Introduction: - Welcome to CFAS! | Central Florida ...



Introduction:Planetary imaging is by far the most instantly rewarding type of astrophotography. Rather than trying to maximize exposure time you are trying to minimize it to capture the small windows of superior seeing. This tutorial will cover every detail and step needed to take a detailed picture of Jupiter (and other planets), from image capturing to final post-processing. Camera Selection:One thing a lot of people do with planetary astrophotography is try to use a camera that is not well suited for the task. For example, a DSI Pro I may be able to get an image of Jupiter on the chip, but the minimum exposure rate and most importantly, download speed, are not well suited for the application. If you are going to go to all the trouble of trying to produce a good image of Jupiter (image post processing is labor intensive!) you should make sure you have the right camera for the job. Here is a list of important camera properties to have for planetary imaging:High Download Speed High frames per second (fps) are crucial for planetary astrophotography. The more frames you can capture during the imaging session, then the more frames you can stack and the more processing you can do to pull out fine details in the planet without creating noise in the image.SensitivityIt’s imperative that the camera you use has a very high sensitivity. The higher the sensitivity, the lower the exposure needed during capture and thus the more frames you can capture and stack.Monochrome or Color?The choice between a color and monochrome camera is an endless debate on the Cloudy Night forums. However, if you look at all of the most spectacular images created by amateur astronomers, they are made using a monochrome camera. That is not to say monochrome cameras do not have their drawbacks and disadvantages. Monochrome cameras are more expensive due to the need to have filters, but result in an image with better resolution. Color cameras do not need expensive filters and the post-processing can be much simpler. Take a look on the Cloudy Night forums and see example images posted by other amateur astronomers. My Planetary ImagerI use the ZWO ASI120MM USB 2.0 camera. This camera is currently the most popular planetary imagers on the market today. If you search the Cloudy Nights forums, this camera is even what the “amazing” pictures come from. There is a guy down in Australia that takes the most amazing pictures with this camera and a C14. It has many advantages, from the “relatively” low price range (~$300) to being incredibly small and compact. It also comes with a neat 150 degree lens that screws directly on the camera. You can get some great night videos by placing the camera on a tripod and setting the exposure to every few seconds. It can also function as guide camera (even has a built-in guide port). This camera does not have thermoelectric cooling, but that is really not needed for planetary imaging. Figure SEQ Figure \* ARABIC 1: Photograph of the ASI120MM cameraTable 1: ASI120MM DetailsCamera SpecificationsSupported ResolutionsSensor: 1/3" CMOS AR0130CS(Color) / MT9M034(mono)Resolution: 1.2Mega Pixels 1280x960Pixel Size: 3.75?mExposure Rage: 64?s-1000sROI: SupportedInterface: USB2.0Bit rate: 12bit output（12bit ADC）Adaptor: 2" / 1.25" / M42X0.75Dimension: φ62mm X 28mmWeight: 100gWorking Temperature: -5°C—45°CStorage Temperature: -20°C—60°CWorking Relative Humidity: 20%—80%Storage Relative Humidity: 20%—95%1280X960@35FPS 1280X720@46FPS1280X600@55FPS1280X400@80FPS960X960@46FPS1024X768@54FPS1024X600@69FPS1024X400@101FPS800X800@66FPS800X640@82FPS800X512@102FPS800X400@108FPS800X320@158FPS640X560@98FPS640X480@113FPS512X440@123FPS512X400@135FPS480X320@165FPS320X240@215FPS2X2Bin：640X480@35FPSImaging Equipment Once you have selected an imaging camera specifically dedicated to planetary imaging, there are several other pieces of equipment needed in your optical train for planetary imaging. Barlow LensDepending on the native focal length of your imaging telescope, you will probably need to have a barlow lens in order to capture the finer details of the planet. It is important to have a magnification that matches the ideal arc seconds per pixel of your camera. The arc seconds per pixel of your camera and imaging setup can be calculate by:arcsec/pixel=P(206265)F (1)where P is the pixel size of your camera (mm) and F is the focal length of your imaging setup (mm).Depending on what barlow you have, there is some adjustment on the amount of increased magnification you will be getting. Once you have filter wheels and other accessories between the camera and the barlow lens, your magnification will increase. The figure below shows the magnification of TeleVue barlows with varying distance from the imaging chip. I typically image at 3.0X magnification (filter wheel in front of the barlow), which is pushing my little 8 inch Meade SCT a bit. You can either put the filter wheel before or after the barlow, depending on what magnification you want. Figure 2 below shows the effective magnification of several TeleVue barlows depending on the distance from the chip of the CCD camera.Figure 2: Magniciation of various TeleVue barlows at varying distances from the imaging surfaceFiltersIf you choose to use a monochrome camera, you will be using three filters for basic imaging of the planets; R, G, and B. There are high end filters (Astrodon) and low end filters (Meade) available today. If money is tight, the Meade filter set (~$35) is the way to