Michael A. Covington    Michael A. Covington, Ph.D.
Books by Michael Covington
Consulting Services
Previous months
About this notebook
Search site or Web

Daily Notebook

Popular topics on this page:
How to keep astronomers warm
How to predict the stock market
Why you can't interconvert between word processors
Sky-Watcher illuminator fits Celestron, can be brighter
M35 and surrounding objects
Quasar CTA102 (4C 11.69)
Horsehead and Barnard's Loop
Seagull Nebula (IC 2177)
Rosette Nebula
Rosette Nebula (wide-field color)
Rosette Nebula (wide-field narrowband)
Field of Xi Puppis
Many more...
This web site is protected by copyright law.
Reusing pictures or text requires permission from the author.

For more topics, scroll down, press Ctrl-F to search the page, or check previous months.

For the latest edition of this page at any time, use this link: www.covingtoninnovations.com/michael/blog


Sky-Watcher polar scope illuminator fits Celestron, can be brightened

The polar scope illuminator for Sky-Watcher equatorial mounts fits the Celestron CGEM and AVX (and maybe others). Sky-Watcher and Celestron use the same Chinese manufacturer, and although the products are designed separately, there is a surprising amount of interoperability, of which this is an example.

The illuminator is sold as a separate item in other countries but not the United States. I bought mine, secondhand, from a Briton.

The recommended battery is a 2032 lithium coin cell. The 2025 also works. But if you want more brightness, my unofficial recommendation is two 2016 cells, stacked. The current through the LED is still only 5 mA, which shouldn't cause damage, and you still have adjustable brightness. (Caution: Do not stack two 2016's in place of a 2032 in any other equipment. I have made tests of this one.)

Background: If you're wondering what this is all about, let me explain. Many equatorial telescope mounts have a tiny telescope built in through which you can sight Polaris to get the axis of the mount aligned with the axis of the earth. If you do this, I strongly recommend Polar Align Pro software to tell you where Polaris should be relative to the reticle, which consists of crosshairs and circles.

Against a dark sky, you can't see the reticle. That's where an illuminator comes in. There's no way to illuminate the reticle directly (to see red lines against a black sky), but you can inject some scattered red light into the field so that you see dark lines against a reddish sky, dim enough that you can still see stars.

Homemade illuminators are easily improvised, but Sky-Watcher makes an especially compact one.


Narrowband Monoceros

My astrophotography project last night (January 26) was just one picture — and all I did was set the camera on the iOptron SkyTracker and leave it running, with the exposure timer going, for an hour and a half.

The picture shows gas clouds (nebulae) in the constellation Monoceros, and especially the Rosette at bottom. Compare it to my recent color picture of the same fieldand narrowband picture of Orion.

Today's picture is a stack of 50 1-minute exposures with the Canon 60Da, Sigma 105-mm f/2.8 lens, B+W 091 deep red filter, and iOptron SkyTracker. Although presented as a black-and-white image, it uses only the extreme red end of the spectrum to avoid sky glow from city lights.


Field of Xi Puppis

Here is a rich field in the constellation Puppis, southeast of Sirius. The brighter star in the pair in the middle is Asmidiske (Xi Puppis). Above it is the star cluster M93; to the left, the cluster NGC 2482; and below it, the cluster NGC 2467, which includes nebulosity. Stack of 26 30-second exposures, Sigma 105-mm f/2.8 lens, Canon 60Da, ISO 1600, iOptron SkyTracker.

This is a star field that I will revisit with other instruments. There are several interesting objects.

Clusters and nebulae in Gemini

Here you see the star cluster M35 at the upper right, with the fainter cluster NGC 2158 immediately to the lower right of it; the nebula NGC 2175 (the Monkey Head Nebula) at bottom; the star Tejat (Mu Geminorum) at the left; and to the right of Tejat, the faint red nebula IC 443, the Jellyfish. The upper part of the picture is in the constellation Gemini; the lower part, including the Monkey Head, is across the border in Orion.

Stack of 30 30-second exposures with the same equipment as above.


Rosette Nebula in context

Here is a wider-field view that includes the Rosette Nebula at lower right and, to the upper left, the Christmas Tree Cluster and various gas clouds. Like the pictures below, this was taken in my driveway in Athens, Georgia, but with this one I used a Sigma 105-mm f/2.8 lens on an iOptron SkyTracker. Stack of 25 30-second exposures with a Canon 60Da at ISO 1600.

