Mars and Siding Spring

Ben Davidson put out this video last evening:

https://www.youtube.com/watch?v=5E9SZce6pzM

I immediately emailed him back with a response, explaining everything that he was seeing (and speculating about) except one series of images which I'm still baffled by (although I'm leaning toward an optical effect).

No response, and as of his latest video this morning, he's either ignored my email or not seen it.

Anyhow, here is my email to him in the hopes that he'll see this blog entry:

Ben -- I've been an observational and theoretical astronomer for 35
years.  All of the things you show on your video are things I've seen
before except one.  Let me go through it all in some detail:

00:34 - 00:37 Those donut shaped objects are NOT caused by the
atmosphere, but rather internal reflections in the optics of the
telescope.  The atmosphere causes stars and other non-resolved objects
in the sky to twinkle, but that wouldn't be detectable in these
images.  The donut shape is actually an image of the primary mirror of
the telescope with the secondary mirror blocking the center (the
"hole" in the donut).

00:50 - 01:34 All optical effects.  Optical coatings create some very
beautiful effects like this.  Seen 'em a million times.,,

01:35 - 02:00 more about this below

02:00 - 02:37 this is the one you're most confused about, but it's
obvious to those of us who have used cameras on telescopes what this
is.  In order for the camera to see the background stars, a fairly
long exposure needs to be taken (in this case I'd guess it's about one
second).  The camera, of course, is in "live" mode so it's just
shooting pictures continuously and displaying them -- but not at video
rates otherwise we wouldn't see those background stars.  The next two
frames that distort the image of Mars is the telescope slewing while
the exposure is being taken.  This causes the streaks and blobs that
you see.  While the telescope is slewing, there really isn't anything
interesting to see so they cut the feed.  I've attached some data
images ('boops') I've taken of the Jupiter system (for my photometry
work) where I've bumped the telescope while an exposure is being
taken.  As you can see -- the same "mysterious" extensions.  Jupiter
is overexposed but you can see the effect on the moons.

boops

Here is more proof.  Look at the attached image named 'mars3'.  These
are two images taken from your video tonight.  In both images, there
is a 'TOP' blob, and a 'BOTTOM' blob.  First notice that the distance
between the TOP blob and BOTTOM blob is the same in both images --
indicative of the same exposure time.  Also notice that the position
of the BOTTOM blob on the left image is in the same place the TOP blob
on the right image.  This means the telescope was slewing at a
continuous rate (as expected).

mars3

So no big deal -- I'd probably do the same thing and show people
something else other than the telescope slewing.  I hope my
explanation makes sense.

The one that sorta has me baffled is the image from 01:35 to 02:00.
The bright object in the middle of the field of view not only is
flickering like I haven't seen anything astronomical flicker, but it's
also moving against the background stars.  In the short time span that
this apparently covers, we would NOT see Mars move against the
background stars.  I can't account for the flickering at all, except
to say that maybe it's some kind of off optical effect (it usually
is).  Do you have any information on how this video was taken????

Hope this clears up most of it.

Great work.

Peace, CL

IRIS MgII measurements 2014-09-09

From 2014-09-09 05:48:07-07:44:33:

Figure 1: IRIS view of the sun at 2796A

Figure 2: The MgII k (left) and h (right) spectral lines

Third set of IRIS data

Some new measurements from data taken by IRIS on 12 September 2014

I've added an additional column to the data.  Column 11 is a measurement of the "power" in the given spectral line.  Here's a plot showing the MgII k and h power.  The x-axis is the power and the y-axis is the number of vertical arcseconds from the center of the Sun.

The total power corresponds as expected to the shape of the spectral lines: less power where the spectrum is dark, more power where the spectrum is bright ---

k (left) and h (right) MgII spectral lines

Although now apparent when looking at the image, the plot of total power shows that the h lines are typically brighter than the k lines.

More IRIS Data

I took some more recent data and ran it through my code to make these measurements and I'm very happy with the results.  I made a small modification to the peak finding code that made a huge difference in terms of positive detections.

Anyhow, the latest measurements are from data taken in 'Sit and Stare' mode on 2014-09-19 05:17:12-07:14:54.  This calibrated data is marked 'OBS 3860608353' and is available here.

My measurements are here, which is a zip file containing the data values.