go. However, these filters are not parfocal (meaning you have to refocus between each filter) and they do not have an IR blocking filter built into each color filter. That means you have to use two filters in series for each channel you image, i.e. clear IR filter + Red filter, etc. This causes you to have a slightly higher exposure since you are losing more light. I started out with the Meade filters and have since upgraded to Astronomik. I have found that the post processing involved with Astronmik is much less than the Meade filters and the amount of light that reaches the chip is noticeably increased. A comparison of the light curve for the Meade and Astronomik filters are shown below in Figures REF _Ref386958531 \h Figure 3 REF _Ref386958533 \h Figure 4.Figure 3: Spectral transmission of Astronomiks LRGB Type IIc filter setFigure 4: Spectral transmission of Meade’s LRGB filter set with the IR filter placed in seriesIn order to switch between the R, G, and B channels quickly enough during an imaging session a filter wheel is a must. Jupiter rotates so quickly that you are limited to how many frames you can take per channel before rotation becomes too much. This can be mitigated by using programs like WinJUPOS which we will talk about later on in this tutorial. I use a basic manual Orion 5 slot 1.25” filter wheel. There are much nicer filter wheels available that have the ability to automatically switch between channels via a USB connection to your computer and your image capture software. I believe the biggest advantage of a motorized filter wheel is that the filters are completely sealed from dust in the imaging train. I currently use some electrical tape on my manual filter wheel to help prevent light leaks and dust from entering at the manual slide. MountUnlike imaging DSOs you don’t need an incredible mount or a guiding camera/telescope. I have seen amazing pictures (much better than mine) from people using MANUAL tracking. However, it is greatly advised to have a mount capable of tracking since you will be so incredibly zoomed in on the planet you are imaging. There is enough headache already! Therefore it is important to have a good polar alignment, but you don’t need something that will keep Jupiter smack dab in the center of the chip for hours on end. Small movements of the planet you are imaging over time can easily be corrected with a slight slew of the mount. If the planet you are imaging is slowly drifting across your chip, then when you stack the images you are effectively reducing any dust spots that you might have.FocusFocus aids are very important. It is beneficial to have a Bahtinov focus mask and a micro focuser. The primary mirror focuser on SCTs is very course and moves the object to be imaged around the chip, making you re-center each time after focusing. This can make it almost impossible to know if you have achieved perfect focus. Therefore it is worth buying a microfocuser that attaches to the back of your SCT and allows fine focus adjustments without any image shift. Optical Tube AssemblyIt is beneficial to be imaging with a telescope that has a natively long focal length. SCTs are by far the most popular choice for planetary imaging for amateur astrophotographers. The larger the aperture, the more precious photons you will get and the more details you can pick up. Some of the best planet photos I have seen were taken from a C14. However, a larger SCT also means you will suffer from “mirror flop” and several other problems, but this does not outweigh the benefits of a larger aperture.My SetupMy complete setup for planetary imaging consists of the following: ASI120MM camera, 3X TeleVue Barlow, Orion Manual Filter Wheel (with Astronomik LRGB), and an 8 inch Meade SCT on a Celestron CGEM. Imaging SoftwareThere are many free programs available today that can be used for image capture with almost any USB camera. FireCapture is one of the best, allowing you to program specific settings for each color channel that is to be shot. It has many customizable options and if you are going to use WinJUPOS, it also has a file saving convention for that which can save you hours. For this tutorial, the image capturing process will be discussed in detail using FireCapture in section 6. Collimation and Preparation Seeing is king, as they say. It is important to have a good grasp of what the night sky conditions are going to be like and if it is worth it to go to all the trouble to try to image. You can have the best camera and telescope and on a mediocre seeing night your image will look something like an image from a toy telescope. If you are using an SCT, then collimation is IMPORTANT! You cannot get detailed images of any planet if you are not properly collimated. Collimation is something that is greatly dependent on sky conditions, which can easily limit how well you can collimate your telescope on a given night. There are several tools that can help you when collimating your SCT, some more useful than others. Below are some methods to achieving good collimation with your SCT. You should check the collimation of your telescope before EVERY imaging session!Basics of SCT Collimation:Collimation involves adjusting the secondary mirror so that it is perfectly aligned. On most SCTs there are three adjustment screws on the front of the telescope that when turned, adjusts the alignment of the secondary mirror. A slight adjustment of collimation can mean the difference between seeing details in Jupiter’s bands or a vague smudge. If you plan to