Blog

ADVENTURES WITH FOCUSERS AND COLLIMATION

I recently discovered/realized that my last 3 images with my LX200 scope were made when it was significantly out of collimation - and thus the images appear very fuzzy (not well-focused). It took over a month to realize what the problem was, why it had occurred, and fix it.

It was late June, and I was taking images every night (most of the nights during the summer are clear); the images were OK, but the stars in my M16 (hydrogen alpha) images seemed a bit bloated. I realized that I had not collimated my 12" LX200 OTA for more than a year (since it was installed in the observatory in April 2010). As it was a full-moon period, I decided to bite the bullet and spend time in the observatory collimating the scope. It's not that bad of a process, and I enjoy being out in the observatory ... but I was nervous, considering some of my prior collimation experiences.

COLLIMATION

Collimation is the process of ensuring that the mirrors of the telescope are centered on the optical axis, and orthogonal to it. For an SCT like the LX200, the centering of the mirrors cannot be adjusted, but the tilt of the secondary (mounted on the corrector plate at the front of the scope) can be adjusted so that it is parallel with the primary mirror. This results in the best focus, and thus smallest star sizes (full width half maximum, or FWHM). As Richard Bennion has shown, smaller star sizes also mean brighter and higher resolution nebula images, as well. The ability to focus a star to a small spot on the CCD chip is dependent upon a number of factors - including accuracy of automatic focusing program, seeing (atmospheric stability), tracking accuracy, periodic error of the mount, and collimation, to name a few. Thus, with a poorly collimated scope, it is impossible to focus to a very small spot, and the spot may not be round. As an example, my LX200 images which should have star sizes of 2-2.5 arcseconds FWHM could measure 3.5-4 arcseconds when the scope is out of collimation. This makes a huge difference in the quality of the resulting image.

The original collimation adjustiment mechanism on the LX200 consisted of three hex screws that are adjusted tighter of looser to tilt the secondary mirror ... and, when completed, to lock the mirror in place (i.e., all three collimation screws must end-up fairly well tightened). These are difficult to adjust - requiring a hex wrench, and inserting it into the small hex screws in the dark. And, if you drop the hex wrench, it could land on the corrector plate of the scope (potentially scratching, cracking, or breaking the corrector). About two years ago, I replaced the Meade hex screws with "Bob's Knobs", a set of screws that replace the Meade hex screws, and provide a nice knurled cap that can be used to adjust the collimation easily. You can learn more about Bob's Knobs here.

The first time I tried collimating the scope, not long after acquiring it about 3.5 years ago, I tried to be very careful to follow the instructions given on various websites - namely, to only loosen one knob at a time, and not to make any of them too loose, or the entire secondary structure could fall out of the corrector plate, and (potentially) fall onto the primary mirror - pretty much destroying the scope! While there are various laser collimators that can be used with an SCT, I was just using a magnified, out-of-focus star image - which looks like a donut. The central "hole" (black portion, corresponding to the secondary mirror, which blocks light coming into the telescope in the center) needs to be precisely centered in the donut (bright diffraction rings seen in the unfocused star image). By visually centering the hole in the donut, you are collimating the secondary mirror of the scope. As I have a very good eye for angles, alignment, and centering, I felt that this would be good enough. In fact, it was - as I have been imaging for the past year based on that visual collimation. However, I did not complete the collimation that first time without incident ...

Although I knew to keep each of the hex screws tightened - making only very small adjustments by untightening one side, and tightening the other - there was a stiffness in the screw threads that "faked me out" - making me believe the screws were tightened when they were actually loose. So, at one point in the collimation process, two or possibly all three of the screws became very loose ... and I totally lost alignment of my scope. Practically, that means that no light was coming out the back of the scope! At that point, there is no way to know which way to turn the knobs to regain the alignment. I removed the CCD camera, pointed the scope at the moon (brightest object I could find, as we're lucky not to have streetlights), and looked through the opening in the back of the scope. While none of the moonlight was coming directly down the focus tube of the scope (on which the primary mirror rides), I could see the bright glare of the moon off to one side (looking directly into the back end of the scope with everything - focuser, rotator, camera, etc. - removed). Once I saw the glare, I could roughly align the secondary to get the moonlight to come through the back of the scope. As I did this, I re-tightened the collimation screws so that I was ready to re-start the collimation process. The second try worked out well.

TCF-S FOCUSER ISSUE

So, in late June 2011, I decided to re-collimate the scope to minimize my star sizes and improve my images. Rather than using only my eye, I decided to try CCDInspector to provide feedback during the collimation process (which is one function of that software). I got everything set-up, started collimating, and had the collimation close ... when I decided to check focus of a star (it was out-of-focus during the collimation process), and found that my focuser was not moving! The TCF-S is a temperature controlled focuser system that can be driven by FocusMax (sofware) to automatically focus each image; it is based on a Crayford design, where the focuser tube (to which the camera package is connected) moves in and out on four roller bearings, driven by a geared stepper motor. I had seen this issue (focuser not moving) a few times previously - and, in fact, the manufacturer of the TCF-S, Optec, Inc., sent me a replacement, higher-current, power supply, thinking that I may not be supplying enough power to the focuser to move my heavy camera package (and since the same power supply also powers the Pyxis rotator).

As I had done previously when this had happened, I supported the camera (tried to "un-cock" it, which I thought might be binding the focuser tube), tried it again, removed the camera and tried it, re-started the focuser, etc. etc. ... to no avail - the focuser would not move (although I could hear the stepper motor running, I did not see the tube moving). So my collimation experience abruptly ended, pending repair of the focuser. The next day, in the light, I went into the observatory to try again to get the focuser to work. I finally got so frustrated, that I decided to remove one of the two accessible covers for the roller bearings. The scope was fairly level, but the bearings are mounted to the each side of the TCF-S, so when I removed the cover ... the entire bearing fell out in pieces! I didn't know what the bearing was supposed to look like, so I thought I had broken the focuser by having the bearing fall out into a dozen pieces on the observatory floor.

Jeff Dickerman at Optec provides absolutely fantastic service, and helped me work through the problem. He informed me that at one point Optec had bought some bearings that were prone to breaking - and I must have had one of those units. So he sent me a replacement set of bearings. As Murphy's law was still fully in effect, I received the package about a week later and it had been run over by a forklift (we could see tire tracks!), and the envelope was torn open ... with the bearings missing! So I re-contacted Optec, and they again sent out a set of bearings. These did arrive safely, and I replaced the bearings in the TCF-S (very happy I didn't need to actually send it in). Once the new bearings were installed, and the covers replaced, I tried the focuser, and it worked perfectly!

POOR COLLIMATION LAST 3 IMAGES

Now by the time the focuser was working, we were not only into late July, but again had moonless nights. I had been using the FSQ to take wide-field images, as the LX200 was down, but wanted to get back to my galaxy imaging with the LX200. So I immediately started imaging with the LX200 every clear night again - taking approximately 50 hours of data - including M16, M17, NGC6960, M13 and NGC6992 (Fish Hook part of Veil nebula). My stars were not very small, but I attributed that to bad seeing conditions (as seen on the Clear Sky Chart). However, I also noticed that the stars were not round - they were slightly oval-shaped. I originally thought this might be coma in the corners (as my CCD chip is large compared to the image circle of the LX200 scope), but realized the out-of-round stars were occurring all over the image. It wasn't until early August that I realized that I had never finished collimating the scope, after repairing the focuser!

I decided this time to use my visual centering technique, and spent the time (during the bright moon phase during early August) to get a well-centered donut image. This time, I checked the focus, and got near 2 arcsecond FWHM stars. I have now started re-imaging these objects, but may have lost some of them - as they precess through the sky with the passing of the months, eventually rising and setting too early for imaging. As of mid-August, I am now getting very small stars (typically in the range of 1.9-2.4 arcseconds, depending on seeing conditions), and very happy with the collimation.

I have left the last 3 images on the "Recent Images" page of this website, but am taking additional data which I will use as new Luminance images for these objects. I'm also now moving on to other objects with the LX200. Of course, now that I have the LX200 working again, the Andromeda nebula is rising (earlier and earlier, now around 10PM), and I'll soon want to switch to the FSQ. In any case, it is nice to know there was a reason for the fuzzy images in the past 6 weeks, and even nicer to have the scope working well, including the focuser, and what appears to be excellent collimation.

I will be collimating more often in the future, and will also try using CCDInspector again to provide a more scientific basis for the collimation. [Top]

WHY MY IMAGES STILL SUCK COMPARED TO THOSE OF THE GREAT ASTRO- IMAGERS

Although I have worked quite diligently (obsessively?) on my astro-imaging skills over the past 3 years, my images still seem quite far from the quality level of the images coming from the great contemporary astro-imagers. Some other imagers have climbed the learning curve much more rapidly than I have - for example, Rogelio Bernal Andreo, who produces stunning deep widefield images ... and only started about 3 years ago. I have the technical (optical, electronic) background for this hobby, but am still struggling with several aspects of astro-imaging that are limiting the quality of my images.

