select the background RGB layer and adjust the saturation and other parameters in the knowledge that the fine detail of the image is being safely retained in the Luminance channel. Once you have made changes to the RGB layer, it is often beneficial to blur this layer a little using the Gaussian Blur filter. From PhotoShop, select Filters - Blur - Gaussian Blur and select a pixel radius of between 1 and 3. Remember, the crispness and detail of the image are held in the Luminance channel so blurring the RGB layer will not destroy detail, however, some over-processing of the RGB layer can lead to unpleasant edge effects and other artefacts but the blurring of this layer can alleviate these. It is well worth experimenting with this technique using a range of alterations to the RGB layer and making very small changes to the sharpness of the Luminance as well as this can also be very effective. I shouldn't mention this here in an article about deep sky imaging but this technique is incredibly effective at producing highly saturated images of the Moon! Soft Light/Luminosity Colour Boost - PS Start by creating a duplicate layer. From PhotoShop, select Layer - Duplicate Layer, set the name to 'soft' and change the blend mode to 'Soft Light'. Duplicate this new layer again, setting the name to 'lum' and change the blend mode to 'Luminosity'. You will immediately notice more vibrant colours appearing without the introduction of any artefacts. Select 'History', right click on the last entry and create a new document. Flatten all three layers, by selecting Layer - Flatten Image, select and copy the entire image and then delete this new document as it is no longer required. This should bring you back to the original image after the first bit of colour enhancement. Now paste the copied data onto this image creating a new layer and delete the two layers called 'soft' and 'lum'. If the colours are too extreme for your taste, change the opacity of the top layer to get the effect that you want. If you want to further increase the saturation, you can repeat this process several times but I would recommend that you carefully stretch the image between each enhancement. LAB Colour Boost - PS Unlike RGB colours that are screen-dependent and CMYK colours that vary with printer, ink and paper characteristics, Lab colours are device-independent. This means that the visual characteristics of these colours remain consistent on monitors and printers. What is of interest to us though is that Lab space is capable of delivering superb improvements in contrast and saturation. The Lab mode comprises of three colour channels. The first channel is the L (Lightness) channel also known as luminance and this has a value range from 0 to 100. A Lightness value of 0 equates to black and a value of 100 equates to white - the higher the value, the more vivid the colours. The other two channels, a and b, represent the actual colour ranges. The a channel contains colours ranging from green to red and the b channel contains colours ranging from blue to yellow. Working in Lab mode is straightforward. Select your current image, copy it and create a new document, pasting the data into it. Flatten the image by selecting Layer - Flatten Image. Now change the mode from RGB to Lab by selecting Image - Mode - Lab Colour. To adjust reds and greens, go to channels and select the a channel. To adjust this channel select Image - Adjustments - Brightness/Contrast. Increasing the contrast will boost reds, decreasing the contrast will boost green. Select all the channels again by clicking on Lab and check the result against the original image. To adjust blues and yellows, go to channels and select the b channel. To adjust this channel select Image - Adjustments - Brightness/Contrast. Increasing the contrast will boost yellows, decreasing the contrast will boost blues. Select all the channels again by clicking on Lab and check the result against the original image. Selecting the L channel will alter the Lightness channel and increasing the contrast will weaken colours, decreasing the contrast will strengthen colours.

