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
The ads that you see here are not controlled or endorsed by Covington Innovations.
They come from Google based on your browsing history.
If offensive ads appear, click on "AdChoices" and then "Ads Settings" and make changes, and/or turn off cookies in your browser.
Click hereto see books and other items
(mostly astronomy-related)
that I am selling through Amazon. More will be added in coming days.
2015 May 28-31
More hasty notes
After this, I'll see you in June. I'm too busy with too many things to write much
until then, so I'll close out May with these short notes.
(2) I'm trying out a
Tele Vue Dioptrx astigmatism corrector
(on a 32-mm Tele Vue Plössl eyepiece).
It seems to work well. (My correction has been measured variously
as 0.50 to 1.00, so I ordered 0.75.)
The crucial difference is that the DioptRx is centered on the eyepiece,
which your glasses aren't.
When you observe through glasses (as I always have), the eyepiece, eyeglass lens, and eye
are rarely all aligned. The DioptRx gives a bit of extra crispness to the view.
Granted, I haven't tried it on my EdgeHD telescope yet, only on the C5 under poor conditions.
And this leads to another point:
EdgeHD is a game-changer as regards eyepieces.
Like refractor owners, and unlike users of reflectors and SCTs,
we finally have flat fields that are sharp from edge to edge.
So if the eyepiece isn't, we notice.
More about this later.
The post-Memorial-Day frenzy is upon us. The business world is making up for lost time.
My consulting business is booming — in fact there may be some really big developments by
the end of the summer — when there's anything to announce, I'll do so.
Very short notes to be followed up later:
I've discovered that super-pixel mode in DeepSkyStacker was producing color fringes
around my stars. (The star images were sharper than it assumed!) So I'm going to reprocess
a lot of recent astrophotography and get better color. Stand by for pictures.
The pictures from my astronomy session on Saturday are also yet to come.
On Friday and Saturday I got enough mosquito and chigger bites (and one tick bite) to last me
the rest of the year. So I responded with chemical warfare — bifenthrin the first day
and malathion the second. There will be more chemical warfare before my next observing session.
There will also be 25% DEET on me and permethrin on my shoes, which are going to become a death
trap for chiggers.
(Before you ask: It was a small male Lone Star tick, a species that does not carry Lyme disease,
and was not engorged. It looks like it made two trial bytes, both of which itched for two days
afterward, but didn't make contact with my blood.)
And my newest toy is a Tele Vue astigmatism-correcting eyepiece. Naturally this brought
a long rainy spell, and I haven't been able to try it out yet. If I weren't so busy, I'd try
it out looking at treetops or something.
On the night of the 23rd I had another very good
galaxy quest
(no, not the kind the link refers to!)
and shall be posting pictures soon.
In the meantime, here's a picture of the setup I used:
That is, Celestron C8 EdgeHD telescope on Celestron Advanced VX mount, and, roughly
from top to bottom, Kendrick dew remover on the C8 with homemade controller;
Orion 80mm ShortTube guidescope; ADM rings and dovetail;
homemade additional support for autoguider camera using an ADM clamp and
a piece of 3/4-inch square aluminum tubing;
Orion 50-mm eyepiece tube extender; SBIG STV autoguider (only the camera
is shown); Celestron EdgeHD f/7 compressor lens; Canon 60Da camera; and a
cheap imported exposure timer connected to the Canon.
Here you see the crucial role of zip-ties in my setup. I put them on after
setting everything else up, then cut them at the end of the session.
They ensure that any tugging on the cord won't move the autoguider camera
relative to the dovetail that supports it (which is firmly attached to
the telescope). Lots of mischief took place when I didn't use zip-ties.
I have also opened up the autoguider camera, added a layer of rubber tape to
make the strain relief grip the cable tightly (which it wasn't doing), and
tightened everything up, so the zip-ties might not be strictly
necessary now, but they can't hurt!
Did Constantine throw out some books of the New Testament?
Recently, people have been asking me about
"the books of the New Testament that the Emperor Constantine threw out."
That left me scratching my head. I knew of no such event.
I knew that the formation
of the New Testament was a gradual process,
but all indications are that it was mostly a matter of gradually adding
books, not throwing them out.
I knew that "rejected" books did exist — people kept writing books
that tried to combine Christianity with Gnosticism — but these were
pretty widely rejected right from the start.
