In the winter of 2002, I got this crazy idea that I would like to take pictures through my telescope. Even though I have a good friend named Jeff Ball who takes excellent pictures, and even though I've watched him build his mount and then spend 1-2 hours polar aligning it - I thought that I could use a video camera to take "snapshots" with my telescope and sort of short circuit the process. On this page I hope to detail what I have learned both by trial and error and by reading things.
Well just what have I learned?
Astrophotography is harder then you think.
How can you image (take pictures) with your telescope? There are several methods.
- Use your telescope or its tracking mount to hold a camera - this has been primarily used with 35mm film cameras to image the Milky Way. Though now people are attaching Camera or other lenses to their CCD or video cameras and imaging large areas of the sky in a digital format.
- Use a film based camera to image through the telescope
- Use a digital camera to image through the telescope
- Use a CCD camera (very sensitive/expensive astronomical digital camera) to image through the telescope
- Use a video camera to image through the telescope
For more information on this take this link to a brief discussion on each of the above.
Why Video and not a digital camera?
Well:
- The StellaCam series of cameras have a very sensitive CCD chip
- Most deep space objects look grey and white through the telescope anyway
- I thought that polar alignment would be less critical with video because of its shorter exposure times
- I thought it would be easier to focus (focus is a bane of astrophotography)
- I thought I could beat seeing (more on this below ... this has to do with imaging planets)
Early Stellacam images from an original Stellacam
The Little Dumbell Nebula, a planetary nebula in Perseus
|
M 36 - open cluster
|
Seeing:
The atmosphere is often the limiting factor in both visual observing and in imaging. In WV, where we spend a lot of time observing from mountain tops, the seeing is often poor. High power views of Jupiter look like a blurry ball that occasionally distorts into an oval. But then ... occasionally ... the sky will become less turbulent and the air will steady and the observer is rewarded with crisp detailed views of the king of the planets ... then it turns to mush. The idea with video is that if you keep recording you may capture those instances when the seeing steadies up. You can either save images to tape using a camcorder or you can capture images directly to the hard drive of a laptop computer. This allows you to take hundreds or thousands of exposures ( at 30 frames a second they add up quick ) and then select only the best to process.
Capturing planets or moon
Video works well on bright objects like the sun, moon, and planets. You can capture 100's or 1000's of video frames, select the best ones (either manually with a program such as AVI2BMP or automatically with a program such as Registax 2), average them together (with a program such as Astrostack or Registax 2) to improve the signal to noise ratio (clean up the image) and produce a nice picture.
If you do this manually, you must sit there and click through all those images one at a time to select the best ones! This is tedious when you have thousands of frames to view. This is why many opt to use programs such as Registax 2 to do that work for them.
When I got started I was recording to High 8 metal video tape on my SONY cam corder. I captured images from the tape with a SNAPPY 4.0 device. This grabs images only one at a time. I had a hard time as I needed to keep the tape running in order to grab a frame and I could not always "catch" the one that I wanted. You might get around this if you have frame advance on your VCR/Camcorder. But it is CHEAP!
My next step was a USB instant video which grabbed 15-24 frames/sec. This lets you get enough frames to use the above technique. Eventually I realized that the USB device was capturing at 320x240 resolution not 640x480.
This led me to switch to a Imperex Video Capture Essentials Card (the new one can capture 30 frames, visit AVA and look for video capture devices). This gets only 4-5 frames a second, but at least it gives me full chip resolution. Full chip resolution is necessary for high resolution imaging of planets in order to capture the most detail. So, now, I watch the image of the planet on the computer screen, when the seeing improves and the planet image sharpens I hit record. I usually grad 10-60 seconds of video. You are limited in what you can grab by you computers memory. Full chip resolution AVI's grabbed at 5 frames a second eat up disk space quickly. It is not unusual to have AVIs that are 300-500(+) MB in Size! And with this system, I can still record on tape and capture to the laptop computer later (though this does cost some resolution - about 20% - high 8 tape is 400 vertical lines while the direct lap top capture gives 480 vertical lines!). Recording to tape is better if you want to collect a lot of video to process later. The tape holds much more "data" then the laptops disc.
Once the data is on the computer in the form of an AVI you can begin to process the data to make a picture. Originally I'd use AVI2BMP to select 10-40 frames and convert then to sequentially numbered BMP files. Then I'd align and "stack" (average) those with Astrostack to make an image. Now I process larger AVI files with Registax 2 which I use to quality screen, align and "stack" a larger number of frames (100-600). Stacking improves the signal to noise ratio. Registax also lets you do "wavlet processing" (I have no idea what this is doing) to bring out fine details.