collimate your scope a lot (it should be done before every imaging session) then it may be useful to pick up some of Bob’s Knobs to allow for easier and finer adjustment. The first step of collimation is to view a bright star near the zeneith at prime focus with the telescope (no barlows, etc). As you go in and out of focus, you will see the donut get larger and smaller. The initial step of collimation is to make this donut round and uniform. The concentric circles within the donut must be aligned and symmetric. Once the rings in the donut are fairly centered, you will want to switch to a higher power eyepiece. It is also recommended that at this point you collimate with the imaging train you will be using. It is important to allow your telescope to come to thermal equilibrium with the ambient environment, or you will have tube currents wreaking havoc on your image. After you have centered the rings do the best of your ability, you should then achieve focus and see the “airy disk” if sky conditions are great. Final tweaks are made to center the star within the elusive airy disk. The Duncan Method:Collimation may also be achieved by using the focus mask shown below. Find a fairly bright star near the zenith and place the mask over your telescope so the gaps are opposite of the collimation screws. By using a 400X magnification eyepiece you will see curved images of the gaps. As you adjust focus they will flip to lines pointing at the center. Adjust the opposite screw accordingly, and once collimated all of the lines will perfectly cross. This procedure is shown in Figure 5:Figure 5: Collimation process using a Duncan focus maskMetaGuide:When seeing isn’t great, or even if it is, this program will allow you to get much more accurate collimation of your telescope. MetaGuide stacks images in real time to help you see the airy disk when collimating. It will also tell you in what direction you need to adjust the secondary mirror. Figure 6 shows an example of MetaGuide during the collimation process. This program is free online. Figure 6: Screenshot of the MetaGuide UI ()Finder scope This may seem trivial, but it is imperative that your finder scope is VERY accurately aligned with your OTA. Either align your finder scope during the day on a distant object or use a crater on the Moon. Once you have a rough alignment, it is best to slew to a bright star and use a low magnification eye piece to further align the finder scope. Increase the magnification of the eye piece as needed for increased accuracy. The go-to on your telescope you will let you down 9 times out of 10 when at high magnifications. Image CaptureOnce your telescope has been properly collimated then you are ready for image capture! At this point you will already have your final imaging train hooked to the telescope. Go ahead and close whatever software you were using for collimation (MetaGuide, etc.) and open up FireCapture. This program is fairly self-explanatory, but we will go over some of the basic features. One of the best parts of FireCapture is that you can change the size of the image you are actually saving and downloading to the computer. You can enter a dimension in pixels for your imaging size by using the ROI option. Make the image size as small as possible while still allowing breathing room for the planet to move around the chip. The smaller the image size, the higher the fps you will be able to capture. Gain, gamma, and exposure are easily adjusted in the program and are remembered for each filter channel you select (top right). It is important to change the file naming convention to be compatible with WinJUPOS. GO to settings>parameters and select winJUPOS file naming convention. I prefer to use the saved file type as SER rather than AVI, just because I have found that with large video sizes (8 gigs) programs prefer SER. Turn on the histogram feature, this will allow you to see in real time the histogram value and what the average value was for previous channels imaged. As a rule of thumb, you should try to have Jupiter at a histogram value of ~70% for each channel. After collimation you will still be centered on a fairly bright star near the zenith, this is perfect for your initial focus. Throw on a Bhatoniv focus mask and adjust the focus accordingly. Seeing will go in and out so take your time, as good focus may be hidden during a period of bad atmospheric conditions. The next hurdle, once you have made it here, is actually getting the planet to be imaged on your chip. This can be easier said than done. Make sure you have an excellent alignment between your finder scope and imaging OTA. Slew your telescope to Jupiter, 9 times out of 10, it will not show up on your laptop screen. This is no surprise considering the incredibly small FoV you are using. Go ahead and take the imaging camera off the telescope and insert an eyepiece. You will see an out of focus donut, but do not adjust the focus! Center the out of focus donut to the center of the eyepiece and then place the camera back on the telescope. Several iterations of this may have to be done, but eventually you will get the planet on the chip. Now, there are several options you can do with regard to focusing at this point. If the star you focused on with the Bathoniv mask was close to the planet you are imaging, then it might be worth the time to run a series of image captures at the focus you are at. If not, it is at this point you can either slew to a moon of the planet and use a bhatoniv focus mask, or use your own judgment when