There are four particular reasons why I believe my latest images (e.g., over the past 3-6 months) are still not close to being "great" images (in order of priority):

1. I am still struggling with making good "flat-field" corrections to my raw images
2. I have continued to collect (and use) data despite non-optimal weather/sky conditions
3. My image processing skills, although improving, are still at a fairly basic level
4. My optics (at long focal lengths) are not on par with available top-end telescopes

FLAT-FIELDING

The raw images collected by the CCD camera must be "calibrated" prior to actually combining them into a viewable image. Calibration includes subtraction of "Bias" and "Dark" frames from each image (subexposure). This takes into account the pixel-by-pixel variations of the CCD, in terms of bias levels (pedestal) and dark current - which results in the CCD showing increasing noise with temperature (and can be subtracted out, if a suitable "dark frame" is taken at the same temperature and exposure time). These calibrations are relatively easy to do - collect bias and dark frames (with the camera shutter closed) at various times and temperatures, combine these into Master Bias and Master Dark frames, and then use these to calibrate each subexposure. The library of Master Bias and Dark frames does need to be updated occasionally (every 3-6 months?), as the camera characteristics can change with time; but overall, I have gotten good results with Bias and Dark frame subtraction.

However, for images which will be highly "stretched" - e.g., to bring out very dim portions of the image - it is also necessary to do a "Flat-Field" calibration - which reduces the vignetting effect of the optics (my scope image circle doesn't fully cover the CCD area), as well as reducing or eliminating "dust motes" - circles on the image which are caused by dust on the optics near the CCD (cover slip, inside of filter, outside of filter, inside of AO glass, outside of AO glass, etc.). In the case of my LX200 scope, which has a central obstruction (by virtue of being a reflector-type scope, in this case a Schmidt-Cassegrain catadioptric scope, having both a corrective glass element and secondary mirror at the front of the tube), the dust motes are dark donut-shapes on the image. The size of the donuts depends on exactly where the dust is, and the dust location can be determined by measuring the size of the dust motes in pixels, as described HERE.

The process of doing "flat-fielding" has been well described previously, such as in this OVERVIEW. Basically, you take a picture of a uniformly lit area - so that any variations due to the optics, dust, etc. can be seen (i.e., are greater than the variations in the "flat" illumination). As stretching can make very small variations look huge, it is necessary that the flat illumination is uniform to about 1% (see the Alnitak Astrosystems discussion HERE). There are several common ways of taking flat exposures - 1) during the daytime, with a T-shirt over the scope aperture (assuming the T-shirt is uniformly white and of uniform density), 2) using an electroluminescent panel (such as the Flip-Flat by Alnitak Astrosystems), or 3) by taking "sky flats" at dusk or dawn. As I use CCDAutoPilot, and it has a great capability for taking sky flats, this is how I usually make my flats.

However, only recently did I realize that one must make separate flats for each filter - Luminance, Red, Green, Blue, and Hydrogen Alpha. Not only this, but it's also necessary to make separate flats for each position angle (i.e., rotation of the camera, with respect to the scope). Up until a few months ago, I was only taking simple luminance flats, and using these to calibrate all of the subexposures (i.e., from every filter). Obviously, each of the filters will have specks of dust in different positions, and so images taken through each filter must be flat-fielded using flats taken through that specific filter. In the case of my FSQ-106EDX widefield scope, I do not (yet) have a camera rotator - such as the Astrodon "Takometer" - so the camera is rotated manually at the beginning of the session, and left in place (through the dusk flats, target images, and dawn flats). Therefore, the position angle (rotation) of the flats will match exactly to that of the images. If there is a meridian flip (i.e., imaging both to the East, and - after the transit - to the West), then the Western images will be upside down from the Eastern images ... but it is only necessary to flip the flats by 180 degrees, and the calibrations can still be made very well.

In the case of my 12" LX200R scope, however, I do use a rotator (the Pyxis) - in order to place a guidestar on the guide CCD, and also to better frame the overall image. This introduces a complication: the rotation angle must be identical between the images and the flat frames, or else a "3-dimensional" donut will appear in the final images (picture to come). This is even more difficult to correct (e.g., in Photoshop) than a single-dimensional donut. While dusk flats may be taken, and the rotation angle unchanged for the evening images, in the even I want to continue taking exposures past the transit (highest point of the target in the sky) - whcih sometimes occurs early in the evening - the mount must do its meridian flip, and the camera must be rotated 180 degrees to get the guidestar back onto the guide chip. I am finding that the Pyxis may not provide precise enough rotation to merely flip the dusk flats by 180 degrees and then use them to calibrate images after the meridian flip. A solution would be to take dusk flats (to the East), image on the East side in the evening, do the meridan flip, image on the West side in the early morning, and then do dawn flats. In this way, the evening subexposures will have exactly the same position angle as the dusk flats, and the early morning subexposures will have exactly the same position angle as the dawn flats. This works fine, if there is only one object being imaged, or if two or more objects being imaged have the same position angle. However, there is still a problem, if I want to image two or more targets, each with a different position angle (which is usually the case, as there is only so much time that an object can be "high in the sky" - and avoid obstructions, such as trees in our yard, so I often image one target earlier in the night and another target later in the night; I then repeat this for several days to collect enough data for each image).

Even if I limit my imaging to one target per night (or two with the same PA), I must take both dusk and dawn flats, comprised of exposures through L, R, G, B, and Ha filters. And, in order to remove noise from the flats, it is usually good to take 3-6 flats (per filter), and use a median (or more sophisticated SD Mask) combine. Thus, something like 15-30 flat exposures should be done each evening and morning. However, as the light is rapidly changing, it is tricky to get this many flats in one session. If it is too bright outside, the exposure time must be extremely short, which may create a non-uniform flat due to the camera shutter not providing even illumination. If it is too dark outside, the exposure time must be extremely long, which allows stars to be imaged in the flat - which ruins the "even " illumination. While the stars can be elmininated through suitable dithering (moving the scope between exposures) and using median combines, the practicality is that only about 12-15 flat exposures can be taken in one session.

I will add a page to show the result of poor flat-fielding, including not only dust motes and vignetting effects, but also terrible color gradients across the image. These are very difficult to correct - even using a program such as Russ Croman's "Gradient Exterminator" (although I have just purchased PixInsight, and hope that this very technical program may be able to do a reasonable job of correcting most of the gradients).

It should be obvious that the best way to eliminate dust motes is to eliminate the dust (!). So I have taken the camera apart, taken the filter wheel out, meticulously cleaned the filters and windows and re-assembled everything. However, as I don't have a cleanroom (or even a laminar flow bench), more dust inevitably gets on the filters during the re-assembly. I am considering sending the camera to SBIG for a cleaning - which hopefully would include cleaning the filters. However, if I want to start doing narrowband imaging (e.g., with Ha, OIII, and SII filters ... and assuming I don't buy a $1200 8-position filter wheel), I'll still have to regularly open the camera and change filters - potentially exposing them to more dust.

At this point, I am still struggling to get good flat-field corrections of my images, and feel that this is a MAJOR source of my image degradation. If there are unsightly dust donuts (especially when they overlap part of the target), I must use brute-force techniques (in Photoshop) to correct them and, ultimately, end-up reducing the brightness of the background (i.e., increasing the black point - thereby clipping data from the image) to hide the dust motes. By doing this, I am eliminating many of the faint details in the background, including galaxies, faint nebulosity, and other interesting things that make up a top-quality astro image. So, by not having yet solved the flat-field issue, I am purposely degrading my images - thinking that very noticeable dust motes are worse than losing some of the dimmest details on the image. For my images are to improve dramatically, I must solve this issue.

WEATHER ... and how selective I am about using non-optimum images

Somehow, I didn't realize that earth-based astronomical imaging is so totally dependent upon the weather conditions. Not just rain and clouds - which we expect (usually in the winter months, in Northern California) - but also high-altitude winds (the jetstream), high hazy clouds and contrails, and atmospheric stability. We have had horrible weather for astronomy for the past 3 mnonths, with huge rainstorms and high winds through much of the winter. However, having a permanent observatory, I am able to image on any particular night that happends to be clear (as I don't have a Boltwood Cloud Monitor, I usually do not start imaging on a night when I know there will be rain before morning). Over the past couple of months, there have been quite a few "clear" nights, in-between the storms. I set-up my imaging session in the late afternoon or early evening (in the evening case, I may have already missed taking dusk flats), and then hope for the best through the night. I rely heavily on the Clear Sky Clock to decide whether conditions through the night will allow imaging.

