The Starry Night, 84

:: home ::




              <<   41   42   43   44   45   46   47   48   49   50   51   52   53   54   55   56   57   58   59   60c   60   61   62
              63   64   65   66   67   68   69   70   71   72   73   74   75   76   77   78   79   80   81   82   83   84   85   >>   SRCH

1/29/2012. M81 and M82 emerge from Birdfeeder Tree just before midnight, and I've been waiting for them with the AT10RC telecompressed to F5.1. I've been doing the first focus run of the night in Nebulosity, then I image and guide with the ST2000XM in dual-chip mode using Maxim. Bin the guider 2x2, autostretch its images using the "low" or "medium" preset, calibrate in the vicinity of the subject, compose, find a guide star, designate it, select "track" and start. Give it a few cycles to settle down. Then under the "expose" tab, select "autosave," customize the filename, the sequence of exposures, the length of the exposures, "apply" or "OK" and "start." Fidget some as the camera reports Filter Wheel errors (if the filter to be used is already in place then for some reason I get these errors -- when they come along, just go to "locate DSO," select a different filter, and reinvoke the sequence using the autosave button).

Recent tracking plots have been just about dead flat: 0.4 pixels RMS in X and Y. Half an arc second guiding will let a lot of faint stuff collect quickly. Here's M81:



M81, a spiral galaxy in Ursa Major
20x900s L (5 hours)
AT10RC @ F5.1
guided and acquired using Maxim DL


And because no one ever images M81 without thinking about M82 nearby, here's M82. There is, of course, a lot of H-a activity around that one. I didn't go after that tonight because I can do that under moonlight. Some RGB on each would be good, too:



M82, a galaxy in Ursa Major
24x900s L (6 hours)
AT10RC @ F5.1
guided and acquired using Maxim DL
Begs for RGB and H-a, no?


The background is a little brighter with M82 because I wanted to keep a suggestion of the red filaments extending out of the nucleus. As a bonus, I think there's just a hint of "galactic cirrus" over on the far right.

I began the night with a "vertical" composition of M33. Nothing special, so let's save the bandwidth. And out of curiosity, I collected 15 minutes of H-a from M1 two different ways: 20x45s and 1x900s. There may be better ways to combine the first set of subframes, but a sigma clip of the former does not begin to compete with the latter. I want to work with the short exposures more before making sweeping announcements, and I want some quantitative descriptions of noise and signal levels. Until then... nothing to see here, move along, people. And when you have a choice, avoid very short subframes.

The Horsehead Nebula (B33 against IC434) skimmed between Greenwich Pine and the next one in line while I waited for M82 to emerge into open sky, so I gathered some H-a of it:



The Horsehead Nebula (B33 against IC434)
1x900s H-a, 1x730s H-a


This wasn't exactly a throwback to the days of sitting with my eye at a reticle eyepiece, but I did have to watch the guiding numbers and the image of the guidestar with more than the usual attentiveness. When the numbers diverged strongly and the star dimmed, I presumed the target had moved behind twigs. When the divergence persisted through a few guide cycles, I stopped the second exposure and worked with whatever I had. Not bad.


1/29/2012: So, I spent the evening hours revisitting IC342, then slewed to M82 to gather H-a from midnight until dawn. Stars were a good bit softer than I wanted in the first few H-a frames, so I did something stupid. I have more success focusing in Nebulosity than in Maxim, so I minimized (but did not close) the latter and launched the former. And when I tried to connect to the CCD, didn't I create a cats' cradle of software and who-owns-whats. I had to restart the remote computer and go outside to cycle the power on the camera. I spent an hour getting the show back on rails. I left the telescope taking 15s guide exposures to cling to a 10th magnitude guide star through the 7nm Baader H-a filter. I would not have bet that it would hold on, but, six hours later, it was still right where I left it (with only 0.4 arcsec RMS guide errors in x and y). The result was not as dramatic as I had in mind; the H-a emission does not isolate the extensive filaments out of plane in M82 nearly as cleanly as I expected. Maybe I'm doing it wrong. The histogram suggests that maybe I should try much longer subs.



24x900s L, 18x900s H-a
Channel-mixed in Photoshop



18x900s L



Maxim DL screenshot, cropped, showing lovely tracking performance. I saw better, I saw worse as the night went on, but I'll take the RMS values shown which are typical of recent results. I don't know where those peak values came from; next time, reset the extremes before capturing a graph for bragging rights.