adjusting focus on the planet itself. When using visual judgment, lower the gamma to 0 and slightly angle the laptop screen back to increase the contrast of the visible features of the planet. Keep the gain near 50. This is where the microfocuser shines, if you were to adjust the primary focus knob (moving the primary mirror) you will also be moving Jupiter around the chip and will have to re-center after each adjustment. You will also be able to notice mirror flop, which adds another layer of difficulty. Adjust the focus until you feel you can see the best contrast of the planet visual features. After each focus adjustment, wait several seconds to make sure that it is not a change in seeing rather than focus. If your microfocuser has a digital readout display, then you can find the correct focus for each color channel and easily adjust to that number when you switch filters. The blue filter should need the most focus adjustment compared to red and green. Once focused, it is time to setup each color channel in FireCapture if you are using a monochrome camera. A screenshot of FireCapture during an imaging session is shown below in Figure 7. Start by selecting the red filter and make sure your actual red filter is in the imaging train. Adjust the gain to somewhere near 75 or so (I have used as high as 85). Change the exposure accordingly to achieve a histogram value of nearly 70%. We are also going to want to automatically limit the time for each channel. In this case since we will be using WinJUPOS for post processing we will do 300 seconds per channel. Once you have done this for the red filter, switch to blue on the pull down menu and then change to the blue filter in your optical train. Repeat for the green filter. FireCapture will remember these settings for future use, but you will have to do a preliminary check of them at the start of each imaging session (and during) as seeing conditions will vary. Figure 7: Screenshot of an imaging session with FireCapture using the green channelYou will need a significant amount of hard drive space available for each imaging sessions (~100 gigs). Each 300 second video will be somewhere near 7 gigs in size. Seeing will constantly change throughout the night. It is important to check your focus between every image capture or every few. Imaging ProcedureIn short, here is the basic imaging procedure:Polar Align MountCollimationFocus (either Bhatoniv focus mask and/or visually on the planet)Setup FireCapture settings for each channel (RGB)Run a capturing run (R, G, and B)Adjust focus on planet by visual inspectionRepeat steps 5 and 6Image ProcessingTo get the most detail (or any) out of your image involves heavy post processing. Here are the programs I use for image processing in chronological order, some are repeated:Autostakkert! 2 (Free)Registax 6 (Free)WinJUPOS 10.1.2 (Free)Autostakkert! 2 (Free)Registax 6 (Free)Astra Image (Free Trial – Optional for Processing)WinJUPOS 10.1.2 (Free)Photoshop (Can get free trial)The first thing to do after your imaging session is pick through all of your raw red channel videos and see which ones have the best raw quality. To do this, we are going to use Autostakkert! 2 (AS!2) to stack all of the images and create a lightly processed stacked image for each set. AS!2 is a very basic but powerful program, the settings menu is shown in Figure 8.Figure 8: Screenshot of Autostakkert! 2Click open, and select the first red channel video file you would like to stack. Select planet (COG) and dynamic background. For the quality estimator, select gradient for large planets, such as Jupiter and Saturn, or Edge for smaller planets like Mars and Venus. The noise robustness depends on just how noisy your raw images are. Typically this should be left at a value of 3, but if your image is very crisp (little gain used) then you can use a higher value. Click Analyze. This will order your frames in terms of quality and buffer them in the program. The quality graph will show in grey the image quality over the imaging session in chronological order, while the green line represents the order of quality that AS!2 put the images in. Select TIF under stack options. Use the slider, shown in Figure 9, to go through your sorted images. Try to decide what percent of the images you would like to use. I usually do not use any images that are below a quality of 50%. In this example shown, that would mean stacking about 50% of the frames. It is sometimes better to stack less images than you think, as adding poor images will decrease your noise but ruin any hope of seeing fine details of the planet. It is worth some experimentation. Figure 9: Screenshot of the image viewing window in AS!2Select normalize stack and set it to 85%. This will account for background variations between different frames if seeing conditions change. For the sharpened image, you can set it to 50% or try other values. This is simply going to output a slightly sharpened image of your stack. Do NOT use this image for processing later on, only as a comparison between stacks. There will be image artifacts if you are going to use wavelets in Registax later on. Select HQ refine and then click Stack. Repeat these steps for all of your raw red channel videos. Once done, compare all of the slightly sharpened stacked images and decide which ones are the sharpest. The fewer you select the better, as the rest of the image processing is incredibly lengthy, tedious, and will give you a headache. Once you have selected the best set of images to use, go