But most of the "clear" nights since the beginning of the year have had poor "transparency" and/or poor "seeing" as determined by the Clear Sky Clock. In some cases, "transparency" can mean low clouds or even fog (completely obliterating any view of the sky), but often it means high, hazy clouds - usually moving through a mostly clear sky throughout the night. While I can see these high clouds during the daytime, it is nearly impossible to detect them at night (these clouds aren't thick enough to block the light from the stars). When I set-up in the evening, I watch the focusing (usuallly automated via FocusMax and my TCF-S focuser) - and often I'm getting nice small stars (e.g., 2-2.5 arcseconds FWHM). And, when I look at the guider (AO) window, the star may seem to be quite stable (i.e., good seeing, or not much "twinkle" of the stars). However, the images made through these hazy clouds are not as high-resolution as they should or could be - like looking at a streetlight through fog , which smears the light and puts a halo around it. And, the only way to know which subexposures were taken through haze is to look at the images and measure the stellar widths (FWHM). As the clouds generally are moving through the sky all night, some subexposures may be acceptable, while others look terrible.

The solution to this problem is quite easy: just throw out the bad images (i.e., only combine the good subexposures into a final image, and discard all of the images with poor FWHM). [The "ultimate" solution for this could potentially happen in a few decades ... or longer: for amateur astro-imagers to put their telescopes into space - like the Hubble!]

The challenge I have been facing is this: if I include the bad subexposures, my images will be reduced in quality; if I discard the bad images, I'm left with very few subexposures, which also means that my final image will be degraded ... unless I wait (to produce the final image) until I have sufficient subexposures for a quality image. These days, a "great" astro-image may include 12-24 hours of subexposures. And, even on the best winter nights, I can generally only get 6-8 hours of subexposures for a single object. If half of these need to be discarded, due to hazy clouds, then I might only end-up with 3-4 hours of subexposures on those nights. Which means that I might need 4-6 nights to complete a single image. That would be no problem, if we had clear weather most of the nights; but in the winter, we have only had 1-2 clear nights per week or less. So, I have the choice of waiting for a month to hopefully collect enough data for a single "great" image ... or go ahead and process the data from the one or two partially clear nights into a final image - even though 1) I may be using non-optimal data (i.e., some of the subexposures taken through high clouds, reducing the effective resolution of the image) and/or 2) I will not have sufficient data for a 'deep' image.

If I know that I have only one or two nights to image, and then bad weather for at least the next week, I usually process the data I have - probably not discarding enough of the subexposures - and end-up with a 3-6 hour total exposure time, much of it through haze. So my images lose in both resolution and depth (which means seeing dim things, as well as having overall lower noise in the image). The advantage is that I'm gaining experience with imaging different types of objects, and continuing to develop my portfolio ... with the downside being that most of my images (at least during this bad-weather period) are not nearly as good as they could be, if we had clear skies. Of course, I could also be complaining that my skies aren't dark enough to easily capture very faint objects; that can only be corrected by moving my observatory to a dark site, or using a portable set-up and traveling to dark sites for imaging. The weather situation is one that I cannot change, but which I can mitigate by being very selective about when I image, discarding a larger percentage of the bad images, and not producing a final image until I have sufficient data.

Up to now, I have been more interested in racking up a variety of new objects (and seeing the results of my imaging sessions quickly, by doing the processing to make a final image ... and then often not wanting to go back to collect more data and start the processing over again) than patiently collecting data on very few objects to make really great images. As I am seeing the limitations of this approach, I am motivated to be more selective about when I image, and which images I use in a final composition. However, I remain frustrated by only getting sporadic nights of 'reasonable' weather when I can image, and then finding out that even those nights had relatively poor conditions that affected my image quality. Hopefully, as we approach the summer, and the weather improves, I'll have more consecutive nights to make images, and much clearer skies, without the high haze that we have seen recently.

Just to be clear: I realize that almost everyone fights the weather (except for those in the Atacama desert!) ... but my problem has been a lack of patience in waiting until I collected sufficient good data to make a great image. Given that half the month is moonlit (and unsatisfactory for taking RGB exposures), and that objects are really only in an optimum position for imaging for 2-3 months of the year, the "patient" approach could take several years to finish images of winter objects (we do have clear skies for nearly 6 months in the summer). So, I may need to delay the processing and hold back on posting new images until I have sufficient data - which will certainly result in much better, but many fewer, images being produced.

PROCESSING SKILLS

A year ago, I would have said that this was my number one challenge. But, in the meantime, I've attended the AIC (advanced imaging conferences), and attended tutorial sessions from the masters, and I feel my processing skills are improving with each new image that I process. I now make regular use of masks in Photoshop, and use many of the techniques taught at the AIC. In the past year, I have obtained and learned to use Gradient XTerminator, Noise Ninja, and Noel Carboni's Actions. Recently, I've obtained and am learning to use eXcalibrator (for color calibration based on G2V stars) and PixInsight (a very advanced technical image processing program). The fact that I'm not in a dark sky location increases the challenges of image processing, but I think I'm doing a fair job, but being mostly limited by my poor data (due to the flat-fielding issue described above). I will continue attending the AIC and learning from every tutorial I can find; please take a look at the many articles I've referenced in the Tips & Tricks section of my Links pages.

Despite my improved skills, I am nowhere near the level of the masters. As I have stated elsewhere, I have no doubt that any of the great astro-imagers could take my data and make beautiful images - far superior to what I'm currently producing. I have considered asking some of them (e.g., Neil Flemming, Adam Block, Jay Gabany, Ken Crawford, et.al.) for help on an image ... but have been hoping to first solve the "flats" issue, so that I can provide them with a full set of data (including the calibration images). In the meantime, I am trying to learn how to remove gradients in PixInsight - which will help my images, but still not get rid of the dust donuts. For the same reasons, I have refrained from posting more than one or two images on the SBIG Yahoo site - where many experts will see the images and provide advice. At this point, I think I know some of the limitations in my system, and am trying to address those prior to asking for help to get to the next level of processing skills.

BETTER QUALITY OPTICS

This issue is much lower in priority - there is always better equipment that one can use, including telescopes, cameras, and filters. My STL-11000M is a great camera, and was state-of-the-art only a few years ago, and the filters (while not being the top Astrodon series) have cuased few issues (except occasional blue halos around very bright stars). But the details (resolution) seen in my images depend not only on the sky conditions (and mount tracking, etc.), but also on the quality and design of the telescope optics. Until recently, I have been fairly happy with my (cheap) 12" LX200R scope - I think it has performed admirably, and I have obtained some decent images. But looking at these images (especially galaxies, requiring high resolution to eke out the details), they are MUCH less sharp than those from the great astro-imagers. While my FSQ-106EDX is a top-end widefield scope, used by many (most?) of the great astro-imagers, by LX200R long focal length scope - which is used for "high magnification" of small objects like galaxies - is a rather common consumer-oriented high-volume commercially-made telescope of only mediocre design and quality.

Many of the top professional telescopes and most of the great astro-imagers' long focal length scopes use improved optical designs - such as Ritchey-Chretian or corrected Dall-Kirkham - which produce higher resolution, flatter-field image circles than my Meade "Ritchey-Chretian-Like" or "Advanced Coma Free" design. These scopes produce pinpoints of light (from a star) that are typically 6-9 microns in diameter - i.e., smaller than the individual pixels of the imaging CCD; and they do this over a wide field - e.g., an image circle of 52 mm or greater, which will cover large-format CCD cameras (e.g., the KAF-16803, which is a 36x36 mm chip). The 12" LX200R is barely able to cover the STL-11000M chip (39 mm diagonal), with a great amount of vignetting being seen in the corners. My scope does do a pretty good job in terms of FWHM across the image, with stars in the corners being very similar in size to stars in the center. However, it does not seem to have the central resolution of the better scopes - which cost typically 5-10X more (for the same size aperture).

I should note here that the "relatively lower" resolution of my images could, at least in part, be due to the sky conditions (haze, seeing) as discussed above; could be due to imperfect focusing; or could be due to inadequate collimation of the scope optics. However, I have had some very clear and steady nights; my focus is often as good as I've ever seen it - using FocusMax for automated focusing; and my collimation looks reasonable (from a visual assessment of centering of the "hole" in the "donut" when the scope is inside or outside of optimum focus). Perhaps I could do a bit better collimation, and better optimize FocusMax. However, while I can often get FWHMs of 2.2-2.5 arcseconds (not bad, and probably close to my typical seeing conditions), I virtually never seen FWHMs lower than about 2 arcseconds - which is 4 pixels on the CCD. [I should also note that I am regularly getting 2.2-2.5 arcsecond stars even after 10-20 minute subexposures, which indicates that my tracking, periodic error correction, guiding, and AO operation are all performing quite well.]

When the weather improves, and we have a week or two of very clear skies, I will have more opportunity to test whether my apparent decrease in resolution (compared to the great astro-images) is due to sky conditions, focus, collimation, or inherent optical quality. My current 12" LX200R OTA is worth about $2000-2400 currently, while a Planewave CDK 12.5" scope sells new for $10K, an AGO 12.5" scope sells for $10K, a Takahashi CCA 10" scope sells for $15K and a Mewlon 12" sells for $16K, Optical Guidance Systems 10" and 12" scopes sell for $15K and $20K respectively, and an RCOS 12.5" sells for about $22K. So, "improving my optics" will probably cost $10-20K, even if I buy used.