But what about accurate polar alignment?
No matter how you image, a tracking mount makes things a lot easier. The more accurately it tracks, the longer what you want to image will remain on the video chip. This helpful weather you are trying to image a highly magnified planet or a minified deep space object. Good polar alignment of your mount either makes things easier or is essential to obtain a satisfactory image (you can video with a nontracking Dob but it is HARD!).
What are the requirements? This varies by what technique you are using to image with.
- film through the telescope - this requires very accurate polar alignment, and for long exposures it requires either that you manually guide on a a star or use some type of an auto guider.
- CCD - less stringent requirements then for long exposure work with film. If you auto guide (or manually guide) then you may get by with less accurate alignment especially if your exposures are short.
- digital camera - exposures are shorter. But, the higher power that you image with the more refined the polar alignment should be.
- video - hey, some folks do this with an un driven Dob. But a drive will keep the image in the FOV longer which makes things easier. If you intend to "stack" frames to image DSOs, the better your polar alignment, the easier it should be to stack (add together) video frames in a program like Astrovideo (which adds or integrates frames to build a brighter image). How are you imaging? If you are using focal reducers to image as big a field as possible then your tracking requirements are much less then for when you are trying to image a planet magnified with a Barlow lens.
With video, if your alignment is "off", You can compensate for this to a degree with an electronically driven mount. I took a video movie of Saturn being occulted (disappearing behind) the moon. Though my polar alignment was poor I was able to keep Saturn in the camera's field of view by using the mount's "slo mo" controls to keep Saturn in the field of view of the video's imaging chip. The more highly "magnified" your image is or the smaller the size of the chip that you have in your camera, the more desirable a feature this becomes.
Well what about Field of View?
Most CCD chips that aren't very expensive have a small field of view!
Video chips tend to be some of the smallest. The table below describes the field of view of the StellaCam's chip through the telescopes that I have used. This small field of view makes it hard to get an image onto the chip's surface. If your mount is not tracking well (i.e. not well polar aligned) then the object will tend to drift out of the field of view of the chip (unless you have 'slow mo' mount controls to keep it there). And the small FOV will limit the size of the objects that you can image with it.
| Telescope | Average pixel size in arc seconds (approximate) |
Chip field of view in arc minutes |
| 80mm Orion f5 "short tube" |
4.5 | 34x57 |
| 90mm Meade ETX | 1.4 | 12x18 |
| 6" f8 reflector | 1.5 | 13x19 |
| 10" f5.6 reflector | 1.3 | 11x16 |
| 20" f5 reflector | 0.7 | 6x9 |
Another way to look at it is like this. One source said that a "1/2" inch chip (the type in the StellaCam) gives a field of view similar to a 7mm Orthoscopic eyepiece with a relative field of view of 40-45 degrees! Thus the actual FOV of the chip is much less then a degree. With my 10 f5.6 the FOV is about 11 by 16 arc minutes (only 1/6 by 1/4 of a degree) . You can increase your FOV size and increase the amount of light seen by each pixel by using a focal reducer. The folks at AVA recommend a f3.3 reducer for the f10 SCT's.
As the chip FOV is small, it is nice to have a good medium power finder to help place objects to be imaged onto the chip and a well aligned tracking mount with electronic fine adjustments to keep it there. Your finder must be accurately aligned in order to use it to try to place an object onto the chip.
Telescopes/lenses that I've used for videoastronomy:
I have been experimenting with video imaging of planets, the sun and DSOs with an Astrovid Stellacam EX. I have primarily been using my 10" f5.6 on a Tom Osypowski "Compact" Equatorial Platform. The main problem with this mount is its lack of 'Slo mo' controls. Tom makes a dual axis platform that has these. If you consider imaging with a Dob, I'd advise using his dual axis platform. Recently I modified the ground board of my 10" scope so that it can ride on the Dual Axis platform that Tom made for my 20" scope. This has helped in several ways: the scope is now "over mounted" and rock steady and the mount has electronic "slo mo" controls that help greatly when trying to center an object, particularly at high power! I've also had better results with DSO's since I obtained a Steve Mogg f6 reducer which takes my 10" f5.6 scope's effective focal ratio down to about f3. Lowering the focal ratio brightens the image and widens the FOV.
I've taken a few images through the 20" f5! This is harder to use as it has a much smaller FOV.
I have also used an Orion Short Tube 80 on a Alt/Az mount for DSO's. This yields lower resolution but wider field of view images then what I can obtain with the "focal reduced" 10" scope.