1/30/2012: The answer to How Low Can You Go? is F4.8. I added a second Hyperion 14mm extension ring to the stack and came very close to focus but could not quite reach it. By removing the A-P CCD adapter and using the SBIG stock adapter with the two extension rings and the A-P CCD67T compressor, the telescope came to focus with only about 200 robofocus units to spare. EFL = 1220mm. Call it F4.8. That's it. That's almost it, see below. [2/6: And note that Pinpoint says 1232mm, F4.85, 1.24"/pixel. Call it F4.9] Well, that's it with the Teleskop-Services collimating ring in place, and it will stay in place. I also pulled the guide telescope and the bronze counterweight off the mount. Eventually, the Lunt will find a nest up top, and the bronze weight will rejoin the team. In the meantime, let's see if a lighter load makes an appreciable difference to guiding (ans: not so's I can tell).

Do not aim the telescope in its widefield mode sans dewshield at a large, low-surface-brightness galaxy 18 degrees from the bright, almost-first-quarter Moon. Nothing good will come of this except that you will appreciate what a gradient really is. You will waste valuable time speculating about how glare finds its way down the tube, past the baffles and onto the chip. Just don't.

Wth the rest of the night, I did as I have often threatened and took several RGB frames of IC443. And some luminance. 15s guide integrations let me see through all the filters with little drama. Even with these longer integrations, guide errors remained in the 0.4 pixel range:



6x900s L, 3x900s RGB
AT10RC @ F4.9


It could use a good underpinning of luminance, hours and hours of it, to smooth out the noise. I might have enough H-a and L already in the can to make some progress. Don't use this image to measure the field of view -- it's been cropped slightly after the first few blue frames (I lost the guidestar before increasing the integration time and things settled down on a slightly different center).


I noticed more than the usual tilt in these images. CCDI indicated that collimation was off, too. By daylight, I saw that the motor cover on the filter wheel had collided with the Robofocus motor and that the camera was neither seated firmly nor inserted fully. Maybe I can get it to come to focus with the preferred adapters after all, provided I avoid the small range of angles where this collision can occur. I don't think it will make any useful difference to minimum EFL, but it would be nice not to have to change all the connectors when switching from F8 to F5.1 and faster. Checking...

Yes: thread the A-P tapered nosepiece into the Mandel Muscle plate, then thread on two Hyperion 14mm fine tuning rings, and cap the stack with the A-P CCD67T (which does not thread far into the Hyperion ring). That yields an EFL of 1180mm, F4.65, and an image scale of 1.29" per pixel at the 7.2 micron pitch of the ST2000XM. Call it F4.7, and really, that's as short as it's going to get. [2/6: I got Pinpoint Astrometry LE to run again and need to revise these "manual" measurements. Pinpoint says 1211mm, F4.76, 1.26"/pixel. Now let's call it F4.8] For extended objects, that should let me achieve the same image density three times faster than at the native F8 (e.g.: in 20 minutes instead of 1 hour).

To check the field of view and image characteristics, I took a series of 30s exposures of M52 right after nightfall and before the advent of clouds. Puzzlingly large and useless FWHM measurements came out of CCDI, but the analysis program showed very low curvature (only 8-9%) with very little tilt (2-3 degrees). I walked the field down to NGC 7635, the Bubble Nebula, and waited to see if the sky would permit anything useful. Eventually, I collected six 300s frames and two 900s frames:



NGC 7635, the Bubbe Nebula, in Cassiopeia
6x300s, 2x900s
AT10RC @ F4.8


I lost the Bubble Nebula in the trees to the NW before I could get color data. I grabbed an RGB series of the Horsehead to add to the H-a shown above:



The Horsehead Nebula (B33 silhouetted by IC434)
H-a data above with 1x300s RGB added
(RGB at F4.8)


... and then turned to another big, nebula-rich field that I thought might show the widefield mode of the AT10RC to good effect: NGC 2244, the Rosette Nebula and cluster. Here's what a single 300s Luminance exposure gets at this minimum EFL:


NGC 2244, a nebula and cluster in Monoceros
1x300s L
AT10RC @ F4.8


If the weather holds up, I should have something spectacular in a few hours. I've got Maxim DL rotating through the Baader LRGB filters. I doubt the weather permits the series to run to completion, but the sequence is set to capture three hours of L in five minute gulps and half an hour each of R, G, and B also in five minute subs. I'd use much longer subs, but I really don't want to grossly over-expose the bright stars.