And there were, and are, solidly Christian books dating from just after the
New Testament, which were classed with Scripture by a few people in early times.
These books, such as the Didaché, the Epistles of Clement, and the Epistle of
Ignatius, were not "thrown out." They are still available;
you're welcome to read them,
and they don't say anything scandalous. They just aren't included in the
New Testament because they date from a slightly later time, as is obvious when
you read them. They describe the early functioning of the Church.
Maybe people were thinking of the Council of Nicaea, convened by Constantine.
But that Council did not reject any New Testament books.
We know, because councils publish their decrees.
Maybe people were thinking about reports that Constantine had a number of
Bible manuscripts copied and distributed. But there are no reports that he
threw out any of their contents.
Then I found out what was going on.
Constantine's "throwing out" of NT books is fiction,
found in The Da Vinci Code.
It is fiction. It is no more real than Sherlock Holmes or Mickey Mouse.
It never happened.
Some people might ask me to prove it never happened.
That would be like asking me to prove that Mickey Mouse doesn't live in Burbank.
Fiction is fiction, folks.
Modern people can easily fail to realize how much we know about ancient and
medieval history.
It's not guesswork.
Surviving records are abundant, and
(before you ask) a lot of the material has never been controlled by
the Vatican. Plenty of it is preserved in secular museums, libraries, etc.,
and in the Eastern Orthodox churches, which are not centrally controlled.
In fact, a lot of it was common knowledge among better-educated people a couple
of centuries ago!
For a more detailed treatment of the question,
see this essay
by M. James Sawyer (about whom I know very little, but what he wrote is solid).
While it's Bible study time in the Daily Notebook, I hereby reprint, for your amusement
and with his permission, a witty response I got from my old friend, Rev. Doug Beyer:
Last night (May 22) I had a very good astrophoto session with my 8-inch
EdgeHD telescope and new f/7 reducer, in my driveway under magnitude 5 skies
in somewhat unsteady air.
Each of the following is a stack of 148-second exposures at ISO 1250.
I chose the exposure time and ISO setting primarily to give the autoguider
a good test rather than to get the best possible pictures.
Here's the galaxy M66 in Leo, a stack of 3 148-second exposures:
This is a particularly pleasing picture, a stack of six 148-second exposures of
M102, the galaxy I wrote about last year
and about which I have an article coming out in Sky and Telescope.
I reconstructed how Pierre Méchain discovered this galaxy and then failed
to realize he had done so, leading to the long-standing belief that
"M102" was an unassigned number, a duplicate of M101.
In this stack of 5 148-second exposures, you see M51 (the first spiral galaxy
to be discovered) and its companion galaxy NGC 5195, to which it is linked by
a streamer of stars and gas. But you can also see the low-surface-brightness
galaxy IC 4263 (15th magnitude, at the bottom right) and the distant galaxy
IC 4278, about which little is known.
For my non-astronomer friends I should emphasize that
no telescope in the world gives a view like this; it relies on the
ability of the camera to accumulate light in a long exposure.
The spiral structure of M51 was discovered telescopically, but it was
a close call; Lord Rosse, with his 72-inch telescope, could barely
see the spiral.
The view through even a big telescope is very dim,
apart from the central "fuzzball".
This picture almost needed a smaller telescope, but it's a good example of the
edge-to-edge sharpness of the EdgeHD optics. (It's not quite the full frame;
you'll have to take my word for it that the stars are just as sharp all the way
to the corners.) At only 7200 light-years, this globular is unusually close to us;
typical bright globulars such as M5 and M13 are three or four times as distant.
Someone is finally making a gadget that I wished for, and almost invented,
twenty years ago: A pill-bottle cap with a clock built in that tells you how
long since it's been opened. Click here and buy some.
These are ideal for pills that are taken when needed, but then must not be repeated
for a certain length of time — such as the pain pills we give our arthritic dog —
and also pills you might wonder if you forgot to take.
There is nothing to set. The timer starts when you re-close the pill bottle, and counts up
to (I think) 99 hours and 59 minutes, and then shuts off.
The battery is not replaceable (at least by ordinary mortals; I have a well-equipped electronic shop,
so when the time comes, I'll try).
It lasts about a year in use, or much longer if the timer is shut down (i.e., past 99:59 or with the
cap off the bottle). The caps cost less than $8 each, and considering that a year's supply of almost
any medicine costs much more than that, they're a bargain.