What you can do for your country

Many people are responding to the inauguration of an unusually unpopular president by re-fighting the campaign, as if denying reality would change it. The time for campaigning for or against individuals is past. Instead, we need to turn our attention to issues and actions, pull together, and try to make our country better.


Rosette Nebula

The Rosette Nebula is a region of active star formation in the constellation Monoceros. It is basically an irregular roundish gas cloud whose center has condensed out to form stars, leaving a hole in the gas. It is hard to see in a telescope because so much of its light is in the deep red, at a wavelength to which the eye is not very sensitive, and was discovered piecemeal by visual observers who found various bright regions within it. Using binoculars I find it easiest to see as a dark nebula, because the gas cloud hides the stars behind it; at a sufficiently dark site I can see that the nebula glows faintly.

This is a stack of thirty 1-minute exposures with a Canon 60Da (specially built to have high sensitivity in the deep red), AT65EDQ refractor, and AVX mount, in town (Athens, Georgia).



This is the globular cluster M79 in the constellation Lepus. The first picture is the cropped and enlarged center of the second. This is a stack of twenty 1-minute exposures through the AT65EDQ refractor and Canon 60Da. The second picture also shows, at the upper right, the faint galaxy NGC 1886 (13th magnitude).


Why you can't interconvert between word processors


"Your paper is written in LaTeX, and we need it in Microsoft Word. Can you convert it?"

The answer is "no," not without editing it by hand. Even much less drastic conversions, such as between Word and Perfect Writer (remember that?), are imperfect.

Of course, it depends on what you mean by "convert." If you mean "convert the text and some of the formatting, so we don't have to re-type every word," then yes, you can convert between word processors. After converting, you'll have to work through the document to get the formatting exactly the way you want. But you won't have to re-type it all.

If you mean "produce something that will print exactly the same way without any hand-editing," no, that's what you can't do.


In particular, why, considering that we can inter-convert graphics? You can turn a PNG file into a TIFF file with absolutely no visible change. Why not a written document?

The answer is document structure, or more generally document semantics.

A word processor file does not just say what characters to put on the page, in what positions, in what sizes and fonts. If it did, it would be perfectly interconvertible. In fact, PostScript and PDF are that kind of file — just character positions, sizes, and fonts — and they are interconvertible.

No; a word processor file also makes claims about the structure of the document. As soon as you allow paragraphs to re-wrap, you have to represent something that is not just the position of the letters on the page. You have to have the concept of "end of paragraph." The end of a paragraph is not something you can see. It's a meaning attached to a particular point in the text.

So far, so good. But any word processor worth its salt is going to include aspects of document structure that go much further, such as "chapter heading" and "table." And we often want to say how a chapter heading should look separately from saying that a particular phrase is a chapter heading.

And that's the problem. Writing and editing is an art, and no two artists have exactly the same idea of how to represent every possible part of a document — or even what the possible parts are.

RTF (Rich Text Format) is helpful but is not a complete solution. It is a sort of least-common-denominator that represents simple word processing functions portably, but not more advanced ones.

The only way to make word processors completely interconvertible would be to forbid people to have any more new ideas about how to structure documents.

On top of all that, different word processors don't even have the same printing capabilities. LaTeX has its own mathematical symbols, for example, and its own built-in knowledge of how to lay out formulae. Word can type emoji that LaTeX can't. And so on.

The reason your picture interconverts perfectly between PNG and TIFF is that it has no internal structure. It is supposed to look the same all the time. Written documents are not like that; they rewrap when we change the margins, and the software is supposed to know how the document is organized.


Seagull Nebula flying out of the picture

This is supposed to be a picture of the Seagull Nebula in Monoceros, which you see at the very right. In the middle is the star cluster NGC 2343, and throughout the picture are fainter gas clouds. I'm not sure why I didn't aim the telescope straight at the intended object; obviously this is one to revisit.

Stack of 26 1-minute exposures, AT65EDQ refractor (6.5-cm f/6.5), Canon 60Da at ISO 1600, Celestron AVX mount, in my driveway in Athens.


Narrowband Orion

First astrophoto of 2017, resuming (with great success) an old experiment. Here you see Barnard's Loop (a very faint nebula stretching across the left side of the picture), the Horsehead (the small notch in the nebula below center), and other gas clouds in Orion.