IRIS Measurement Data At Last!

The first set of IRIS MgII h and k spectral feature data is available here.  The ZIP file is here.

The input calibrated data can be accessed here.

Measurements of MgII h and k lines

The singly ionized Magnesium h and k lines are a source of strong ultraviolet emission in the spectrum of the sun.  Each of these lines has features in their profiles that have been identified as k1r, k2r, k3, k2v, k1v, h1r, h2r, h3, h2v, and h1v.  The measurement of their wavelengths, intensities, and spacial locations is the purpose of this project.  The k1r, k1v, h1r, and h1v features are not included in this current project.

We start with a portion of the sun that is being observed in 'Sit-and-Stare' mode (see Figure 1) at a particular time.  Different data sets may have a different time interval between successive images/spectra.  In this figure, the thin black vertical line is the entrance slit for the spectrograph.  Notice that the slit intersects a sunspot about 2/3 of the way from the bottom of the image.

Figure 1

The spectrograph produces spectra at many different wavelength bands of interest, one of them being a band that covers the Mg II k and h lines, which are at approximately 2796.4A and 2803.5A respectively.  Figure 2 shows a typical h spacial-spectrum.  The x-axis corresponds to wavelength and the y-axis correponds to spacial location along the slit.

Figure 2

Figure 3 shows the identified spectral line features.  The red, green, and blue points identify the h2r, h3, and h2v features, respectively.  The x-axis is wavelength (in angstroms) and the y-axis is the number of vertical arcseconds away from the center of the sun.  Compare this to the input data (Figure 2) and you'll see that the algorithm is generally doing a fairly good job at feature identification.

Figure 3

Figure 4 shows a typical line profile for the h line, identifying the three features that will be measured.  Figure 5 is the same data, but zoomed in to show just the spectral features and their calculated pixel positions.

Figure 4

Figure 5

The algorithm scans the profile and calculates the difference in intensities between adjactent points.  When the difference flips between a positive difference and a negative difference, a peak has been found.  When the difference flips from negatice to positive, a trough has been found.  The exact location where the difference has a value of zero is the location of the peak or the trough.  Figure 6 shows the differences.  The black dots indicate where the differences are zero.  There is a 1/2 pixel offset between Figure 5 and Figure 6 values because the intensity differences are measured from the "centers" of the pixels while the intensities themselves are measured from the "edges" of the pixels.

Figure 6

Because of various factors, not all line features can be measured.  These features disappear, for example, in the location of the sunspot.  This algorithm is looking for a peak-trough-peak sequence and will fail proper identification if this sequence is not encountered (for instance if it finds a trough before the first peak, or if just a single peak is found).

Once the pixel location of the peak or trough has been calculated, it is translated into an spacial offset (in arcseconds away from the center of the sun) and wavelength (in angstroms), by using the "CRVAL1" and "CDELT1" metadata values provided in the FITS headers.  The intensity of the peak or trough is also measured and reported.

This is done separately for the k and h lines.

Currently, the software will only work with 'Sit-and-Stare' input data.

Data Example

The data is organized into separate text files corresponding to spectra taken at each raster epoch.  For example, if there are 144 raster images, then there will be 144 separate text files with the following type of data:

6912.522659 30 120.884 144.357 2803.386 1519.708 2803.537 1114.734 2803.614 1178.763 31539.500
6912.522659 31 120.884 144.523 2803.389 1504.800 2803.530 1235.172 2803.615 1337.240 31642.000
6912.522659 32 120.884 144.690 2803.406 1400.550 2803.518 1242.861 2803.607 1341.820 31097.500
6912.522659 33 120.884 144.856 2803.397 1384.531 2803.487 1195.102 2803.596 1341.308 30745.000
6912.522659 34 120.884 145.022 2803.394 1409.741 2803.474 1225.024 2803.580 1337.178 30260.750
6912.522659 35 120.884 145.189 2803.391 1395.913 2803.494 1173.014 2803.596 1392.362 29817.250
6912.522659 36 120.884 145.355 2803.389 1359.357 2803.485 1160.290 2803.597 1352.109 29595.250
6912.522659 37 120.884 145.521 2803.372 1312.136 2803.493 1140.555 2803.593 1333.213 29225.250
6912.522659 38 120.884 145.688 2803.358 1268.717 2803.488 1131.498 2803.601 1314.044 29207.000
6912.522659 39 120.884 145.854 2803.354 1316.627 2803.457 1084.482 2803.596 1335.140 29375.250
6912.522659 40 120.884 146.020 2803.350 1356.074 2803.476 991.637 2803.601 1320.630 28961.250
6912.522659 41 120.884 146.187 2803.342 1408.039 2803.470 968.719 2803.603 1325.508 29083.750
6912.522659 42 120.884 146.353 2803.343 1409.933 2803.469 967.708 2803.612 1352.689 29782.750
6912.522659 43 120.884 146.519 2803.344 1391.391 2803.465 1012.583 2803.611 1395.858 30408.750
6912.522659 44 120.884 146.686 2803.334 1380.299 2803.457 1033.055 2803.612 1392.879 30696.000
6912.522659 45 120.884 146.852 2803.355 1393.129 2803.467 1131.581 2803.593 1397.156 31680.500
6912.522659 46 120.884 147.018 2803.350 1509.836 2803.470 1211.112 2803.609 1413.840 32370.000
6912.522659 47 120.884 147.185 2803.350 1448.126 2803.471 1135.210 2803.612 1331.672 31034.500
6912.522659 48 120.884 147.351 2803.341 1351.033 2803.499 1047.807 2803.615 1278.129 29607.000
6912.522659 49 120.884 147.518 2803.350 1321.970 2803.485 1096.239 2803.618 1293.102 29079.750
6912.522659 50 120.884 147.684 2803.356 1274.051 2803.460 1117.601 2803.462 1130.555 28155.250
6912.522659 51 120.884 147.850 2803.358 1190.512 2803.453 1041.789 2803.478 1055.543 26629.000
6912.522659 52 120.884 148.017 2803.368 1084.548 2803.455 971.335 2803.467 978.182 24838.750
6912.522659 53 120.884 148.183 2803.399 1035.587 2803.500 984.569 2803.561 1060.511 24682.750
6912.522659 54 120.884 148.349 2803.383 1124.145 2803.387 1111.945 2803.420 1152.975 25633.500
6912.522659 55 120.884 148.516 2803.395 1224.889 2803.485 1129.833 2803.529 1163.413 26514.750
6912.522659 56 120.884 148.682 2803.393 1250.542 2803.437 1184.945 2803.467 1204.757 27016.500
6912.522659 57 120.884 148.848 2803.368 1227.481 2803.479 1162.977 2803.587 1244.195 26974.500
6912.522659 58 120.884 149.015 2803.378 1213.820 2803.453 1178.464 2803.480 1184.556 26797.500

The files will be named in the following format:

obs_AAAAAAAAAA_BBBBB[h/k].txt

where the A's correspond to the IRIS observation tag, and the B's correspond to the raster image number starting at '00000'.  'h' or 'k' corresponds, of course, to the spectral line.

Column Descriptions:

1:  Julian Date (JD - 2450000) of observation
2:  Raster image row number
3:  X position (arcsec) from center of Sun
4:  Y position (arcsec) from center of Sun
5:  Measured k/h2v position (angstroms)
6:  Measured k/h2v intensity (data units)
7:  Measured k/h3 position (angstroms)
8:  Measured k/h3 intensity (data units)
9:  Measured k/h2r position (angstroms)
10: Measured k/h2r intensity (data units)
11: Total power

IRIS is a NASA small explorer mission developed and operated by LMSAL with mission operations executed at NASA Ames Research center and major contributions to downlink communications funded by the Norwegian Space Center (NSC, Norway) through an ESA PRODEX contract.

IRIS Data Products -- Almost

I was about to release my first set of IRIS data products and realized that I need to add a bit more data to the final product, AND most importantly I need to write up a document describing how I created the product.

I'm very excited about getting this data out, but it'll take a little more time to get it out.  The cool thing about being an independent researcher is that I can make my own timeline.  This way I can be satisfied with what I do rather than trying to finish on a schedule with only decent results.