Have computers replaced star atlases? Not at all. I think there is still a place for a well-crafted
paper atlas, for use at the telescope, and this one introduces some useful new features.
For starters, it's the right size, about 11 by 11 inches. The pages are printed on both sides so that you
view a two-page spread but can fold the atlas back to hold it in one hand.
The classic atlases of Bečvář [you didn't know I could type that, did you? I can even pronounce it]
and Tirion are too big to handle easily at the telescope. The
Pocket Sky Atlas from Sky and Telescope
is very handy — in fact I use it constantly — but it only goes to magnitude 7.6.
Enter Interstellarum. Developed in Germany by the group that publishes an excellent magazine
(Interstellarum) under the leadership of distinguished
deep-sky expert Ronald Stoyan, the Interstellarum Deep Sky Atlas efficiently presents
the information a telescopic observer actually wants.
Objects are classified by visibility (4-inch, 8-inch, 12-inch, or larger telescope required),
size, and type. Popular names as well as NGC numbers are given.
M102 is identified as NGC 5866 (thank you!).
And stars are shown to magnitude 9.5.
Double stars are first-class citizens (finally!
almost
for the first time since Norton!) and
are shown with rough graphical indications of position angle, separation, and magnitude difference.
I field-tested this atlas by having a ramble through Corvus and Virgo with my 5-inch telescope
during a recent break in the clouds. The atlas showed me stars I was seeing in the telescope,
not just in the finder. I was pleased that as I approached M104, it showed me the Stargate Asterism
and another asterism shaped like a hockey stick. It also drew my attention to a few double stars
that I would otherwise not have gotten to see.
The only thing that puzzles me a bit is the name. Inter stellarum is not a complete phrase
in Latin; "among the stars" would be inter stellas. I can only hazard two guesses.
One is that this is short for inter stellarum X where X is some other word;
"among the X of the stars." If this is a literary quote, I haven't been able to track
it down. The other is that maybe the authors made up the word
interstella for "object between the stars," i.e., deep-sky object, and then of course
Atlas interstellarum makes perfect sense.
But then, I'm so old I once had to decipher Atlas Coeli Skalnaté Pleso, which is half
Latin and half Slovak —
a book we remember with great fondness because it
inaugurated the modern era of star atlases.
This is another of my all-time favorites among my color slides.
If memory serves me right, it was taken out of my window at Clare College
(Thirkill Court) in the spring of 1978, with an Olympus OM-1, a 100-mm lens,
and either Agfachrome or Kodachrome, probably the former.
Ordinarily, data structures in C# are limited to 2 GB in size.
Although I haven't looked into the details, I presume this is because the
pointers used internally are 32-bit integers with the top bit unused.
In .NET Framework 4.5, there is an exception. Arrays, and only
arrays, can be bigger than 2 GB. This is true even if the array is
within another object (i.e., there is a reference to it in another
object; in C#, that's what happens if an array is part of
a struct or class).
There are still two limitations: An array can't have more than about 4 billion elements total,
nor more than about 2 billion elements in a single dimension. Here by "2 billion" I mean Int32.MaxValue.
To use giant arrays, you must do the following:
Compile for x64 architecture (not "any CPU") and make sure "Prefer 32-bit" is unchecked.
Include the appropriate declaration in the App.config file.
This will cause an .exe.config file to be generated along with your
executable .exe file.
Actually have the .exe.config file in the folder with the .exe when
you run it.
The App.config file is one of the project files you've probably been
ignoring, but it's there (under Properties) and you can edit it.
Here's mine, with the crucial code marked:
A file with the same contents is generated with the name XXXXXX.exe.config (where XXXXXX is the
name of your program) and copied into bin\Debug or bin\Release alongside your .exe file.
It needs to be present whenever the .exe runs.
In fact, I recommend having your program trap the out-of-memory exception and warn the user
that the .exe.config file may be missing if the program runs out of memory.
As far as I know, there is no limit on the size of an array you can allocate this way,
at compile time or at run time.
You can create one that is bigger than your computer's RAM, and Windows will happily
spend a lot of time swapping to disk.
I don't often get to blog about my day job, but today I want to write about a new
technology that is important for almost everything I work on.
It's a way of getting statistical information about relationships between
words in a human language.