This was taken in town, with the gibbous moon just 30 degrees away. I used a Canon 60Da and a B+W 091 deep-red filter; the picture was taken entirely with wavelengths of light between 630 and 700 nm. That excludes almost all of our streetlights and much of the moonlight, but the nebulae, shining at 656 nm, are practically unaffected. Stack of 50 1-minute exposures with a Sigma 105-mm f/2.8 lens on an iOptron SkyTracker. Only the red pixels in the Bayer matrix were used.



A new etiquette controversy: "Cutting off"

Sharon points out to me that there is starting to be a kerfuffle about the practice of "cutting people off," declining to communicate with them, on social media.

This is especially claimed to be a characteristic misdeed of "millennials." Well, I deplore the practice of sawing society up into "generations" and claiming that every generation but your own is obnoxious. But we'll save that for another time.

The main point is, in a world of social media and instant communication, we've lost the middle ground between people we see all the time and people we never see. It is hard to be in occasional but regular touch with someone.

And this leads to difficulties. The person who used to exchange greetings with you twice a year may now demand a constant audience for his political opinions, which you cannot provide. What to do?

My hat is off to Facebook for managing this to some extent, the only way they can: They don't show you all your friends' postings, only as many as you seem to have time to read, giving preference to the people to whom you frequently reply.

That works, except for the people who say, "I want to see who my real friends are. If you are really my friend, copy and repost this..." and Facebook does not show this to all of their friends in the first place because it is a copy of a widely circulated message. Then they get upset because you didn't reply to things you never saw.

More generally, social media make it easier for people to be unreasonable, sometimes never realizing they are doing so. Then the only alternative is to stop replying to them, and they'll say you've cut them off.


Photographing a quasar with a 1.5-inch instrument

Quasars were discovered with radio telescopes and first observed optically with the 200-inch telescope on Mount Palomar. They are believed to be black holes that are absorbing matter, causing it to emit light and radio waves, and are the most distant objects in the known universe.

Right now, we have an opportunity to observe a quasar with much smaller instruments. One of them, CTA102 (4C 11.69), in the constellation Pegasus, has flared up to magnitude 11.5. You could see it visually with an 8-inch telescope. I photographed it with an aperture of just 1.5 inches.

The picture is a stack of 13 30-second exposures with a Canon 60Da at ISO 1600 and a Sigma 105-mm f/2.8 lens on an iOptron SkyTracker. This is my ultra-portable setup — all of it, including the tripod, can be carried in one hand — but I was using it at home, in my driveway. The brightest star in the picture is HIP 111601 in Pegasus.


Snow's coming...

Northern Georgia has a winter storm warning for the weekend. The last time we had one, 3 years ago, people ignored it, and Atlanta had a massive snowjam. Key points:

(1) Southern snow falls on warm ground, melts, and refreezes. It's not powdery like up north. Nobody can drive on it.

(2) No matter how good your car is for driving on snow, it can't drive through an ice-locked traffic jam. That is how Atlanta snowjams happen. A few cars that can't move on snow and ice block everyone else.

(3) Many people's ordinary Atlanta commute is a 2-hour road trip over very hilly terrain, not safe to make when there's snow starting to fall. Everyone, please be wise and be safe.

Ektachrome coming back; Showcase Photo going away

The last really good Atlanta camera store, Showcase Photo, will close February 28. I will be sad to see it go.

But late in the year, Kodak Alaris will resume selling Ektachrome E100 film. The market hadn't totally dried up. I could have told them that; Fuji was still selling very similar products! In fact, right now, you might have to process Kodak's film in Fuji's chemicals, unless Kodak comes up with something.

As you might guess, they needed it for motion pictures, not just color slides.


Weber had a word for it
(3 reasons for being loyal to a regime)

A while back, I mused that many people's attitude toward politics and voting was far too much like football-team fandom.

If I had been better-read in political science, I would have recognized something. Max Weber, one of the founders of the field, named three reasons why people are loyal to a ruler or government:

(1) Charismatic: They admire their valiant leader.

(2) Traditional: The regime has been there for a long time and they rely on it for stability — that is, they rely on each other to respect it.

(3) Rational-legal: They believe the regime was set up legitimately, is fulfilling its responsibilities reasonably well, and can be improved as needed without overthrowing it.