I've also tried a 25 mm F? "C" mount lens attached directly to the camera and mounted on a photo tripod.
Random thoughts on things to consider when imaging:
In order to image planets you want to achieve a large image scale. My understanding is that you want to magnify the image so that you get a large image of the planet on the chip. This way the planet is imaged by more of the chip's pixels and the resulting image is higher in resolution. I achieve image scale by using a barlow (2x) or a Televue Powermate (2.5x, 5x).
However, at baseline, the FOV of the chip at prime focus is small. When I add a 2x Barlow to get some magnification and improve the image scale the FOV becomes smaller. Now it becomes harder to get and keep the planet on the chip - especially if you use a dob! Good polar alignment of your tracking platform/mount is needed to keep the object in the FOV of the chip as the coverage of the chip as (at prime focus) the FOV is only 11 by 16 arc minutes (about 1/6 by 1/4 of a degree!). With the 2x Barlow the FOV is only 1/4 this size! Just getting an Object on the chip with the Barlow in place is a challenge with the Dob mount. This is where the electronic "slo mo" controls of the platform help a lot. This is where a good fork or equatorial mount is a big help! No wonder SCT's are popular for imaging, you can polar align them, they have electronic controls, they can be focal reduced to f3, and then have good focuser travel.
Generally, I'll get in the area of the planet with the finder, use a low power Eyepiece to center, refine my object centering with a high power EP, focus with an EP that is parfocal to the Stellacam, insert the Stellacam and fine tune the focus. Then I View or image. If I want to try for more image scale (higher magnification) I: insert a barlow or powermate, recenter with a high power EP, focus with a parfocal EP, insert Stellacam and fine tune the focus. Then I view or image.
Focus is an issue. I've found that it is easier to find a bright star (Orion's "belt stars" are good for this) and focus on that. If the object is too dim you can't see the diffraction rings of the out of focus object and it is hard to focus on an object when all you see is a black screen. It is helpful to make the camera as sensitive as is possible. With the StellaCam, I use the ELC setting with the AGC on. After the target star is focused, I move to the object that I wish to video. Then I adjust the camera settings to make the object look "good".
Viewing the real time video:
The Stellacam EX has both a "S" video and a coaxial output. When I record to the tape on my camcorder, I use the "S" video to hook up to the camcorder and send the coaxial output to a 9" TV (I covered the screen with a red plastic report cover attached with Velcro tape - my wife has tolerated applying of Velcro tape to the front of the TV quite well - your spouse's response may vary). It is much easier to focus and see what I was recording by looking at the TV rather then the small screen of the video camcorder; though the video camera is easier to take in the field.
If I record to my computer I just use the S video output and feed that into the ImperX VCE adaptor which is placed in the laptop's PCI slot. I use the computer screen as a display and to focus with. But, you can still take the other output from the Stellacam EX and send that to a TV monitor or camcorder. Some folks feel that objects are better viewed on a TV or B&W monitor (that has brightness and contrast controls) then on the lap top. I don't have an opinion on this as I usually use either one or the other and I haven't done a "side by side" comparison. I'm usually imaging and now "viewing".
video tape
When I tape my video output with my "high 8" camcorder (digital video would be better), I use good quality metal high 8 tape. If I use the computer, I capture directly to the disc of the laptop computer with the Imperex VCE PCI card. I can record longer with tape then the computer. The tape is good for meteors as it can last for hours! 336 frames is about 300Mb of memory. So the more frames that you record, the faster you use up your memory. With the VCE card, I record 640-480 video at 0.5 to 5 frames/sec.
Processing planets:
IF I'm using Astrostack to process, first I'll load the AVI file into AVI2BMP (free ware, US version) and review the individual frames. I check the ones that look "good". Then I batch save checked frames only. Then I load those selected frames at 640x480 resolution and add the images in Astrostack (freeware) using 2x oversampling for planets and no oversampling for DSO's (full frame selected for DSO's). Some degree of Unsharp masking 3-12x using a 5x5 or 7x7 Gaussian Kernal seems to help to sharpen all images. My default is 5x5 gaussian filter with unsharp masking set at 3. Reviewing the "grabbed" images (in Astrostack) prior to stacking is helpful as the effects of seeing dictate that some images are better then others. Astrostack allows you to assign a weight to images of different quality and to toss out the really bad ones. (Thus I screen for image quality twice - both in AVI2BMP and in Astrostack.)