Couple of hours on: the weather did fall apart. Guiding held up very nicely except for sudden, disastrous excursions of 10-15 pixels. The guider always recovered in only a few guide cycles, but with so many bright stars in the frame, the damage was done quickly. Excursions were so sudden and severe that I honestly imagined that the raccoon who routinely takes down the bird feeder had decided to play with the telescope. I swept the area with a MagLite from inside, saw nothing, then stepped outside for a look around. The sky was 90% roofed over wtih altostratus. I'm astonished the guider was doing so well.

Here's what I got, and it demonstrates why it was worth working out a means of using the AT10RC at a fast focal ratio: you collect a lot of light in a not very long time.



NGC 2244, the Rosette Nebula and cluster in Monoceros
7x300s L, 2x300s RGB
AT10RC @ F4.8


Now that I've made the telescope very short, maybe I should try making it very long. GSO makes a Barlow lens that can be disassembled and put together with spacers and adapters to produce reasonable amplification with a variety of imaging trains. I would be very surprised if I couldn't use it to get an EFL of around 3000mm, F12, for tiny galaxies and (Tom's damned) planetaries, as well as for lunar and planetary photography. It always seems worthwhile to multiply the usefulness of expensive gear using inexpensive means. Leverage, that's what that is. [Agena Astro supplied the GSO ED in 3 days flat, free shipping; it must be excellent, to judge by the clouds it came wrapped in.]


2/1/2012: I added a RoboFocus motor to the Lunt solar telescope. I'm a month and a few days from having the Sun back above the pines, however briefly. The motor bolts to the bracket I made for the 5-inch refractor a few years back. The assembly should move without much fuss or bother between the Lunt and the 200mm Nikkor [Ha! keep reading], either of which can ride on the top rail of the Ritchey. The bother with the Nikkor is having to change the faceplate to use it with the CCD, which is not so bad, come to that. It can also be used with my Canon bodies. I have not dry-fitted the motor to the telephoto, but I don't see any reason it won't work. I dry-fitted the motor to the 200mm Nikkor and ended up cutting, drilling and otherwise performing extemporaneous machine shop romances to try to find one arrangement of the motor and bracket that would work for both instruments. Did. But it took a while. At a guess there are at least six binary choices (motor on the right side of the bar, on the left side; connector forward, connector backward; bracket on the left side of the Nikkor, bracket on the right; motor behind the Lunt, beside the Lunt...) which means there were at least 2^6 or 64 possible arrangements to check for compatibility with both the Nikkor and the Lunt (sufficient belt tension, secure grip, reasonably compact, no conflicts between the timing belt and other mechanicals...). I'd say less than half the possible arrangements were obvious nonstarters, so that left something like 3 dozen to check for suitability before tightening the bolts and calling it done. So, yeah, it took a while, but after trying several positions and then slicing off a little more aluminum to refine the fit, I found a very nice, compact arrangement that works for both instruments (it's perfect for the Lunt and better than merely adequate for the Nikkor). Photos to follow when I finally get one or the other out under the stars or under the returning Sun. I'd like to see it work before bragging about it any further.


Except where noted, deep-sky photos are made with an SBIG ST2000XM CCD behind a 10-inch Astro-Tech Ritchey-Chretien carried on an Astro-Physics Mach1GTO. The CCD is equipped with Baader LRGB and 7nm H-a filters. A Meade DSI Pro monochrome camera looking through a modified Orion off-axis guider keeps the OTA pointed in the right direction. The imaging camera is controlled via Nebulosity 2; the guide camera is operated by PHD Guide 1.13, both by Stark Labs. The stock focuser on the AT10RC has been augmented with Robofocus 3.0.9 using adapters turned on the lathe downstairs. Maxim DL5.12 performs image calibration, alignment, and stacking; Photoshop CS4 and FocusMagic 3.0.2 take it from there. Gradient Xterminator by Russell Croman and Astronomy Tools by Noel Carboni see their share of work, too.


:: top ::



                   © 2011, David Cortner