Obviously, modern society is mostly (3), somewhat tempered by (2). In fact, Weber points out that the growth of bureaucracy is not altogether a bad thing — it shows that power is spread among people who are accountable to each other and have to follow rules.

Of course (1) will never completely die out, especially in political campaigns. But it has its excesses. "Ooooh, he's Rich... he's a Star!" Did you hear any of that during the recent presidential campaign? Was it supposed to be a good reason to vote for someone? But on the other hand lots of people legitimately admire real achievements, and any political movement may need to put up a charismatic figurehead in order to attract a certain kind of supporters.

There are also excesses of (2), which we hear occasionally from people who think the Constitution is of divine origin and cannot have any flaws. Compare the divine right of kings — a transitory theory that was not held in Britain when the monarchy was originating, nor today.

Football-team loyalty is a mix of (1) and (2), of course. Part of the game is to pretend to be a tribal society. But nobody thinks this is actually a good way of choosing governments, I hope — it's just a game.




I didn't realize quite how much my new cold-weather suit resembles a late-model space suit until granddaughter Mary started calling it my "astronaut suit." Hmmm...


How to predict the stock market (or anything else!)

One of the most important reasons people analyze data is to make predictions. In what follows I want to quickly run through some of the pertinent concepts. I should explain that although this is not my main specialty, it's something I like to keep up with.

The simplest way to predict the stock market, the weather, or any other measurable quantity is to guess that it will be the same tomorrow as today. That is almost always close to true; enormous sudden changes are rare. So you can predict the stock market (or the weather) to a high percentage of accuracy by just doing that.

And that kind of prediction is basically useless. What we really want to predict is the change from yesterday to today. That's much harder.

We should also ask whether we are predicting something (stock prices, air temperature, or what have you) just from its own previous values, or from something else too. The math is similar in both situations except that in the second one, there are more input variables.

It's also worth remembering that "previous values" can often be split up into components. Maybe changes in the price of small-cap stocks predict changes in large-cap stocks, or the other way around. Maybe rain in Alabama predicts rain in Georgia. (It certainly does; that's the way storms move in the South. Economists would say that rain in Alabama is a leading indicator of rain in Georgia; that is, it "leads," or happens sooner.)

Now then... how do you make predictions?

One thing that might be going on is a steady trend. That is the territory of linear regression, a technique overused by amateur statisticians ever since it became easy to do on pocket calculators. The problem is that the steady trend might not be a straight line. Once, very early in my career, I was asked to predict the (steadily rising) number of students in a particular program. Someone else had already done a linear regression. I looked at the data, saw that the rate of change was increasing, and cobbled together a quadratic equation. My prediction was more accurate.

And of course there is always some random noise. One way to remove random noise is to replace the data with a moving average; that is, average each point with a fixed number of previous ones. You get a graph that still wiggles, but much more gently.

Moving averages are popular with stock market analysis. They work well with quantities that are not likely to undergo sudden changes. In image processing, though, they blur the picture too much, and random noise is removed in several other ways, all built around the concept of maximum entropy (maximum simplicity) — finding the simplest underlying data that could be the true image without the noise. (You've been told that "entropy" means "disorder," but it might better be described as "lack of features." We don't want to imagine that all those little noise bumps are something real when they're not.)

Beyond long-term trends and random noise come the interesting things. Time-series data very often includes periodic trends (summer versus winter, Monday versus Friday, etc.) which can be found by autocorrelation and Fourier analysis. Sometimes these are of great scientific importance. For instance, a periodic variation in one planet's orbit can indicate the gravitational pull of another planet — even an undiscovered one — and tell you its orbital period.

There is also the tendency to "bounce" after a sudden change ("ringing," like the way a bell vibrates after being hit). This bounce has a particular frequency. That is, there may be a persistent tendency for the quantity to be changing the opposite of the way it was 3 days earlier.

And even variations that seem random may have a preference for changing at some speeds and not others (detectable by Fourier analysis). In essence, a change builds up momentum and will take time to reverse itself, even if an external cause that should reverse it appears suddenly.

Last, the whole thing may be gradually changing with time, so that even the most abstract rules and parameters are different now than they used to be. That is an argument for not using excessively old data, or models built on excessively old data.

It happens that there is a set of mathematical modeling techniques that capture all of these things, together, in a single computation. One of them is called ARIMA (autoregressive integrated moving average) and can often do impressive things.