6 Hot Pixel and Blemish Removal - PS Despite your best attempts at errant pixel control, some bad (hot or cold) pixels can still appear in your processed images and if these are intrusive at the resolution you want to view your final image, correction will be required. Most image processing software will have a very useful tool called the 'clone' tool which is identified by a 'rubber stamp' icon. By selecting the correct size of brush and selecting a source point ('Alt' and click in PhotoShop) a brush sized portion of the image can be duplicated by clicking anywhere on the image. This feature will allow you to paint over the hot pixel or other blemish with a piece of nearby clean sky. In PhotoShop, once you have made the first 'destination' click, the source and destination points move together as the mouse is moved around the image and clicking the mouse will continue to copy the area under the source crosshair onto the area you are clicking on. The most effective way of cloning 'invisibly' is to use a fairly small brush and to change the source position relative to the destination position regularly so that no nearby pattern is reproduced over the area you are replacing. Star Rounding - PS Use the Magic Wand tool to select a poorly shaped star choosing a tolerance value that just encloses the visible region of the star (I find 120 works well). Select Select - Modify - Smooth and choose a sample radius of 2-3 pixels and expand the selection by 5 by selecting Select - Modify - Expand. Now select Select - Modify - Feather and change the feather radius to 3 pixels. Select Filter - Blur - Radial Blur and adjust the parameters as follows:- Amount = 90 Method = Spin Quality = Best Press OK and the star will take on a better round shape. Red Star Rim Removal - PS Use the Magic Wand tool to select a red rimmed star choosing a tolerance value that just encloses the white visible region of the star (I find 120 works well). Select Select - Modify - Expand and expand the selection by as many pixels as required to completely encompass the red rim I usually find that between 2 and 4 pixels is sufficient on an APS sized image. Select Image - Adjustments - Hue/Saturation and deselect 'Colourize'. Finally, reduce the Saturation level slider until the red rim just disappears and press OK. Reducing Star Size - PS Sometimes a particularly bright star can spoil an image as it can overshadow the deep sky objects nearby. The following is a simple but effective method of reducing the size of an individual star. This process should be left to the end of the session because it only works on 8 bit images so all your stretching should have been completed by now. Copy the image and paste it into a new document. Select Image - Mode - 8 Bit to change the new image to an 8 bit image. From the Tools palette, select the Elliptical Marquee Tool and whilst holding down the shift key, draw a tight circle around the star you want to reduce. Feather the selection by selecting Select - Feather and change the feather radius to 3 pixels. Now select Filters - Distort - Pinch and choose the percentage by which you want the star to reduce and click on OK. The star will reduce in size uniformly. What is nice about this simple method is that if you have diffraction spikes on the star these will also contract uniformly keeping the star's original appearance. Do this for each star you want to alter and then copy the altered image and paste it on top of the original which will convert it back to 16 bit. By flipping between the two layers on the Layers palette you will be able to decide if you are happy with the result or not.

7 Correcting Elongated Stars - PS It is a sad fact that if you have elongated stars then your main object must also be elongated. You can improve the appearance of the stars to some extent but this will always give an artificial result and will not repair the elongation of the main object. The star 'repair' process starts by creating a duplicate layer and changing the blend mode to 'darken'. Next select Filter - Other - Offset and select 'set to wrap around'. Finally with 'preview' selected, shift the duplicate layer in the Horizontal and Vertical axes until the stars take on a good shape. A positive horizontal number will shift the duplicate layer to the right, a negative number to the left whilst a positive vertical number will shift the duplicate layer down and a negative number up. Best results are achieved by making small changes (1 to 2 pixels) and repeating the procedure again if necessary. In all cases, the stars will become dimmer. If quite a lot of offset has had to be made to get good star shapes, you may find that you have lost detail in the non star portions of the image and that you prefer the appearance of the original image. If this is the case, you can use the 'eraser tool' to reveal sections of the main object from the original image by choosing a suitable brush size and brushing over the portions that you want to reveal. Adding a Picture Frame - PS If you want to display your image to others, say on a forum, a frame with minimal details about the object and image capture data can really add the finishing touch to your work and impart useful data for others. The excellent plugin 'Astronomy Tools' from Noel Carboni has an excellent 'Astro Frame' feature that will produce the actual frame for you and all you have to do then is add the text in a contrasting colour. Adding a frame can also be done manually. Open your image select it, copy it to the clipboard, create a new image by selecting File - New and click on OK. Paste the contents of the clipboard onto the new blank image. Add some extra 'canvas' around the image by selecting Image - Canvas Size. Select 'Relative' and enter the amount of extra canvas (this is your frame width) in the Width and Height boxes in your choice of dimension (remember that the values for width and height will be halved and added to each