Finally, it should be noted that many of the great astro-imagers are now using much larger scopes - for example, 16" scopes are quite typical, and some amateurs now have 20" and even 24" scopes in remote observatories in dark sky locations. Larger scopes do collect more light (and therefore allow faster exposures and/or deeper imaging) and also provide higher resolution. For example, the limiting (i.e., greatest) resolution of a telescope is approximately 138/D (Rayleigh) or 116/D (Dawes), where D is the scope aperture in millimeters. Thus, a 12" scope has a maximum resolution of about 0.38-0.45 arcseconds. This is obviously much less than typical seeing conditions, so scope resolution (based on size) should not really be a great issue; I believe, however, that the quality of the optics (e.g., accuracy of mirror curvature, Strehl ratio, etc.) does make a difference in actual imaging resolution. Therefore, an amateur with a small, but high-quality scope should be able to come close to matching those with a much larger scope - if the seeing conditions are great, and longer total exposure time is used to offset the reduced light collection of the smaller scope.

It is true in astro-imaging, as it is in other fields, that the experts can make good images with just about any equipment, while beginners can often not make good images, even with the best equipment; so equipment is usually not the 'answer'. I have no doubt that any of the great astro-imagers could take my data and produce much better images - due to their greater skills in image procssing. That is why I have not been concerned about getting a better scope at this point - I have a great mount and a great camera, and still coming up the learning curve in using the equipment I have, and especially in processing the image to get the most out of my data. However, I can see the day when I will not be satisfied with my long focal length scope, and will want something that will help me produce much better images. Whether this means a better-quality scope in my current observatory, a bigger scope (which will require a larger observatory - hopefully in a dark sky location), or even renting time on some larger scopes (and saving the capital expense), I look forward to being limited only by the weather, my processing skills, and my patience. [Top]

TRIALS AND TRIBULATIONS OF WEB SITE DEVELOPMENT

I am not a web site developer. I've written a little HTML just as an exposure, but never developed or maintained an entire web site. There were many choices to be made in the initial design of this site - primarily to use Frames (so that images could scroll without losing the menus and website title), using a dark background (to better view the images), and to generate the site myself so that I could easily improve it and maintain it. I also made many assumptions in the definition of this site; mainly, I assumed that most people looking at this site would be already interested in astro-imaging. And I assumed that most of these site visitors would have a high-resolution monitor. However, I'm now finding that all kinds of devices, browsers and resolutions are used - from iPhones, to iPads, to older small-screen computers, to ultra-high-resolution 30" displays - and my web pages look very different for many of these combinations of browsers and resolutions.

One of the design elements you will note about this site is that many of the pages (Astro-images, Equipment, Software, Observatory, etc.) use images as "buttons" to select the various content. By using large images, I am able to easily navigate this site on my iPhone. The iPhone scales everything well, but does not show both horizontal frames entirely (however, it does show the main content frame perfectly). While I am sympathetic to those site visitors with smaller displays, I also want to provide a good experience for those with large high-resolution monitors. It's not that simple...

I have been using Dreamweaver to develop the site, without any prior knowledge or training in this software. Most things are pretty self-explanatory, but now that I'm more familiar with the structure, I would do many things differently (e.g., greater use of templates, styles, etc.). I do have many plans for this site, including making all of the images into thumbnails, and having various resolutions available (to optimize the experience for all screen resolutions). However, at this point, I keep discovering seemingly simple things that should be "fixed", that are not as easy as they seem; even with online tutorials, etc., I have finally fixed several things - most of which have to do with "Frames":

                    * Having the title of each page show up in the browser window header, rather than the generic Frame title (DONE)

                    * Being able to re-load current pages (by clicking the browser reload button), rather than re-loading the Home page (DONE)

                    * Being able to bookmark individual pages, even though they are shown in Frames (DONE)

                    * Ensuring that there are no "orphan" web pages from browser searches - all pages are in the Sierra Skies Observatory frames (DONE)

In addition to these improvements, I am still struggling to make sure that this web site is viewable on most platforms with most browsers at most resolutions. At this point, my images are fixed resolutions, but put in tables which should scale with the screen resolution. However, my frames are currently fixed, and do not scale with screen resolution. Furthermore, although the images (e.g., on the home page) scale fine, I cannot seem to get the title (banner) in the top frame to scale (even when in a table). I have realized that not only should the frames be scalable (i.e., a percentage of the screen size, rather than a fixed number of pixels), but that the content inside each frame needs to also be scalable ... I think (!??!??).

Most of my development is either on my MacBookPro with a 1920x1200 resolution screen, or my 30" Apple Cinema Display, which has a 2560x1600 pixel resolution. For these screens, if I make the images, buttons, etc. too small (so they fit on a 1024x768 screen, for example), they are tiny on my higher-resolution screens. I'm still learning how to properly scale these pages and their content.

I am currently using a "Guestbook" provided by another web site (with ads - ugh!) but now have half a dozen guestbook scripts that will be hosted on my own site. However, now I need to learn how to incorporate php pages into my html pages/frames.

This web site will undoubtedly evolve, as I learn more and have time to incorporate more improvements. As with my astro-imaging, it is a "work-in-progress"!     [Top]

HOW DO CONTEMPORARY MID-LEVEL AMATEUR ASTRO-IMAGES STACK UP TO PROFESSIONAL IMAGES OF 50 YEARS AGO?

A friend of mine, looking at some of my astro-images, asked "How do your images stack up to images from, say, Palomar from the 1950s?" We had grown up together in the Los Angeles area, and both were astronomy enthusiasts who were constantly amazed and dazzled by the incredible astrophotos coming from Mt. Wilson, Mt. Palomar, Lick and the other great observatory sites. These images enthralled us in much the same way that the great Hubble Telescope images impress current generations of astronomy-aware kids and adults.

It was fairly easy to tell my friend that my images are really not very far up the learning curve, but that today's great imagers can make far better "pretty pictures" than anyone could do even a few decades ago. The revolution in digital imaging - not just CCDs, but powerful and sophisticated processing of images using Photoshop and other tools available today - has enabled people with relatively modest equipment to produce outstanding (and astounding) images.

I have collected a decent astronomy library, so I started looking in a few of my books for older astro-images. I have some astrophotography books from the 1980's that are totally outdated! (of course, nothing is obsolete, if you know how to use it - film can still provide great results ... but I would submit that images like those currently being made would not be possible if the film weren't scanned and the images digitally processed as are images from DSLRs and astronomical CCD cameras.)

The images shown by the "expert" amateur astrophotographers back in the 1980-1990 period (only 25 years ago) were amazing in their day (I can't even imagine sitting in the cold, looking through an eyepiece and guiding the scope for an hour!). However, by today's standards, these images don't look so great. One of the key things amateurs have been able to do - enabled by robotic scopes and software that allows unattended operation - is to go "deeper" in their images; that means very long exposure times, usually by stacking a large number of images that each have relatively long subexposures (5min-30min). In the old days, a 30-minute exposure was heroic; now, you must take dozens of them to see very fine details in nebulae, galaxies, etc. - details that nobody had seen until very long exposures were possible. Amateurs have evolved from photographing bright galaxies and nebulae, to imaging very dim emissions from diffuse hydrogen gas, to imaging dust! Of course, really dark skies help greatly, if dimmer objects are to be imaged.

I have a great set of books that provide images from the large telescopes of the 50's and 60's - Mount Palomar, etc. This is three-volume set of "Burnham's Celestial Handbook: An Observer's Guide to the Universe Beyond the Solar System". In looking through this set of books, many of the images are similar (in scale, framing, etc.) to those that I've been making. I decided to answer my friend with a FUN (not serious) comparison between my images and those of the great observatories 50 years ago.

To see the presentation, please follow this LINK...     [Top]

WHAT IS THE BEST WAY TO GET INTO ASTRO-IMAGING?

This depends on what your goals are - DSOs (deep space objects), wide-fields, planetary, and solar all require different and very specialized equipment. For example, DSOs require a long focal-length scope to obtain any detail; these, in turn require sensitive, cooled astronomical cameras to gather sufficient light in these (typically) slow f# scopes. Wide-field shots require a short focal-length scope with a suitable larger-chip CCD camera. Guiding errors are much less important for these wide-field shots, making in-the-field imaging and less-expensive mounts possible. Planetary imaging utilizes totally different equipment, usually based on a webcam (or higher-end high frame rate - and usually relatively low resolution - CCD), and taking tens of thousands of frames, using specialized stacking software to evaulate and stack only the sharpest frames, etc. Solar imaging is again a totally different field, requiring specialized solar filters (to protect the eyes, camera, and scope from damage) and narrow-bandwidth devices to view sunspots or solar flares, usually at Hydrogen alpha wavelengths, but also at other wavelengths.