I can't share anything proprietary, so I can't exactly tell you what I've been
working on, but I can tell you some things that are public knowledge.
I need to start with just a bit of mathematics.
A vector is a sequence
of numbers, such as (5, 4, 9, 36, 33, 0.5).
Vector similarity, also known as cosine similarity,
is a way of comparing vectors that ignores the size of the numbers and pays
attention only to their proportions.
For example, (0, 2, 4) and (0, 20, 40) are perfectly similar, with a similarity of 1.0, because
they are alike except for the sizes of the numbers. (One of them has numbers exactly 10 times as
big as the other.) The vector (0, 1, 2) is also perfectly similar to these.
The vectors (0, 1) and (1, 0) are as different as can be, so they have a similarity of 0.0.
Now for some computational linguistics. A term-document matrix or word-frequency table
is a table of how many times each word occurs in several different documents (texts).
Here's a made-up example:
Document 1
Document 2
Document 3
aardvark
5
0
1
abacus
0
3
6
integer
1
10
5
lion
13
2
1
At a glance, you can see that Document 2 is more similar to Document 3 than to Document 1.
(Documents 2 and 3 seem to be about mathematics; Document 1, about animals.)
You can also see that aardvark is more similar to lion than to integer.
(It occurs in more of the same documents.)
What I've just asked you to do is estimate vector similarities.
Words are vectors (rows in the table)
and documents are vectors (columns in the table).
Measuring whether documents are alike is a basic technique
of information retrieval.
Measuring whether words are alike is important for building dictionaries of
any kind, or for any task that wants to treat related words alike.
And if you want to know whether words are alike, you probably want to look at contexts
smaller than whole documents. Ideally, you'd like to move a 10-word-long window through your
whole collection of texts, and treat each 10-word sample as a separate document.
But if you do this, you get an enormous matrix — maybe 50,000 rows for
the words, and almost a million columns if you have a million words of text.
Accordingly, the challenge is to reduce the dimensionality of the matrix in such a way
that essential information is preserved.
An older method, used in information retrieval, is
Latent Semantic Analysis (LSA, also known as Latent Semantic Indexing), based on
singular value decomposition of the matrix.
If you turn the matrix sideways, to compare words instead of documents,
LSA doesn't work too well (I tried it).
What is new is that some new, fast techniques have been developed for
building sets of word vectors from huge amounts of text.
Not only the algorithms and software, but also the actual vector sets,
created from multi-billion-word text collections, are freely shared.
There are at the moment two rival contenders,
Google's Word2vec
and Stanford's GloVe.
Both produce vector sets with 300 or fewer columns.
Having one row per word is unavoidable, but at least the columns
don't number in the billions.
Here's an example of Word2vec finding words that occur in contexts
similar to parakeet:
Not all of them are synonyms — look at cat, for instance —
but all are words that Word2vec found in similar contexts.
And here are words similar to however. This is a handy way to
get a set of words that indicate contrasting ideas:
Again, there's some chaff in with the wheat, but the output is nonetheless useful.
That's not all. Word vectors of this kind can solve analogies.
Prince is to princess as actor is to what?
To find out, you subtract the vector of prince from the
vector of princess. You then have a vector that means,
roughly, "female counterpart." Add it to the vector of actor
and look for words close to the vector that results.
Here's what you get:
The best fit, actress, is perfect. Some of the other guesses
are a bit flaky, but they also have much lower similarities.
(Why similarity is called "distance" in this program is puzzling;
somebody's thinking backward.)
Think how useful such a thing would be for question answering.
If it didn't give you the right answer as its best guess, it would
certainly give you a set of guesses that are close.
Bottom line? This is no substitute for logical reasoning, but
it's a powerful way to extract usable knowledge from text automatically.
And one thing we learned is that vector sets need to be built from
billions of words of text. Anything less than that, and they
don't see enough examples of most of the words.
I want to report a great astrophotographic triumph.
Using this setup with one minor change,
I got obstinately good tracking (100% of more than
fifteen exposures well tracked).
The change was to firmly zip-tie the autoguider camera cable to the dovetail
near the saddle where the mount holds it.
Here's a stack of three 150-second exposures of M102
(the galaxy I wrote about last year)
at f/6.3 and ISO 1250. The small f-stop and low ISO number enabled me to make longer exposures to test tracking;
ordinarily I would have worked at f/4 and ISO 1600 and exposed for 30 seconds.