The other Charles and Hazel Covington

I am posting this to help genealogy researchers avoid confusion.

My parents were Charles Gordon Covington (1923-1966) and Hazel Roberts Covington (1925-1985). Apart from my father's military service before he got married, neither of them ever lived outside Georgia. When I was growing up, we lived in Valdosta.

My daughter Cathy, researching genealogy, has found another Charles and Hazel Covington with a Valdosta connection. Here's her report:

Charles E. Covington (22 May 1886-?) married a woman named Hazel, and they lived in Nebraska. He had five children, Helen, Bonita, Phillip, Richard, and Mary. The youngest daughter, Mary Catherine, moved to Valdosta right around the time you moved out [c. 1973] and became a prominent member of the local community under her married name (Crane) and is now buried in Lowndes county.

This Charles was the son of Asa Johnson Covington, born in Texas to a William Covington born in NC around 1829. Here the trail is lost because there were three William Covingtons born in NC that year, but I suspect these are our distant relatives since our Covingtons came to GA from NC and I've not traced any of those descendants yet.

Since this is the only other case of a Charles and Hazel Covington I've found and there's a Valdosta connection, I suspect future researchers may be confounded. I certainly was surprised to see "the youngest of five children born to Charles and Hazel Covington" pop up in my search results.


How to keep astronomers warm

Astronomy in the winter is a chilly pursuit. We sit still, outdoors, in the middle of the night. In winter, that exposes us to colder weather than anyone else.

Over the years I've learned that the clothes I wear for walking around in the daytime (even for long city walks at 25 F) are not sufficient for winter astronomy. Of course, everyone knows the secret to being warmer is layers (long underwear, maybe more long underwear, heavy trousers, thermal undershirt, shirt, sweater, jacket...). But that can be awkward! Maybe I'm lazy, but I don't want to spend twenty minutes layering-up for an astronomy session. I also don't want to be clothed in things that are way too hot for indoors and can't be taken off easily! So I want something that is very warm, but simple.

For some years I relied on hunting coveralls. They're handy; they give you a warm garment for the whole body in one piece; but they're not really warm enough for winter astronomy in North Georgia. I find them comfortable down to about 50 F over regular day clothes. Of course, with more layers, they can be much warmer. They are normally used as the outermost of several layers.

Recently several people clued me in that what the astronomer needs is a freezer suit, a special coverall made for people who work in the frozen-food industry. It is very warm by itself, not just over other layers. It's quick to get in and out of. And it's available in bright colors for visibility. Frankly, I don't enjoy being camouflaged on the observing field. Bright orange is good.

These last two nights I had two long astrophoto sessions while wearing my new Refrigiwear coverall. Verdict: Success. It was quite comfortable down to 30 F, at least, and didn't seem to be near its limit. Now I need better socks and a better hat, but the coverall is excellent!

If you want one, go in through their main web page because there may be a discount coupon. Study the size charts, and if you think you need a "tall" size, you do. (I did; the legs and sleeves are a bit too long, but they bunch up, trapping more air, and don't get in the way.) Refrigiwear's customer service is excellent.

How warm a suit do you need? Was it foolish of me to buy one rated to -50 F for use in Georgia? No; I didn't get too hot, and I could even wear it while taking fairly long breaks indoors without getting too hot. Here's why.

First of all, Refrigiwear says to allow an extra 20 degrees F for sedentary work. Astronomy is definitely sedentary. Second, we lose a lot of heat by radiation to the open sky (the same reason telescopes get colder than the surrounding air, causing dew to form on them). I suspect that's worth another 10 degrees F. Accordingly, a suit that is rated for -50 to +20 F (for active work inside a meat locker) is probably good for -20 to +50 F for the astronomer out under the stars. That more or less covers the North American winter.

Refrigiwear gets very good reviews, even from Alaskan snow machine riders. Their products are said to beat mountain-climbing suits that cost five times as much. But Down Under, in Australia, there's something else — Hepworths. A correspondent tells me Hepworths freezer clothing is de rigueur at Australian star parties, when they're looking at Scorpius directly overhead and shivering. I'm particularly intrigued by Hepworths "freezer trousers" (apparently warmer than Refrigiwear's Coolerwear overpants). Something to investigate...

If what you are looking for is not here, please look at previous months.