I will assume here that the reader is interested in deep space or wide-field imaging of the type seen on this web site (or, preferably, on the web sites of the masters - many of whom can be found in my Links page: Astro-Imagers). This still leaves two possible directions: DSO imaging, using longer focal lengths, or wide-field imaging, using shorter focal lengths. In both cases, there are both inexpensive and expensive options, in terms of scopes, cameras, and accessories.

DSO imaging (galaxies and nebulae) can be done with any scope, but in order to get some detail with these small objects, it helps to have a longer focal length system - for example, 2000mm or more. This focal length happens to be what a Celestron or Meade 8" scope with f/10 SCT optics will provide. Of course, larger scopes provide much greater light-gathering power, so make it easier to image faint objects. As the light collection (assuming this all gets to the CCD chip) goes up by the square of the aperture, a 10" scope will have about 56% more light gathering capability than an 8" (0.2m) scope; and a 12" (0.3m) scope has about 2.25X (i.e., 225%) the light-gathering power. You can see why the great imagers often use 16" (0.4m) and even 20" (0.5m) scopes - they give about 4X and 6X the light-gathering power of an 8" scope, respectively (and about 2X and 4X compared to a 10" scope).

The nice thing about starting with a Celestron or Meade scope is that they can be purchased as a complete system - i.e., integrated with a fork mount - which makes it easier to set up the scope. The newer models even include GPS receivers, electronic compasses and levels, and other "features" to more automatically align the scope. This makes is easeier, but perhaps the user doesn't learn as much - and eventually, you really must learn to accurately polar align the scope, improve the pointing and tracking of the mount, and understand the "whys and wherefors" of basic telescope usage, before any good images can be made. I should note here that if you use a fork mount, you must place the mount/scope on a wedge, so that it is equatorially aligned (not alt-az); yes, I know there are field de-rotators, etc. which allow use of the scope mounted directly to a tripod - but these do not really enable high-end imaging, which requires long exposure times and minimal star trailing.

However, there are a few problems challenges in using a longer focal-length scope. And if you're talking about a 10" or 12" Celestron or Meade, you're in the 2500mm (100") to 3000mm (120") range of focal lengths. These problems include tracking errors, and even finding and centering the star on your CCD chip.

Longer focal lengths decrease the "plate scale" - i.e., the number of arcseconds per pixel in the image. The formula for plate scale is:

                         Plate scale (arcseconds per pixel) = 206 * CameraPixelSize (microns) / FocalLength (mm)

CCD chips these days have pixels that range typically from about 5.4 microns (8300 chip) to 9 microns (11000 chip) - for unbinned images. Thus, a 500mm scope might have a plate scale of 2.2-3.7 arcseconds per pixel. And, on a night with great seeing and a well-focused image, you might have a FWHM (full width at half maximum) - i.e., star diameter - of 2 arcseconds. So there is a reason why short focal length scopes can produce "pinpoint" stars: the star images are smaller than 1 pixel on the CCD camera!

In contrast, a scope with 3000mm focal length will have a plate scale (depending on the pixel size on the CCD) of 0.4-0.6 arcseconds, or 6 times smaller than for the 500mm scope. Thus, any tracking errors (which make "lines" instead of "points" over a long exposure) of more than 0.5 arcseconds during the exposure will spoil the image. There is a handy calculator that can determine required polar alignment accuracy, inputting various parameters, including the "field rotation" figure in microns (i.e., length of star trails):

                         http://celestialwonders.com/tools/polarMaxErrorCalc.html        (with credit to Frank Barrett)

For a 5-10 minute exposure (typical in astro-imaging, if not longer) with a 3000mm scope, this means that the mount must be polar aligned to about 2 arcminutes. For the same exposure time with a 500mm scope, one would only have to polar align to about 12 arcminutes (i.e., 6X more tolerance). This is much easier to do, and would be far less frustrating for beginners.

An even more basic issue is finding the DSO on your CCD chip. The FOV (field-of-view) of a camera (CCD chip) is just the plate scale (arcseconds per pixel) times the number of pixels on the chip. Most beginning imagers use less-expensive CCD chips, which are usually quite small - for example, the Meade DSI II camera (available in monochrome or color) has an 8mm diagonal chip. This 4.7x5.6mm chip (with 8.3-8.6 micron pixels in a 752x582 array), with a 500mm scope gives an FOV of roughly 34x44 arcminutes - i.e., about half a degree. This is a large area and makes it relatively easy to find your object on the chip (i.e., pointing accuracy should be about 10-15 arcminutes or better). However, for a 3000mm scope, the FOV shrinks to 5.6x7.3 arcminutes (!); this is a very small area of sky, and pointing accuracy would need to be better than about 2-3 arcminutes to ensure that your object will land on the chip.

So, it is much easier to image with a wide-field scope in terms of both pointing accuracy and tracking accuracy (polar alignment), compared to a long focal-length scope.

I haven't mentioned one other issue that affecs tracking accuracy: periodic error. This is the error in the gears that are evident in the slight 'wobbling' of stars in the image over the period of the RA main gear rotation (about 8 minutes for the Meade LX200 mounts). The periodic error of the Meade LX200 is about 30 arcseconds (peak-to-peak), but can be corrected to about 5-6 arcseconds with periodic error correction (PEC), which is built into the scope - but must be analyzed and loaded using software such as PemPro. If we assume this kind of periodic error (e.g., if a small refractor was mounted piggyback on the LX200, so it used the same mount), then our refractor images would barely notice the error, as the plate scale - or size of each pixel in sky FOV - is about 4 arcseconds, so the total 'broadening' of the FWHM by the periodic error woudl amount to only a fraction of a pixel. However, for the 3000mm scope, the plate scale is 0.5 arcseconds, so a periodic error of 5 arcseconds corresponds to "smearing" of the image over 10 pixels. With a small chip, this can be a major fraction of the image size (e.g., 10 pixels/582 pixel image height = 1.7% of the image). This typicallyl does not make a "pretty" picture.

So if the beginning imager is interested in both DSO and wide-field imaging, I would recommend they start with the easier wide-field set-up. This also makes camera selection much easier, as even small CCD chips will have a 'resonable' field-of-view - i.e., enough sky area to encompass important objects (e.g., Andromeda galaxy, Pleiades cluster).

If your only interest is in DSO imaging, then you will need to have to ensure that your mount can meet the specifications set above. The Celestron and Meade scopes barely do, and these mounts can be used for DSO imaging, as in my Recent Images 1, most of which were taken with a Meade LX200 scope on its fork mount. One of the ways to really mimimize the tracking error effect on the image is to use an adaptive optics accessory, such as the AO-8 (for the SBIG ST-series cameras). This has a tip-tilt element that can correct quick or slow changes in star position many times per second, and can essentially remove most of the remaining periodic error. While these systems are primarily intended to reduce star sizes by attempting to track the "twinkling" of the star - i.e., slight changes in position due to changes in refraction by the atmosphere, AO systems can also minimize tracking errors (i.e., polar alignment errors) in the mount. [this does not entirely correct for polar alignment errors on long exposures, due to rotation of the entire image which cannot be corrected by a simple tip-tilt system]

However, in using an AO accessory, we now introduce a new problem with longer focal length scopes: we need a bright guide star to allow fast exposure times to enable quick corrections using the AO system ... but the FOV of a long focal length scope includes only a small area of sky, and there may not be a guide star in the correct position (when your main target is framed). If using the SBIG cameras, the guide chip is typically the TC-237 with 657x495 pixel array - thus providing a 4x5 arcminute FOV (with 9 micron CCD pixels). It is rather unlikely that you will find a bright enough (Mag 6-9) star in this small field-of view. One solution is to add a camera rotator that can turn the camera so that the guide chip is at the position angle of a bright star. This adds expense and complication to the system. Another approach is to use a guide scope for guiding; this will have a short focal length and provide many bright stars on the chip for normal guiding, but may have too short a focal length to accurately guide the longer focal length scope. Also, it is well known that there can be "differential flexure" between the guide scope and the main (imaging) scope, which makes guiding with the guide scope poor for imaging with the longer focal length scope (as the scopes move around the sky on the mount, the loading may flex the connections between the scopes sufficienty that the guide scope is partially "correcting" the flexure, thereby causing guiding errors with the larger scope).

So now, if the beginning imager wants to image DSOs, the my suggestion is to add a camera rotator and AO system to the scope and camera. The rotator may be a manual unit (e.g., "Virtual View" for LX200) or motorized (e.g., Pyxis by Optec, Inc.). However, the system is now becoming much more complex (i.e., more things to go wrong; not the right direction for a beginner).

And, as is often said - correctly - the mount is really the most important part of the system. We've already discussed pointing, tracking and periodic errors, and these can all be minimized by using a better mount. However, I'll leave discussion of mounts for a later blog ...