I was delighted to find that Celestron's computer agrees with my opinion that M102 = NGC 5866.
As you can see, I also picked up three more galaxies.
And here is the globular cluster M13 in Hercules (with a faint galaxy some distance to its upper left).
This is a stack of three 150-second exposures at f/10. Yes, f/10, so that I could make such a long
exposure against a bright sky. This is a considerably cropped picture. It is downsampled only ×2
from the original; that is, each pixel on your screen is four of the camera's pixels.
As you can see, there's nothing to complain about as regards tracking, even though the autoguider
said it was having a moderately hard time due to atmospheric unsteadiness.
And all of this makes me wonder
why zip-tying the cable made such a difference, when the camera was on such a
rigid support to begin with.
Is there something mechanically loose in the STV camera such that a tug on the cable
makes the whole circuit board shift a little?
Has the cable's strain relief failed, or was it inadequate all along?
Eventually I'll open it up and find out.
In the meantime, I wonder how many of my autoguiding difficulties for the past several
years could have been cured by zip-ties!
I should narrate the events of this busy day [May 8], just so everyone will know what it's
like being a retired man of leisure (hah!).
As usual, I got up, had breakfast, and handled e-mail for about an hour and a half
using my Lenovo laptop at the breakfast table.
Then, at 11, I had a teleconference with a client, during which I nibbled an early
lunch in order to get into the car and head for Atlanta at noon.
Heading down Highway 316, I turned on the radio and heard a news bulletin: a plane
had crashed on I-285 on the north side, killing a whole family
(requiescant in pace) and causing the city to prepare for the mother of all
traffic jams.
But in fact I did not encounter any traffic problems until I got to Briarcliff Road,
which was congested, but not intolerably so. My destination was Camera Bug, in the
Briar-Vista shopping center. Recall that the owner, Tim Nix, died unexpectedly just
before Christmas and the store had been closed up, frozen in time and clutter, ever
since. Today, Tim's sister helped me retrieve a telescope Tim had been trying to
sell for me on consignment (scroll down for details). Another customer came at the
same time to retrieve something, a tripod I think.
That was the end of Camera Bug; now there is no telescope dealer closer than Maryland.
Back in the 1990s, trips to Atlanta always meant visiting a number of specialized
stores (Photo Barn, Austin Electronics, Oxford Books, or wherever
my whimsy took me). Today I stopped by Showcase Photo and Fry's on the way home.
Reaching Athens, I took Melody out to dinner and then set up the Meade telescope
one last time. Placing it on the pier, I did a one-star alignment on Jupiter and
confirmed that everything worked well. Then, for only the second time since I bought it,
I used it in alt-azimuth mode. I removed the blue wooden wedge that had been on its
tripod for about a decade, set the telescope up as shown in the picture below,
took the picture (at ISO 6400 in dim twilight),
did a two-star alignment on Arcturus and Regulus, and confirmed that the telescope found and tracked
stars effectively. In fact, it finds stars more accurately in alt-azimuth mode than
in equatorial mode because it does not make any assumptions about polar alignment.
"Last light" consisted of showing Jupiter to Melody using an 18-mm eyepiece.
Then I put on the lens caps, took the telescope down, and moved it indoors.
Meanwhile I had been composing and circulating ads to sell it, especially
the blog entry below.
And that was May 8, 2015, the day of the plane crash and the telescope retrieval.
I note in passing that the next day, May 9, is the 45th anniversary of my first attempt to contribute
to astronomical research — a rather inaccurate timing of the transit of Mercury
from Valdosta State College Observatory with Edward Van Peenen II.
My trusty old Meade LX200 telescope needs a new home. It has just come back to me
after being tied up in the estate of Tim Nix (of Camera Bug), who was going to sell
it for me on consignment. [Update, May 12: It has found a
new home. I congratulate the new owner and wish him much enjoyment and success.]
This is a computerized "go to" telescope, meaning that
you can use it in alt-azimuth or equatorial mode and have it find celestial objects for you,
once you've aligned it on a star.
This is the very telescope that led me to write
How to Use a Computerized Telescope,
which describes it in detail.
A copy of that book is included with it.
It has an upgraded focuser and has recently been lubricated and collimated.