side of the image). Select the middle of the nine squares to add canvas around all four sides of the image. From the Layers palette, select the background layer and fill the layer with a solid colour of your choice. Now add a texture such as Canvas or Burlap by selecting Filter - Texture - Texturizer. To give the frame a 3D effect so the image appears to be placed behind the frame, select Layer - Style and then tick both Outer Glow and Inner Shadow and press OK. You can annotate the frame with the object name, exposure details and equipment used by using the text tool in a contrasting colour. Star-Glow Effect - PS To enhance the appearance of some images and in particular, wide field constellation images, it can be quite effective to enlarge the key stars in an image. Start by selecting the stars that you want to make 'glow' by using the elliptical marquee tool. Hold down the 'shift' key and draw a circular marquee around each star. Keep the 'shift' key held down throughout and continue selecting all the stars that you want to work on. When you have selected all the stars, copy them to the clipboard and then paste them back on top of the original image. Next, the top layer must be blurred so select Filter - Blur - Gaussian Blur and choose a suitable pixel radius that gives the result that you want. Finally, to give a more realistic appearance, change the layer blend from 'Normal' to 'Screen'. You can now adjust the intensity of this effect by adjusting the 'opacity' of the top layer and when you are happy with the result, flatten the images by selecting Layer - Flatten Image. Crop Rotate and Flip - PS Although I have encouraged you to get the best framing of your image early on in the image capture process, things don't always work out quite as we would have hoped and it is sometimes beneficial to make a final adjustment to either the rotation or cropping of the image. Most image processing software has an option to allow you to rotate the image in either standard increments of 90 and 180 or to an arbitrary angle of your choice as well as flipping the image both horizontally and vertically if required. Make a note of the current width and height in pixels if you want to maintain exactly the same aspect ratio as the original image once you have rotated it as we will use this information shortly. From PhotoShop, select Image - Rotate Canvas - Arbitrary then enter the required angle in degrees (or decimals thereof) and select either the CW (clockwise) or CCW (counter-clockwise) radio button; finally, clicking 'OK' will carry out the rotation. If the rotation isn't quite what you hoped for, undo the rotation and try again with a new number - the reason for this is that a certain amount of interpolation is required during rotation so ideally you only want to do this once. Having rotated the image, the canvas size will increase to allow the whole image to be viewed so now is the time to crop the image to remove the blank areas produced by the rotation process. If maintaining the existing aspect ratio is not of any importance then the cropping can be carried out using 'normal style'. From PhotoShop, select the 'Rectangular Marquee Tool', from the toolbox, select 'normal style' and draw a selection box around the area you want to crop. If maintaining the existing aspect ratio is required, determine the original ratio from the pixel counts made earlier (for example an image of 600 x 480 pixels has an aspect ratio of 6 x 4.8), select 'fixed aspect ratio' and draw a selection box around the area you want to crop - you can adjust the position of the selection using the mouse. Finally select Image -Crop to crop the image to the required shape. Save the image under another file name for security.

8 Combining Smooth and Sharp (1) - PS The following process assumes that you have either Noise Ninja or Neat Image plugins for PhotoShop. There are times when you might well want to apply both smoothing and sharpening to an image. For example, you may have a noisy background caused by too few subframes or poor sky transparency but some portions of the image lack detail and could do with sharpening. These two procedures can often compliment one another. Start by creating a duplicate layer by selecting Layer - Duplicate Layer and name this new layer 'Smooth'. From the Layers palette, duplicate this layer and name the new layer 'Sharp'. Click on the 'eye' to the left of the 'Sharp' layer to deselect it. Select layer 'Smooth' and run your smoothing plug-in on it. Adjust the opacity of the smooth layer to get the amount of smoothing that you want to achieve. Click on the 'eye' to the left of the 'Sharp' layer to select it and create a duplicate layer by selecting Layer - Duplicate Layer. Select this new layer from the Layers palette and set the blending mode to 'Overlay'. The image will take on a high contrast appearance but this is normal. Select Filter - Other - High Pass and adjust the pixel radius to achieve the amount of sharpening that gives the result you want and click 'OK' (somewhere between 2 and 4 pixels normally works for me). Hold down the 'Shift' key and select both the top layers and merge them using Layer - Merge Layers. The next part of the process is to produce a layer mask by selecting Layer - Layer Mask - Hide Selection. It will appear as though the sharpening has gone away but it is still there hidden in the mask. Double click on the 'set foreground' colour square in the Tools palette and set all the CMYK settings to 0% to select white as the foreground colour. Finally choose a suitable brush size by clicking on the brush tool in the Tools palette and set mode to normal, opacity to 100% and flow to 100%. Now gently brush over the areas that you want to show through as sharp. Combining Smooth and Sharp (2) - PS The following process assumes that you have Neat Image or Imagenomic Noiseware Trial software. There are times when you might well want to apply both smoothing