ahead and stack the corresponding green and blue channels. Now it is time to do a little processing in Registax of these processed images before bringing them into WinJUPOS. Open Registax 6, click select, and select your stacked red channel image that was no sharpened. It can be tempting to use the default wavelet filter in Registax, but it is much better to use the Gaussian filter instead. It takes a little more work and is more temperamental, but the end result will be worth it. You will end up with a more natural looking Jupiter. Select Use Linked Wavelengths, pull the slider tab of layer 1 to 100, and then adjust the noise and sharpening values accordingly. I usually have a noise value of 0.25 – 0.35 and a sharpening value of 0.1 – 0.25. Now move the layer toolbar to about 5, set the noise value from 0.2 – 0.35 and the sharpening from 0.1 to 0.15. For other planets, however, it may be better to simply use the default wavelet function in Registax. I have noticed Saturn and Mars have a poor response to Gaussian wavelets in Registax compared to using the default filter. These values will vary from image to image and there is no cookie cutter method, but this gives you an initial ballpark to try. You can then save this scheme by pressing the Save Scheme button at the bottom. Saving many schemes is a great thing to do and will allow you to better compare different image enhancements. Figure 10 shows the initial wavelet processing discussed above. Figure 10: Screenshot of a raw stack with a Gaussian wavelets applied in Registax 6Repeat these steps for the green and blue channels. These three images will not be used for your final compiled image, but instead only serve as an initial image measurement in WinJUPOS. Open WinJUPOS and click Program > Celestial Body > Jupiter. Now click Recording > Image Measurement…, a screen will appear as shown in Figure 11. Click open image (F7) and choose the red image you processed through Registax. If you saved with WinJUPOS file naming convention in FireCapture, then the time of the video will auto-populate. Enter your Geogr. Longit. and Geogr. Latit. Click the Adj. tab and select Outline Frame > Automatic Detection. This will usually get a pretty good outline of Jupiter. It is very important that the north pole of the image outline (N) is actually on the north pole of your image. You can use the ‘n’ and ‘p’ keys to rotate the image. The Page Up and Page Down keys will increase and decrease the size of the outline, respectively. You can also check the box that says LD compensation, and this can help aid you in correctly outlining Jupiter. It is better to make sure you are a little over the edges of the planet rather than directly on. If you fail to outline the entire planet, then you may very well have a sort of “onion ring” artifact later on down the road. Save this image (under the Imag. Tab). Now repeat this same procedure for the Green and Blue channels. Figure 11: Screenshot of the imaging measurement panel in WinJUPOSOnce you have created three image measurement files (.ims) for each channel, click Tools > De-Rotation of Video Streams in WinJupos. Where it says original video, click “…” and select the red channel video. The time fields will auto-populate for you, if not, then make sure you have any program that might have the video open closed. Now click select the image measurement file you created for the red channel by selecting the “…” where it says “Image Measurement of a Preliminary Image from the Original Video”. Click “Start De-Rotation of Video Stream” and WInJupos will start de-rotating your video. WinJupos selects the frame from the middle of your video as a reference to rotate all other frames to. Repeat these steps for the green and blue channels. After you have a de-rotated video for each channel, open AS!2 again. Go ahead and stack the three channels as you did before, but this time using the newly de-rotated videos. Once done, open Registax and process the images as before. You may notice that there is a different response to the same settings you used for the measurement images for WinJupos. After you have all three channels processed, open up Astra Image. In this program you can do slight deconvotultion to each photo. However, this program is pretty optional in post processing unless you are very serious and have excellent data (I only have the trial version of this and do not really use it much). At this point you may also do some editing in Photoshop on the three derotated color channels. Be careful in how and what editing you do, because the RGB combination later can be distorted and it will be hard to achieve color balance. Now open WinJUPOS again and create measurement images of the three final derotated color channel images. Then go to Tools > Derotation of R/G/B frames… Select the three measurement images you just created for the R, G, and B channel. Then select the red channel for the luminance option. You may have to experiment with the LD values, as lower values can help get rid of any “onion rings” you may produce in your image. Change the image type to TIFF and then select “Compile Image”. Some planets, such as Saturn or Mars may look better only doing an RGB compilation. I have found Jupiter to be the hardest in terms of post-processing and achieving proper color balance. With the LRGB image you just produced you can do final touches in photoshop, such as curves, levels, final unsharp masking, etc. Enjoy your final image! ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download