It also includes
this electrical improvement
and
this homemade equatorial wedge
(very sturdy, but only for latitude 34 north, give or take a couple of degrees).
For visual use, alt-azimuth mode (as shown in the picture) is easier.
Complete configuration:
Telescope on computerized fork mount
Upgraded internal focuser
8×50 finderscope
Tripod with homemade wooden equatorial wedge
Wooden plate to help position the telescope in alt-azimuth mode
Hand control box (some wear visible, but fully functional)
Spare cables for hand box and declination motor
Power supply (for powering it from 120-volt AC line) with cable
(Both 18-volt and 12-volt versions; I recommend 12 volts. Also included is a plug for a car lighter socket.)
Computer cable (for controlling with your computer if you wish)
Visual back for standard 1.25" eyepieces
Mirror diagonal for standard 1.25" eyepieces
Two inexpensive eyepieces (Bausch & Lomb 18mm and 30mm)
Instruction book (from Meade)
1 copy of How to Use a Computerized Telescope
1 piggyback camera mount, presently installed at rear top, can be moved
I am asking $500 for an instrument that originally cost four times that much.
This is a telescope for a serious amateur astronomer on a budget.
Purchaser must pick it up in Athens, Georgia as the cost of shipping would be
prohibitive.
I'm testing whether the video capture mode of my Canon 60Da is good enough for
serious planetary work.
(Jerry Lodriguss is the expert on this technique.)
While I do not yet have a strict side-by-side comparison
of the Canon to my ImagingSource DFK and DMK cameras, I do have these.
Both were taken with the Canon last night (May 6) through the 8-inch with a 3x Barlow lens.
Each is a stack of the best 80% of about 3900 video frames, taken at 60 fps
and exposing slightly more than a minute.
The pictures differ only in the processing decisions I made. The lighter picture
looks more like Jupiter as seen from a spacecraft. The darker one is a bit overprocessed
to bring out detail for study.
The advantages of the Canon are that I don't have to take a computer into the
field; the pixels are smaller, so I get a bigger image;
and focusing and framing are easy because I can use the camera like a regular
DSLR to get Jupiter into the field before switching to "movie crop mode" (full
resolution, small field) to take the video.
The disadvantage is that the Canon records compressed video, and the compression is
lossy. To save file size, some subtle differences of brightness and color from frame
to frame are ignored. I think, however, taking a large number of frames tends to
make up for this.
More deep-sky photography under the full moon.
I've been in "drop-everything-and-do-astronomy" mode lately, since
we are having a run of clear nights after a long, cloudy winter and
early spring.
Here you see the globular cluster M3 on the evening of May 4.
Stack of the best four out of seven 150-second exposures through my
8-inch EdgeHD telescope at f/10, Canon 60Da, ISO 1250.
I was testing the autoguider, which performed decently, although not
quite as well as two nights earlier because the air was less steady.
Again, zip-ties were a key part of the system; and again, the last
residual non-roundness was dealt with by deconvolution.
Late on the night of the 4th — actually, early on the morning of the 5th —
I decided to turn the telescope toward Saturn before taking it down for the night,
and the visual image at 100× was so steady that I couldn't resist doing some
video imaging. Stack of the best 80% of about 900 video frames, taken with my
usual setup at f/30.
Last night's astronomy session (May 3) was just a brief session to take video
images of Jupiter.
The satellites are Callisto (on the left) and Io and Europa (on the right, in that order).
Each of these is the best 80% of 3600 video frames taken
with my 8-inch telescope (C8 EdgeHD). The larger one used a 3× Barlow lens
for extra magnification; the smaller one did not. Although the air looked
worse than the previous night, the pictures came out better. A minor puzzle.
The other puzzle is that, with this telescope's Celestron AVX mount, I have to put
in the wrong time of day to get it to "guess right" about where the first
alignment star will be. To be specific, I have to give it a time that is 1 hour into
the future. I tried doing this several ways, but the easiest way was to set the clock
to daylight saving time and tell it standard time was in effect.
I've been beta-testing firmware for Celestron, and this might be a bug in a beta version,
but I thought I was currently running a public-release version.
The answer: Somehow there was a nearly 15-degree error in the
RA switch calibration, which cleared up when I did a full reset. What probably happened
is that at some time in the past I performed RA switch calibration while the time was
set wrong.