Anyhow, here's a sample of the current data product:

20 -145.7258 2803.3411 7178.8491 2803.5112 4874.6631 2803.6743 6886.1235
21 -145.3931 2803.3596 6523.7871 2803.5061 4834.3359 2803.7090 6792.2446
22 -145.0604 2803.3843 7062.9453 2803.5510 5101.1875 2803.5935 5201.3462
23 -144.7277 2803.3965 7286.9478 2803.6162 4424.9214 2803.7390 5431.4185
24 -144.3950 2803.3757 7256.7803 2803.5818 4509.8062 2803.7209 5524.3799
25 -144.0623 2803.3625 7086.9365 2803.5273 3827.3635 2803.6963 5804.1982
26 -143.7296 2803.3557 7059.0205 2803.5303 3544.4858 2803.6841 5574.2578
27 -143.3969 2803.3569 6335.1782 2803.5208 3239.6372 2803.6741 4954.3608
28 -143.0642 2803.3591 5240.8706 2803.5232 2873.1086 2803.6799 4440.0874
29 -142.7315 2803.3716 4632.0991 2803.5444 2698.4321 2803.6804 3886.6016
30 -142.3988 2803.3906 4446.7124 2803.5581 2223.5918 2803.6794 3593.5039
31 -142.0661 2803.3838 4326.3809 2803.5452 1997.5411 2803.6750 3315.3245
32 -141.7334 2803.3789 4018.1577 2803.5532 2099.9224 2803.6716 3081.2000
33 -141.4007 2803.3850 3486.9858 2803.5474 2131.6660 2803.6533 2877.7717
34 -141.0680 2803.3813 2922.0916 2803.5286 1822.3867 2803.6533 2578.5029
35 -140.7353 2803.3828 2700.9463 2803.5337 1802.7117 2803.6450 2478.1128
36 -140.4026 2803.3909 2765.1562 2803.5225 1982.9794 2803.6292 2619.6548
37 -140.0699 2803.3962 2817.1248 2803.5288 2110.8174 2803.6179 2679.0667
38 -139.7372 2803.4019 2863.1106 2803.5273 2071.1104 2803.6245 2625.1670
39 -139.4045 2803.4031 3289.1108 2803.5190 2267.6670 2803.6216 2802.0623

The first column is the y-pixel location.  2nd column is the distance, in arc seconds, from the center of the sun along the y-axis of the image raster.  The next two columns is a pair: the first value (column 3) is the wavelength of the peak value of the MgII k 2v line, and the second value (column 4) is the intensity of that peak.  Columns 5 and 6, and 7 and 8 are the same, except they are for the MgII k 3 and MgII k 2r spectral features.

I need to add the distance from the center of the sun along the x-axis, and I need to add the UTC of the observation.  At the moment I'm only doing this for "sit-and-stare" data which means that the slit isn't scanning the image raster but is stationary.  I still need to work out how the scanning works as far as the data goes.  That'll come in a later version of the data product.

Here's a picture and a plot of the data that I'm working with:

Figure 1: The bright MgII h and k lines

Figure 2: A profile plot through the spectrum image above clearly showing the MgII h and k lines

IRIS is a NASA small explorer mission developed and operated by LMSAL with mission operations executed at NASA Ames Research center and major contributions to downlink communications funded by the Norwegian Space Center (NSC, Norway) through an ESA PRODEX contract.

Buttons

Nearly from day one that I owned my Roland Fantom X8, I've had trouble with the buttons.

I takes me a while to realize certain things, and this is an example.  Sit back for a short story:

When the first button went bad on my board, it was a button that was mounted on a small circuit board that was easy to replace -- and a mere $62.00.  Another two buttons on another circuit board failed.  I replaced that board for $160.00.  Yet another button failed on yet another circuit board and I replaced that for $140.00.  Then another button failed on the first circuit board -- the one that I had just replaced several months earlier.

So it finally dawned on me the other day that I might be able to just replace the buttons instead of the whole circuit boards.  I did a little online research and found that not only is this "button problem" a known problem, but most owners do just replace the buttons themselves.

So, with a soldering station set up, I went to work and replaced bad buttons with good ones.  A continuity test with a multi-meter confirms whether a button is working correctly or not.

Bang, bang, bang and I have a good-as-new circuit board with all the buttons working!

So now I hope to do this same thing with some other bad buttons I have on other circuit boards.  With all of the buttons replaced, I'll finally be able to get into actual sound synthesis.

Here's a picture of some of the bad buttons.  They're about 6mm wide.

Figure 1: Bad Buttons