More recently I've started shooting more frames and the above method was to tedious. I discovered Registax and then Registax 2. Now I process all my planetary video AVI's with Registax 2. This lets me process many more video frames.
My understanding is that Astrostack and Registax are programs that take frames and average together them to improve image quality (improves SNR). These programs are usually used to process video AVI's taken of planets, the moon, or bright star clusters to reduce noise in the video image.
Imaging/viewing Deep space objects
Object size and chip FOV are important considerations. For most imaging I use a focal reducer (f6 in my f5.6 scope) and f3.3 in the club's C14. This brightens the image and buys me a larger FOV.
For most objects I insert the focal reducer, with the camera screwed on to it, into the focuser. I focus on a bright star. Then I locate the object either by "go to" with the club's scope or with a Telrad and finder scope on my 10". With my scope, if it's not in the chip's FOV, I center with a low power EP and then reinsert the camera/focal reducer. Then its time to tweak the focus and then to view or image.
This works for most objects. For Planetary nebula, I localize as above, center on the chip, remove the focal reducer from the scope, remove the focal reducer from the camera (unscrew), focus with a eye piece parfocal to the camera, reinsert the camera at prime focus, tweak the focus, and then view or observe. You can then repeat this with a barlow or powermate, but it had better be a bright object.
Some feel that objects are best viewed (real time video feed) on a TV or B&W monitor with contrast/brightness controls that you can adjust.
Processing deep space objects - original method
Astrovideo actually adds (integrates) frames and thus is better for "building up" frames in an effort to brighten a DSO. Although, it does add up the noise as well. This has become my software of choice foe producing images of DSOs. Astrovideo has a "drift" align processing feature. This allows me to use a dob or a poorly tracking equatorial mount to image and then add those "frames" together to produce a brighter image of a DSO. I have used this to add 50-1000 frames! This gets into minutes of CCD exposure - with a Dob! The program compensates for the scope's drift by shifting the image before adding frames. This tends to "shrink" the effective size of the captured image (the more the drift, the more shrinkage), but works quite well.
I have been capturing AVI files with the VCE's (ImperX capture card's supplied program) software and later processing the AVI files using the "drift align" feature of Astrovideo. After adding video frames in Astrovideo, I tidy up the images in GIMP. I usually crop and then use the histogram to clean the image up.
I have not figured out how to use dark frames with "drift" align processing of AVI's in Astrovideo - so for me noise is a problem. There is a lot of noise in the top left of the Stellacam EX image. Noise from the chip is a limitation when using this camera (Stellacam EX) to image DSO's. To minimize the effect, I always try to position the object to be imaged on the right hand side of the chip. Later when processing with GIMP, I crop out the noise on the left.
More light is good, now I use one of Steven Mogg's 0.6 focal reducers when I image DSO's with the 10" and a f3.3 reducer on the C14, this does help to view/image some nebulous objects (see gallery link below).
Processing deep space objects - current method
I am imaging either with a Stellacam EX or a Stellacam II (the latter I think is better for DSO's based on limited experience) and the camera is integrating frames and producing an image every 1,2 - 8.5 seconds. I collect a frame only once every 1,2, - 8.5 seconds. Rather then saving these as an AVI and "eating up the hard drive", I have begun to save these as FITS files. Astrovideo allows you to either save the individual frames or to save just one summed image. In order to minimize disk usage, I have been producing only one summed image of each DSO that I image. Using the capture dialog I set frames per image to 1, number of images (number of frames to add) to 10-300, autosave images is unchecked, auto add images during capture is checked, and the time lapse timer is set to 1000, 2000, 8500 milliseconds depending on if I'm capturing 1, 2, or 8.5 second frames. This lets me capture 10-300, 1-8.5 sec long frames, and add then to produce an image that is 10-2550 seconds long. Kind of makes your head spin. This lets you use the Stellacam EX or II to image like a B&W CCD camera using a combination of on and off chip integration.
Here are images on M78 with both cameras.
|
M78 Stellacam EX (150x2seconds) C14 at f3.3 |
M78 Stellacam II (30x8.5 seonds) C14 at f 3.3 |
The sun:
I recently acquired a Meade Electronic Eyepiece. This is a "1/4" video chip that is mounted in a camera head which resembles an EP. It has a contrast adjustment but no gain control. I got this as it is the only video camera that I could find which can image with my Coranado NearStar. This 'scope lacks sufficient in travel for other Video units which have the chip located behind rather then within the EP barrel. I use this camera for H-Alpha images of the sun. When I do this I need to put an IR (infra red) filter in order to get satisfactory images. I grab images at 5/sec with the VCE to the laptop. I process in AVI2BMP and Astrostack. Then I tidy the image in GIMP, I find that using the curves tool lets you bring out more detail then with the histogram.