One question that I've asked: can someone "buy" their way in to becomming a great imager? In a word, NO! As we've discussed, good equipment really helps. A good mount will minimize pointing and tracking errors, making images better; a good astronomical camera will have regulated cooling (for better dark frame subtraction), have a sensitive CCD, and a reasonable size chip; and a good scope, of course, may provide better (higher resolution) images - although I think that is the last thing a beginner need worry about. So, with a wide-field set-up, I think someone could buy a great mount (Takahashi, AstroPhysics, Paramount), put a scope on it (ranging from a small APO refractor such as the WO Megrez 90 to a high-end imaging refractor such as the FSQ-106ED), and buy a good CCD camera (e.g., one of the ST- or STL-series cameras from SBIG) buy the best software, and begin making pretty good images. However, the imager must still obtain good polar alignment, good focus (to be discussed in a later blog), and do good processing for the images to go from "pretty good" to "great". Image processing alone has a great deal to do with the quality of the final image.

Finally, I will mention some of the camera choices. As mentioned earlier, the Meade DSI II and DSI III cameras are inexpensive and come with simple software, but have small chips, are not cooled (not to mention regulated), and have no guide chip (only available on SBIG cameras, but a great help to beginning imagers). Other than SBIG, manufacturers of good astronomical CCD cameras include Apogee, FLI, Starlight Instruments, and others. They all use the same chips - usually from Kodak (or Sony), but incorporate different features, such as small size, regulated cooling, guider output, AO accessories, etc. All of these can be purchased used, via a reputable web site such as Astromart. A larger chip camera will make it easier to put the target on the chip (especially with a long focal length system) and will provide a wider FOV - often needed with a long focal length system to include all of the target object. However, larger chip cameras are more expensive than smaller chip cameras.

As CCD chip prices plummet (for given chip size), we are seeing really great cameras being offered at incredible prices (incredible, at least, compared to the "old days" of a year or two ago!). For example, the current crop of cameras based on the KAF-8300 chip from Kodak, are incredible for wide-field imaging with short focal length scopes; their small pixel size (5.4 microns) provides a wide FOV, and the large number of pixels (8.3 megapixels in a 3326x2504 array) provides both wide FOV and high resolution (lower plate scale, vs. the larger pixel cameras). These cameras are selling new for as little as $2000 (which I know may be too much for a beginning imager, but gives an example of future CCD cameras). The ST-2000 (monochrome or color) can be purchased used as low as $1500 - but, of course, the monochrome camera would need a filter wheel and filters in order to make a color image. Used cameras of interest may include the QHY-8, the SXVF-H9, and other similar models.

The only other major decision regarding the camera is whether to buy a one-shot color camera or monochrome camera. If you are interested in making color pictures, the one-shot color (OSC) camera is much simpler to start-off with, as you don't need a filter wheel or filters, and you don't have to stack frames to see a color picture. They can take great images - limited by the same things as better cameras: focus, exposure time, and optical system quality. However, the OSC cameras are not as sensitive (only 1 in 4 pixels detects red light, which is the predominant color of emission nebulae), they don't provide as high resolution (as the pixels are divided between color filters, and then the data is interpolated back to the stated resolution), and they don't teach you quite as much about the image processing steps as a monochrome camera. I think it is a very valid way to start imaging, and have quick satisfaction of seeing a beautiful color image (e..g, first image of the Orion nebula). One can then move up to a monochrome camera, filter wheel, etc. when ready - to obtain better sensitivity, higher resolution, and enable use of narrowband filters.

I hope this brief discussion of how to start out in astro-imaging is hepful to those reading this. There are many great books on the subject than will be helpful, including those by Michael Covington and Ron Wodowski (I'll insert links later). Basically, regardless of the equipment you have, astro-imaging is a time-intensive, detail-oriented, sleep-depriving, and frustrating hobby - just getting everything to work together (and at the same time). However, it is very satisfying to see images being produced of objects out in space (sometimes millions of lightyears away) taken in your own backyard. If you are interested in some of the equipment aspects (rotators, cameras, adaptive optics, etc.), please take a look at my Equipment pages.     [Top]

THE MEANING OF IT ALL ... WHY AM I DOING THIS, AND WHAT ARE MY GOALS?

Basically, I have a general interest in science, a deep interest in astronomy, and a lifelong enthusiasm about system engineering - i.e., getting a bunch of complex stuff to work together and produce a successful result. While one of my "goals" with astro-imaging is to "make pretty pictures" (discussed below), another is just to successfully get everything working together and autonomously. You must remember that I come from a time (NOW, I feel old!) when there were no computers, no CCD cameras, no automation, etc. ... it is VERY exciting to me to pour a beer, sit at my desk in the office, and watch CCDAutoPilot go through all of its steps, finding an object, centering it, focusing, selecting filters, finding the guidestar, starting the adaptive optics, and making exposures - all automatically! That is a big part of the satisfaction to me.

I have been asked by friends why I don't just rent time on great telescopes (e.g.,. via LightBuckets, or others). Well, although I could get great data that way, my role would be reduced (!?!?) to image processing ... which is something very important, that I am learning, but which is not my main interest. On the other hand, perhaps I should provide my data to some of the great astro-imagers, and let them do the advanced processing to make the pretty picture. But with my present set-up, I am trying to do it all: build the system (albeit using commercially available parts, not building it from scratch, as I would have done 30-40 years ago), get it all to work together, solve the multitude of little problems that affect the image or preclude remote operation, and finally to collect and process the data into a result that the average (non-astronomer) person can appreciate.

Perhaps, now that I've built the current system and observatory and gotten some experience, I might do it differently next time ... but I don't think so. I think the experience gained so-far will allow me to build a great remote observatory at some point. Still, doing the building, system integration, and solving the problems is most of the fun for me; I'm sure as I get better at image processing, that will also become more 'fun'.

Ultimately, the goal is to make pretty pictures - basically, a cross between science and art. This has always been important for me; many people don't realize how creative science is and can be, and astro-imaging is taking creativity to one extreme (another extreme may be the creativity to visualize the structure of spacetime, which theoretical physicists do ...). I am not doing science, although I may very well define some scientific projects in the future - or in collaboration with "real" scientists (as Ken Crawford, Jay GaBany and a few other amateurs are doing). However, there is satisfaction in solving the engineering problems, and getting the entire system to work. Some amateurs have ended-up developing new products (hardware or software) based on their experiences solving such problems, and I can imagine doing the same, at some point.

Regarding specific goals, I would like to come up in the ranks of asro-imagers to the point where my work is recognized - such as through the NASA APOD, and in astro-imaging meetings. Most of my work to-date has been 'private'; however, I intend to begin participating more in the Yahoo groups (e.g., SBIG group) and at the various astro-imaging meetings. I am hoping that with continued effort, learning, improvement, and persistence, that I can become a very good astro-imager.

Finally, this question - why do astro-imaging? - has been discussed at the AIC and other meetings recently. A key goal for many is communicating the wonders of the universe to the layperson - through beautiful pictures that show the amazing variety of objects in space, and the incredible beauty of many of therm. In this sense, astro-imaging is a form of public outreach; whether images are published in popular magazines, or just shared with friends and neighbors, it is often an eye-opening (if not eye-popping) experience for many people who don't realize "what's up there", let alone that such images can be made from your backyard.

There are many other aspects of astro-imaging that I look forward to learning: using astronomical cameras to image through Nikon lenses; doing planetary (high speed video) imaging; doing imaging in the field (have done some, but don't have a current "system" to do it reliably); additional wide-field imaging including mosaics; narrowband imaging (I've only done Ha for the past 3 months since I bought the STL-11K, but still don't have OIII or SII filters); and possibly development work in some specific areas - using image intensifiers, developing "real" (deformable mirror phase control) adaptive optics, etc.

In conclusion, astro-imaging to me is just another hobby - but one which takes an incredible amount of focus, persistence, and learning of new techniques to master. It is a hobby that satisfies my need to do some systems engineering and "lab work", but which also results in images which can be shared with people who don't know the first thing about astronomy. And these images open up vistas of new experience for the people seeing them - the realization of just how big the universe is, how many crazy things are in the sky, and how it is possible for amateurs these days to do imaging that eluded even professionals just a decade ago. I think it's an incredible combination of science, engineering, and art, and a very worthwhile way to spend 'extra' time on something that is both educational and results in pretty pictures and, potentially, making images that show things that have never been seen before.      [Top]

HOW DID I GET INTO THIS IN THE FIRST PLACE?