I think I figured out what went wrong with Windows 8.
All of use know that Windows 8 requires the user to learn quite a few new features, most of which,
once learned, are handy. But why are they puzzling in the first place?
Simple: Microsoft assumed all Windows 8 users would be familiar with Apple or Android smartphones.
If you're not into smartphones, a lot of things in Windows 8 are puzzling.
It's as if someone built a car whose controls
were designed with the assumption that the driver knew how to ride a motorcycle.
But a PC is not a smartphone and does not benefit from acting like one.
Windows 8.1 backtracked a little, toward normal PC behavior; Windows 10 will move back a long way,
thank goodness.
The heavy cable of my STV autoguider has been my nemesis; it applies a heavy weight
to the autoguider camera, and the direction of the weight shifts as the camera tracks.
Last night, I tamed it successfully with zip-ties (cable ties), which I cut (carefully)
at the end of the session.
It is easy to get zip-ties very tight so there will be no shifting.
(Light support from a Velcro cable tie wouldn't suffice.)
From now on, in my autoguider case there will be a supply of
zip-ties and a small wire cutter for cutting them.
The cables are tied to the dovetail that joins the autoguider to the telescope. Pulling
on the dovetail, they can't introduce differential flexure. They might benefit from being
tied to the telescope or mount in several more places, too, to bear the weight.
I still need to build a mechanical support for the autoguider camera — either another ring,
or more likely, something that will clamp to the dovetail and engage the camera's tripod
socket. After that is done, the heavy cables will still need to be zip-tied. For now,
zip-tying the cables seems to be just enough.
We had clear weather, for the first time in living memory, and I got the
telescope out last night (May 2). Besides a routine image of Jupiter, I also
got a good picture of the globular cluster M5 — which was right next
to the full moon!
M5 was photographed at the Cassegrain focus of my C8 EdgeHD. It is a stack
of three two-and-a-half-minute exposures. Autoguiding through the piggybacked
guidescope worked well once I zip-tied the autoguider cable to the dovetail to
prevent differential flexure. There was still a slight smear of about two or
three arc-seconds in right ascension, which I cleaned up with
motion-blur
deconvolution.
The final picture cleanly splits a double star with a separation of 6 arc-seconds;
the stars are about 3 arc-seconds in diameter. I think that is respectable.
Polar-axis misalignment can produce drift in any direction
A lengthy disquisition on this subject was here, until I realized I had
written the same thing, better, here.
Please click through to read it.
While thinking through a more
rigorous analysis of this, I thought of a a typographical nightmare.
To be precise, I started thinking about a formula that included a partial
derivative of a declination. That's right:
∂δ/∂t
If I had written out the rest of the formula, would anyone have
been able to read it?
Even at this advanced age, Melody and I were able to get out and have a "big-city date"
this evening, to attend the Cypress South
concert at the Fernbank Science Center planetarium.
(Sue Wilkinson and I were third-grade classmates in Valdosta.)
Melody even climbed the stairs to the observatory, where we viewed Jupiter
through the 36-inch telescope, the largest I have ever looked through.
This is a private web page,
not hosted or sponsored by the University of Georgia.
Copyright 2015 Michael A. Covington.
Caching by search engines is permitted.
To go to the latest entry every day, bookmark http://www.covingtoninnovations.com/michael/blog/Default.asp
and if you get the previous month, tell your browser to refresh.
Portrait at top of page by Krystina Francis.
Entries are most often uploaded around 0000 UT on the date given, which is the previous
evening in the United States. When I'm busy, entries are generally shorter and are
uploaded as much as a whole day in advance.
Minor corrections are often uploaded the following day. If you see a minor error,
please look again a day later to see if it has been corrected.
In compliance with U.S. FTC guidelines,
I am glad to point out that unless explicitly
indicated, I do not receive substantial payments, free merchandise, or other remuneration
for reviewing or mentioning products on this web site.
Any remuneration valued at more than about $10 will always be mentioned here,
and in any case my writing about products and dealers is always truthful.
Reviewed
products are usually things I purchased for my own use, or occasionally items
lent to me briefly by manufacturers and described as such.
I am an Amazon Associate, and almost all of my links to Amazon.com pay me a commission
if you make a purchase. This of course does not determine which items I recommend, since
I can get a commission on anything they sell.
If what you are looking for is not here, please look at
previous months.