and sharpening to an image. For example, you may have a noisy background caused by too few subframes or poor sky transparency but some portions of the image lack detail and could do with sharpening. These two procedures can often compliment one another. Start by creating a duplicate file by saving the file as a TIF file with a different name with the word 'smooth' in it. Open the new file in your smoothing software, smooth it and save the smoothed file. Open the smoothed file in PhotoShop. Return to the original file and create a duplicate layer by selecting Layer - Duplicate Layer and name the new layer 'Sharp'. Select this new layer from the Layers palette and set the blending mode to 'Overlay'. The image will take on a high contrast appearance but this is normal. Select Filter - Other - High Pass and adjust the pixel radius to achieve the amount of sharpening that gives the result you want and click 'OK' (somewhere between 2 and 4 pixels normally works for me). Hold down the 'Shift' key and select both the top layers and merge them using Layer - Merge Layers. Copy this layer to the clipboard by selecting Select - All followed by Edit - Copy then swap to the smoothed file and paste the 'sharp layer' on top of it using Edit - Paste. The next part of the process is to produce a layer mask by selecting Layer - Layer Mask - Hide Selection. It will appear as though the sharpening has gone away but it is still there hidden in the mask. Double click on the 'set foreground' colour square in the Tools palette and set all the CMYK settings to 0% to select white as the foreground colour. Finally choose a suitable brush size by clicking on the brush tool in the Tools palette and set mode to normal, opacity to 100% and flow to 100%. Now gently brush over the areas that you want to show through as sharp. Recommended Additional Software - PS I have included some of my own PhotoShop Plugins on your DVD but I would thoroughly recommend the following purchases/downloads:- Noel Carboni'sAstronomy Tools - all sorts of useful processing actions and Noel Carboni's StarSpikes Pro - for the generation of diffraction spikes Imagenomic Noiseware (Community Edition is free) - smoothing software Neat Image (the demo version is free) - smoothing software PictureCode Noise Ninja - smoothing software Combining Narrowband Images - PS You must first align your three narrowband filter channels using MaximDL to allow for any changes in image orientation due to the filter changing, especially if using a manual filter wheel. Save each file as a 16 bit TIFF file. Open the three mono image TIFF files (Ha, OIII and either SII or Hb depending on your filter choice) and convert them to Greyscale by selecting Image - Mode - Greyscale for each image. Open the Channels palette by selecting Window - Channels. Click on the Channels palette menu and select 'Merge Channels'. Select 'RGB' from the drop down list box, set 'Channels' to '3' and click on 'OK'. Using the list box, specify which file applies to each of the three colours and click on OK. Using the narrowband Horsehead files included for the SGL Imaging Day, you should map Ha to Red, OIII to Green and Hb to Blue. For Ha + OIII + SII imaging, there are two commonly used colour palettes. The Hubble Palette assigns colours in descending wavelength order so that SII = Red, Ha = Green and OIII = Blue. The second commonly used palette assigns Ha to Red, OIII to Green and SII to Blue. Once you have assigned all three colours to their respective channels, your narrowband colour image will immediately appear and should be saved as a PSD file under a new name, along the lines of 'm42_2_210309_ha_oii_sii.psd'. This ensures that if any later processing proves to be faulty, you can always return to the original combined image. SBIGCCDSharp - deconvolution software

9 Combining Narrowband Images - MAXIMDL Open the three mono image FITS files (Ha, OIII and either SII or Hb depending on your filter choice) and decide which of the three channels, Red, Green and Blue you want to map the narrowband data to. Select Colour - Combine colour and then tick the Conversion Type radio button corresponding to your mapping - this would normally be RGB. Use the drop down lists to select which narrowband data to map to each channel. You must first align your three narrowband filter channels to allow for any changes in image orientation due to the filter changing, especially if using a manual filter wheel so click on thealign Button. In thealign Images dialogue box, use the drop down list to select 'Auto - star matching' as the align mode and enable the 'Bicubic resample'. Click on OK to align the images prior to combining them, the images will be aligned and you will be returned to the Combine Colour dialogue box. Click OK and the channels will be combined to produce a single colour image. Save this combined image as both a floating point FITS file and as a 16bit TIF file for further processing in PhotoShop. LRGB Combine - MAXIMDL Open the four mono image FITS files (Luminance, Red, Green and blue) Select Colour - Combine colour and then tick the Conversion Type radio button corresponding to your mapping - this would normally be LRGB. Use the drop down lists to select which filter data to map to each channel. You must first align your four filter channels to allow for any changes in image orientation due to the filter changing, especially if using a manual filter wheel so click on the Align Button. In thealign Images dialogue box, use the drop down list to select 'Auto - star matching' as the align mode and enable the 'Bicubic resample'. Click on OK to align the images prior to combining them, the images will be aligned and you will be returned to the Combine Colour dialogue box. Click OK and the channels will be combined to produce a single colour image. Save this combined image as both a floating point FITS file and as a 16bit TIF file for further processing in PhotoShop.