I have always been interested in science, spending much more time as a kid building things and doing experiments than playing sports or getting together with friends socially. For example, I was about 10 or 11 years old when I built a "planetarium" using my model train, with various flashlights and things mounted on the cars to project on the ceiling. By 12, I had found out what it was like to get shocked with 110VAC (I was trying to put aluminum foil inside a lamp as a reflector, and it touched the lamp contact at the base). In my teens, I became interested even more in astronomy, but also in radio and electronics. At that time, some of my friends were building telescopes (generally from the Edmund Scientific catalog, but also grinding their own mirrors, etc.), and I had to decide whether to follow that path, or spend time and money on electronics. I ended-up getting my "ham" radio license (WB6OOV), and spending much more time with radio and electronics. We didn't know so much about "safety" then ... so I did many dangerous things. Among these were handling a large amount of liquid mercury with my hands; suspending a large ball bearing on a wire from my bedroom ceiling and hooking it to a (surplus) neon sign transformer - then swinging the ball, and watching it draw an arc of 20KV, with plenty production of ozone in my bedroom; lighting magnesium in the backyard, so that a friend across the [San Fernando] valley could find my location so that we could try laser communications; exciting the laser with my ham transmitter (and being exposed to huge amounts of RF that was not being coupled into the laser; and - much later, at the Hughes Research Labs (now HRL Labs), tearing asbestos sheet to use pieces as a beam block for high power lasers (and seeing the asbestos smoke rising into the lab).

In High School, a friend of mine was building a cyclotron (!), and getting many parts donated by big companies. He needed a laser to line things up, and there had been a recent Scientific American "amateur scientist" article on how to build a laser, so I began designing optical benches and my eventual (HeNe) laser. I wasn't very good at machining (I don't have enough patience), so our machine shop teacher assigned the optical bench project to his classes. I found an expert optician who ground the mirrors for free - but he was an alcoholic, and it took many months to get the mirrors finished. The coatings were donated by OCLI, and the glassblowing of the tube and filling of gas mix was done by a local vendor - also for free. (I had bought glassblowing equipment, and started learning, but never would have been able to pull the vacuum and fill the gases at home) I tried many things, such as holography and laser communications - when I was 16-17 years old. I tried to get a summer job in the field, and was actually offered a position at TRW (where the famous picture of a bullet and its shockwaves was made), but it wouldn't start until the end of the summer; I was offered a job at Electro-Optical Systems (EOS), but they didn't have money to actually pay me. So, I looked in the yellow pages, under "research", and got a job with the first company listed - Admerco, a small testing laboratory, where I did electronic and electrical testing and had many interesting experiences (for example, getting a huge multi-rack NASA "time and event" recorder to work, without schematics or a manual; in the process, I connected wires to a 100A power supply, shorting a circuit, and having the wire melt while I was holding it!).

In 1985, the Smithsonian Insitution held a traveling exhibit "The laser at 25: 25th anniversary of the laser". At this time, I was already working in the medical laser field, and was asked by Bill Bridges (among other things, inventor of the argon and krypton ion lasers, and professor at Cal Tech) if I had any medical lasers I could donate to the exhibit. As I didn't have anything at that time, I offered to donate the laser I had built in high school, as an example of what students were doing in the early days of lasers. Note that I built my HeNe laser in 1965, and a HeNe laser had only been developed in mid-1962 (the first laser was operated at the Hughes Research Labs by Ted Maiman and others in 1960). I felt that I was already too late, and missed the field! In any case, the Smithsonian accepted my offer, and my little high school HeNe laser was part of the exhibit that traveled around the U.S. As a cap to this story, I was stranded in Washington, D.C. during the week after 9/11 (I was the last plane to land at Washington National at 9:15AM on 9/11), and in touring the Smithsonian remembered my laser; I was able to find out that it was in the National Museum of American History, called the contact there, and got a tour of the "back rooms" as well as a certificate thanking me for the donation.

When I got married in 1971 (very young), my interest in astronomy was re-kindled, and I became a Celestron dealer - just to support the hobby (!). I ended-up with a Celestron 8 and a C5 telephoto lens (which I still have). I thought my wife would be interested in going to the mountains with me to do observing ... but after one experience with an extremely cold night (and not having the clothes for it), she decided she didn't like astronomy. I did some observing from the backyard, and learned a few things - like polar alignment and finding objects (remember, there were no "go to" systems in those days. I even tried some astrophotography by hanging my Nikon F2 off the back of the scope, but never got it balanced well enough or aligned well enough to track for any length of time ... so became disillusioned and pretty much gave up. I eventually sold the scope when we moved to San Diego in the late 1970's.

In the meantime, I worked at the Hughes Research Labs (1972-1978), and found out about an opening for an instructor to teach a course for the Santa Monica Community College outpost in Malibu (where HRL was located). I got my California Community College instructors credentials, and got the job of teaching astronomy in a night course in Malibu. However, I had never taken a class in astronomy, and really knew very little about stellar formation and evolution, etc. So, I bought lots of books, and started reading and learning astronomy on my own - so I could teach it. Of course, you never learn anything better than when you try to teach it, so I eventually learned quite a bit - about the history of astronomy (I gave lectures on the ancient's interpretation of the heavens), up to recent space missions and interesting data on neutron stars and black holes.

After Malibu, we moved to San Diego (to work for the "Industrial Products Division" of Hughes Aircraft company), and I sold my C8. After 10 years, we moved to Hawaii (Hanalei, Kauai), and had incredible dark skies ... but only dreamed about astronomy (and doing a little observing with binoculars), but did not buy a telescope. Finally, in 2003, we moved back to California - this time to northern California in the foothills of the Sierra Nevada mountains. While I'm still at the edge of the light dome of Sacramento, our skies are fairly dark (around 19-20 mag/arc-sec^2), and I finally decided to get back into the hobby

In late 2007, I bought a Meade 10" LX200 on eBay (I wasn't yet aware of Astromart). After all this time, I thought I knew how to set-up the scope, but it took some re-training to get the hang of it again. I started with alt-az mounting (on a Meade Giant Field Tripod that I bought separately), and finally got a Meade UltraWedge and learning how do accurately polar align the scope. In December 2007, I bought a Meade DSI Pro II and also DSI Pro II Color, and began trying to do astro-imaging. I may decide to insert a picture or two from those early days (2.5 years ago), but am much too embarassed to do that now. I was struggling to get a good 10 second or 20 second exposure, and even to get an object to actually show up on the small CCD chip. However, I did get some small results, so was motivated to continue on.

In early 2008, I found a 12" LX200R (ACF) on Astromart and bought it, selling my 10" LX200 (and beginning the long process of buying and selling things, until I really knew what direction I was going, and what I needed in terms of equipment). Unfortunately, when the 12" scope arrived, I heard tinkling of glass ... Yep! The corrector plate had been cracked into pieces - evidently by Fed Ex standing the box on end, and putting something really heavy (barbell weights?) on it, pressing down on the box, styrofoam, and aluminum cap, and onto the secondary, which finally broke the corrector. Well, it was a long process to send it back, get Fed Ex to agree to pay for the damage, get it shipped to Meade, and then a couple of months waiting for Meade to replace the entire optics set (the corrector is matched to a primary mirror, so both have to be replaced, if either is damaged). However, finally by April 2008, I had the scope back, and starting to set it up in the driveway.

With the 12" LX200 package, I got some JMI Wheely Bars, so had the scope mounted on the UltraWedge, on the Giant Field Tripod, on the Wheely Bars. I could then roll everything from the garage out to the driveway, where I put the Wheely Bar screws onto marks I made on the driveway. I also tried mounting a small laser on the arms of the LX200, and put alignment marks on a neighbor's house (! - but on masking tape, so it could be removed) ... but found that the large beam divergence of the little diode laser pointer was much too large to get good accuracy that way. I finally ended-up designing a better mounting scheme, where I replaced the Wheely Bar screws with 2" aluminum bar stock, machined to sit on a 1" ball bearing which, in turn, sits on a low-profile matching 2" aluminum piece epoxied to the driveway (see Observatory, Driveway mounting). This allowed me to roll out the scope, plop it onto the ball bearings, and get alignment of about 5 arcminutes - with no manual alignment necessary.

In spring 2008 (before the 12" LX200 had even come back from Meade), I bought an SBIG camera - from the same guy from whom I bought the scope; this was a major turning point in my astro-imaging experience. I finally got tPoint working, and the scope centering by merely pushing a button. I learned CCDops and CCDsoft, and bought MaximDL (which is what I mainly use to control the camera). The camera I bought was an ST-4000XCM, which had a high resolution (for that time) CCD chip (2000x2000, 4 megapixels), but was a one-shot color camera - i.e., using a Bayer matrix of color filters over the CCD chip, so that a single exposure could yield an RGB image. While the actual time taken for imaging is similar to a monochrome camera with filters, the time for processing (and mental block over stacking images, etc.) was much easier. I still feel that this is a very reasonable way for someone new in astro-imaging to go: the SBIG camera has an internal guider, and one can learn SBIG software (and MaximDL, etc.), but without some of the bother of stacking images and using filters. Eventually, this became a limitation in my imaging - the ST-4K is not a very sensitive camera and, in the red especially, is extremely unsensitive compared to a monochrome camera (think that in imaging red nebula, which is most common, only one pixel in four actually is receiving light; the rest are only producing noise!). Also, a monochrome camera is higher resolution, as the 2Kx2K resolution of the ST-4K is really split between the different colors.