This camera can also be used to image planets and the moon. It has a contrast control but no gain control so you must magnify the planet image to take up a good portion of the chips surface or the camera's auto gain will over saturate the planet's image and you'll just see a white blob.
My Image Galleries
Thoughts on Settings for the Stellacam:
Stellacam II:
For DSO's with a telescope I like gain at 50%, contrast (gamma) lo, integrations at 128 or 256x for real time viewing. With the 25 mm "C" mount lens I like 50-60% gain, contrast "hi", integrations at 32x.
Stellacam EX
For DSO's with a telescope: ELC mode, gain on MAX automatic, sense up at 128x.
For planets/moon with a telescope: ALC mode - adjust shutter speed to optimize the image, gain set on off, sense up set on off.
Others know more then I do, here are some links.
- Videoastro there are many links off of this page. Many are essays on videoastronomy or how to articles.
- Steve Massey's page, check out his videoastronomy section and his links page
- Craig Zerbe's page want to know the hows and the whys? This is a good place to look.
- Steve Rismiller's solar video astronomy
- Pat Stoker's fine page of images. his work is outstanding. please check out his links page.
- Renato Terabella's amazing images
- Bob Pilz's Planetary / Lunar Imaging Using a Video Camera
- GY's amazing planetary images, he images with a web cam and is one of the best!
- Astrovideo is a program used to do long term exposures of DSO's with a video camera! (not freeware)
- Astrostack freeware program lets you stack your video images to build up detail (freeware)
- Adarondack Video Astronomy (aka astrovid) They sell video astronomy equipment and supplies.
- Videoastro.com Joseph Campbell sells video astronomy equipment too. Check out his FAQ page.
Where can I get a Camera?
- VideoAstro.com - This guy started playing around with low light security cameras and then later began to sell them to others over the web, his designs are simple, he now offers two color cameras as planetary imagers. One is a 1/4" chip the other a 2/3" chip. These are automatic type cameras.
- AstronomyQuest - Their Sky-Cam II made Sky and Telescopes "hot product" list but I've not been "blown away" by the images that I've seen so far.
- Meade or a Meade retailer will carry the Meade Electronic Eyepiece (I have no experience with this vendor, I just liked their page) This has only 320x240 pixels. It is easy to use and it is what I use for H-alpha. It has a "contrast" control.
- Adironack Video Astronomy They make many different types of video cameras with 1/4"-1/2" chips, B&W or color, many have remote controls and all have either electronic or manual controls to change the cameras settings. Their StellaCam series (like the ITE cameras) is built around a Minitron camera with a sensitive B&W Sony HAD chip. PlanetCAM, StellaCam, Astrovid 2000 - a whole line of cameras.
- ITE (ccd imaging aids) They sell Mintron's too. Their cameras have the controls on the camera rather then having remote or electronic controls.
- All kinds of people are marketing video cameras, I saw this for sale from a Canadian store. Focus Scientific Ltd. I know nothing about this one.
- Or you could just modify a web camera. The QCUIAG site can show you how!
How do you attach the camera to the telescope?
The Stellacam (EX or II) comes with a "C" to 1.25" adaptor so it slides into your focuser just like a 1.25" eyepiece. As there is not eyepiece or lens between the telescope and the video camera's CCD chip this is called prime focus photography. You can boost the magnification, increase the resolution of your images at a cost of decreasing your field of view by placing a Barlow lens between the StellaCam and the focuser. I know that with SCT's you can get a "focal reducer" to increase the field of view and these are commonly used by SCT imagers with CCD cameras. Folks with SCTs use a f3 Focal Reducer. Steven Mogg sells a 0.6 focal reducer that screws into the filter threads on the Stellacam's C to 1.25" adapter. You can use this to increase the FOV of the chip and to increase the amount of light seen by each pixel of the chip!
How do I record my pictures?
You have two options: record to video tape or directly to the computer. VHS and 8mm give you about 240 lines of resolution. High 8 and super VHS give you about 400 lines. Digital 8 tape gives you about 525 lines. The direct capture method gives you whatever your full chip resolution is. This is usually the way to get the best resolution. If you just want to tape images you can. But if you want to 'play with the images then you need to get the data off the tape and into the computer. If your VCR/Camcorder is analog then this will require some type of an A/D device. If your camcorder is digital it may be able to act as an A/D converter for your computer.
Back or Home or To the Image Galleries