The ST-4K also came with an AO-8 adaptive optics system. This was an incredible help to me, as I was still struggling to get good polar alignment, and the tracking ability of the LX200 is OK, but not terrific. However, using the AO-8, I eventually also learned that you cannot easily find bright enough guide star - at the long focal length (3000-3200mm) of the 12" LX200 scope. So, I bought a Pyxis rotator. This confused me for a few months regarding Pa and rotation angle (which seemed to be opposite in MaximDL vs. CCDSoft), and actually getting guidestars placed on the guide chip.

During this time, I had my ultimate goal set on doing "remote" imaging (even if "remote" means controlling the system from the house). So I automated the focusing of the 12" LX200 with a RoboFocus, and started learning about FocusMax. Focusing using the primary of the LX200 was not perfect - even replacing the Meade focuser with a Feathertouch 10:1 focuser did not elminate all of the backlash - but it worked fairly well, and allowed me to get consistently better focus than I could get manually. I also learned how to collimate the scope - with some adventures when I replaced the secondary adjustments with "Bob's Knobs", and ended-up completely decollimating the scope! However, I eventually began making some decent images - at least they were focused pretty well. However, most of the time, I was making only single-exposure 5min-20min images.

Throughout 2008, I spent more than 100 nights outside with the scope - and it was really COLD in the winter! We have always lived in southern California or Hawaii, and were not used to near-freezing temperatures; I spent many nights in my ski outfit, with long underwear and a ski cap, sitting by the scope, centering it with the Autostar controls, etc. I had gotten pretty good aiming capability with just the LX200, but decided to use tPoint modeling to improve the pointing accuracy. However, I really didn't understand tPoint, and did not critically look at the details ... I just mapped a bunch of stars (probably 20-30), and found that it gave pretty good all-sky pointing, and stopped there. What I DIDN'T learn, until I bought the Paramount, is that my model was not using the correct terms (I had just picked terms at random, until the pointing accuracy got a little better). The result of this is that while the gross pointing was OK, I could never click (e.g., in MaximDL) on "Center", and get the object to be finely centered on my CCD chip; in fact, when I hit "Center", there was a good chance the object would move AWAY from the center of the field! As I just wanted to image, and was doing it manually (not with ACP or CCDAutoPilot), the centering didn't seem to be an issue - I just sat there and corrected the position using the Autostar controller, sometimes taking 5-6 iterations to get the object centered!

Also throughout 2008, I bought a lot of junk - stuff I thought I would need, but learned that either I didn't need it, or that it wouldn't work with my imaging set-up. For example, Nelson connectors (really neat for quickly changing optical train, but not stable enough for use with a heavy SBIG camera); complete sets of Nagler and Panoptic eyepieces - which I never used, as nobody wanted to sit out with me at night observing, and I was doing exclusively astro-imaging - so in early 2010 I sold most of these eyepieces); a Megrez 90 refractor, that allowed me to dabble in wide-field imaging (although the distortions at the edges of my 15x15mm CCD chip were horrendous, so I pretty much gave up imaging with it); various filters (Meade and Orion light pollution filters, and eventually an IDAS LPS2 filter - which I finally decided was not needed, at least if I imaged towards the East, over the Sierras; and many many more things. Basically, I found that astronomy could be just like owning a boat: a black hole into which you can throw lots of money!

However, in the firsr 1-1.5 years, I did begin at least focusing on where I should spend money. Back from my C8 days, I realized that the very most important thing is the mount - and I was coming up against the limitations of the LX200 mount (although I had been surpirsed how good it is - especially for the price). So, in spring 2009, I bit the bullet, and bought (used, from Astromart) a Paramount ME - which is the premier mount for remote imaging. I de-forked my LX200, and put the OTA on the Paramount and sold the LX200 mount itself. While I thought I would have to design an adapter plate that would mount the Paramount on the Meade Giant Field tripod, I saw a picture of exactly such a mounting; I advertised in Astromart for something like that ... and the person who owned that exact adapter plate (I had put an image of it in my a'mart ad) replied, and was willing to sell me his. So, I ended-up with the LX200R OTA mounted on the Paramount, on the GFT, on the Wheely Bars that I had modified for driveway mounting. My imaging results took another great improvement.

While I could roll the scope system out to the driveway and get it set-up pretty quickly, and while I would leave it out on the driveway (the garage is well above the street level, and not very visible) covered with a beach umbrella during the day, it would still take a couple of hours to really get well aligned each time I rolled it out ... and I would always be afraid of rain, or even fog rolling in, while the scope was sitting in the driveway. It was actually a very useable system for someone who wanted to take "snapshots" - i.e., exposures of 20 minutes to an hours, and several different objects in a night. However, I was really interested in high-end astro-imaging, where total exposures of 12 hours or more were needed. As this requires exposures over several nights - with the scope coming back to the same field repeatedly, the driveway situation was just not appropriate. Also I travel internationally constantly in my business, and I never seemed to be home to do imaging ... except when the full moon was out, or we had clouds!

So, I finally decided to put in a real observatory. I studied the various models, and had pretty much decided on the Technical Innnovations ProDome (ideally, 10', but practically 6'). I nearly bought one of the 6' tall ProDome systems, but finally decided I wasn't ready (I didn't even have a place to put it, other than in the driveway, which would only be temporary). Then, I started seeing ads for the "Gecko Observatory", which had been designed and built by Troy Duval in Dallas. Troy had graduated to an RCOS scope and a remote observatory away from Dallas (for which he selected a clamshell and built the bulding). He originally wanted way too much, and the issue was how I would get it from Dallas to Sacramento. To make a long story short, he came down on the price, and eventually agreed to drive it here in a Penske rental truck! I built a deck in part of our yard not being used (and had to cut down some oak trees to make a "hole in the sky"), and Troy helped me re-assemble the observatory here in El Dorado Hills, in March 2010.

Once I had the observatory, my addiction didn't stop ... so in May 2010 I finally bought a monochrome camera with filter wheel - Randy Nulman's STL-11000M. This again propelled me in the astro-imaging arena, learning how to stack images, create color, RGB vs. LRGB, etc., along with a wider field of view than I had previously achieved. At this point, the weakest link in my system was the Megrez 90, which I wasn't using for imaging - if there was distortion at the edges of a 15mm chip, think how bad it would be at the edge of a 42mm chip! So I sold the Megrez, and bought (again, via Astromart) a FSQ-106ED, the premier widefield refractor (for that diameter scope). Just the Paramount, FSQ and STL-11K forms the core of a really 'professional' amateur astro-imaging system. The LX00R is not bad either - not an RCOS, but fairly respectable curvature and distortion, even on the STL chip. So I am now a "happy camper", just trying to get enough exposure time with some objects to make 'pretty images'. You can see how I'm doing on the Recent Images pages.

Everything is now controlled remotely, via a Mac or even an iPhone; I have a 30" display that shows virtually everything at one time - TheSky planetarium program, dome control, MaximDL camera control, FocusMax, CCDAutoPilot, a web browser showing the webcam images and power controller, and the weather station software. Most nights, I can program CCDAutoPilot and go to bed, ending up with a folder of images in the morning. Then, the process of reducing (calibrating) the images, stacking them, and turning into a color image which can then be edited in Photoshop. The work flow is coming together, although I'm still struggling with good flat-fielding and a few other smaller points.

Finally, what are the limitations of my current observatory set-up? Primarily 1) that I only can see the sky to the South (from East to West), and MANY interesting objects are too far to the North to image; 2) that I don't really have black skies - not bad for near a city, but certainly not what I would see 30-40 miles farther up the highway into the Sierras; and 3) the size of the observatory is perfect for my current set-up, but would not fit a much bigger scope (e.g., RCOS 16" or similar). However, for the part of the sky I can see, it's not a bad set-up. At each stage of this process, I've used equipment until I bumped up against various limitations; I'll do that now with my observatory set-up, learning, getting better and, perhaps eventually, having a REAL remote observatory at a dark sky site.    [Top]

BEGINNING OF THE BLOG ...

Although I re-started astronomy as a hobby - and astro-imaging as my focus - nearly 3 years ago, I have only recently begun building this web site, and only now starting to write a blog. This is unfortunate, as I have gone through many challenges over the past 3 years that should have been well-documented in this blog. In any case, I will begin by presenting a background on my history with astronomy and experience so-far with astro-imaging. I will also comment on my view of astro-imaging as a hobby, and some of the possible goals for this hobby. Finally, I intend to begin documenting the successes and challenges of astro-imaging as they arise in the future. I hope whoever is reading this will enjoy the commentary, and learn something useful from my many experiences.      [Top]