[Message Prev][Message Next][Thread Prev][Thread Next][Message Index][Thread Index]
[vsnet 768] VSNET home page: new items (Sep. 9) / VSSPlot
- Date: Tue, 10 Sep 1996 07:57:21 +0900
- To: binj@bao.go.jp, nova@nova.la.asu.edu, vsnet
- From: Taichi Kato <tkato>
- Subject: [vsnet 768] VSNET home page: new items (Sep. 9) / VSSPlot
- Sender: owner-vsnet@kusastro.kyoto-u.ac.jp
Following new materials have become available on VSNET WWW Home Page.
Some of materials are also retrievable via VSNET ftp
(ftp://ftp.kusastro.kyoto-u.ac.jp/pub/vsnet). All items listed here
are reachable via the "What's new" menu.
We appreciate your regular visits. Please enjoy!
General
-------
* 13th North American Workshop on Cataclysmic Variables
* VSNET light curves, updated on Sep. 5
* HAA/VSS ftp site update
* VSNET CV Circular 1996 Aug.
* VSNET Mira Circular 1996 Aug.
* VSSPlot ver. 1.10 by F. Farrell
Novae
-----
* Probable nova in Crux (Nova Cru 1996)
Dwarf novae
-----------
* International TOAD Watch Home Page (Howell)
* Superoutburst of HO Del (Aug. 31)
* Index to individual dwarf novae, updated
* Paper on AK Cnc and EK TrA (Mennickent et al.)
* HST observation of HS 1804+6753, new WWW page (Marsh)
* Superoutburst of AY Lyr (Sep. 6)
* On V337 Cyg
Symbiotic stars
---------------
* CCD image of FN Sgr
* CCD image of V1413 Aql = AS338
X-ray binaries
--------------
* Optical outburst of V635 Cas
R CrB and related stars
-----------------------
* CCD image of FG Sge
Regards,
VSNET manager group
----------------------------------------------------------------------
VSNET WWW home page http://www.kusastro.kyoto-u.ac.jp/vsnet/
----------------------------------------------------------------------
A useful tool for graphical data display and analysis of variable star
observations, "VSSPlot Ver. 1.10" has been provided by Fraser Farrell.
This program can be retrieved either from the VSNET Home Page (above) or
by anonymous ftp to the URL:
ftp://ftp.kusastro.kyoto-u.ac.jp/pub/vsnet/others/prog/vssplot
[the user's manual]
-----------------------------------------------------------------------
********************************************************
* *
* VSSPlot *
* Version 1.10 *
* *
* Graphical Data Display & Analysis Utility *
* for Variable Star Observers *
* *
* Copyright (c) 1996 by Fraser Farrell *
* All rights reserved *
* *
* PO Box 332, Christies Beach, SA 5165, AUSTRALIA *
* Internet email: fraserf@dove.mtx.net.au *
* *
********************************************************
CONTENTS
--------
1.0 What is VSSPlot?
2.0 Minimum hardware/software requirements & installation.
2.1 What VSSPlot needs....
2.2 Specific operating systems.
2.3 Installation.
3.0 Legal stuff.
3.1 The usual disclaimer in plain English.
3.2 Distribution.
4.0 Associated software.
4.1 VSSOBS
4.2 KSOLO & KSOLOREP
4.3 The VSOLJ text format.
5.0 Main functions.
5.1 Reading a DBF, RAW or TXT file.
5.2 Displaying a lightcurve.
5.3 Displaying a Phase Plot.
5.4 Period Analysis.
5.5 JD calculator.
5.6 Configure VSSPlot.
6.0 Program output.
6.1 Printing the display.
6.2 Saving the display to a PCX file.
7.0 Error messages.
7.1 "But it's not doing anything...."
7.2 Messages from VSSPlot.
7.3 Other errors.
8.0 Wish list and history.
9.0 Credits & acknowledgements.
1.0 WHAT IS VSSPlot?
---------------------
VSSPlot was originally written to read DBF files generated by the
VSS,RASNZ data entry program VSSOBS.EXE; and draw lightcurves, phase
plots & periodograms on my computer screen. VSSPlot can also send
its screen displays to a printer, or to PCX image files.
This version of VSSPlot also works with RAW files generated by the
AAVSO data entry program KSOLOREP.EXE; and with plain ASCII textfiles
structured to the VSOLJ standard. This should make VSSPlot more
useful to the worldwide variable star community.
VSSPlot was deliberately written to run on older computers - if I had
a spare $2000 I would buy a bigger telescope first!
2.0 MINIMUM HARDWARE/SOFTWARE REQUIREMENTS & INSTALLATION
----------------------------------------------------------
2.1 What VSSPlot needs...
* An IBM-XT compatible computer with 640kB RAM.
* A hard disk drive.
* A VGA or compatible monitor. Any modern monitor should work; but
text-only, CGA, EGA and Hercules mono monitors WILL NOT WORK with
this version of the program!
* DOS 3.3 or later.
* A 9-pin or 24-pin Epson-compatible dotmatrix printer; OR
a HP LaserJet-compatible laser/inkjet printer.
* DBF files generated by the VSSOBS variable star data entry
program (or AAVSO RAWs, or VSOLJ TXTs), containing *numerical*
estimates for the star(s) of interest.
VSSPlot will use a math coprocessor if it is installed. Note
that 486DX, 586, 686 and Pentium computers have a built-in
coprocessor. Other computers require installation of a separate
chip.
VSSPlot uses the VGA 640x480 pixels 16 colour mode for the simple
reason that this mode is universally supported by ALL modern
video cards/monitors. Writing code specifically to drive *your*
trillion-colour hyper-accelerated video card to its limits is
beyond my programming ability. Doubly so when manufacturers
expect me to pay big $$ for the BIOS and register details of
their products.
VSSPlot was written to run under the DOS operating system and
expects to be the only program running on the computer. Notes
for specific operating systems follow:
2.2 Specific operating system notes (further contributions welcome!).
MS-DOS 5.0 : VSSPlot may not run from the DOSSHELL task switcher,
depending on what other programs are loaded with it.
DR-DOS 6.0 : If VSSPlot is run from the TASKMAX task switcher,
DO NOT swap out while VSSPlot is drawing, printing graphics, or
reading/writing files. VSSPlot will fail with unpredictable
results.
Novell DOS 7 : TASKMGR task switching (or multitasking) has the
same caveats as for DR-DOS 6. Multitasking VSSPlot with other DOS
programs does work (on my computer); although I recommend you
obtain & install the free update file D70U15.EXE first.
Windows 3.1/3.11/WfW : The file VSSPLOT.PIF should be used to
run VSSPlot correctly under Windows 3.x. Note that VSSPlot runs
much slower in Windows than it does in DOS. Microsoft offer a free
upgrade file to convert Windows 3.1 to 3.11.
OS/2 2.1 and above : Run VSSPlot as a "DOS Full Screen Graphics
Application". Using less than a full-screen window may cause
problems during graphics operations.
Windows 95 : So far, VSSPlot has managed to crash almost half of
the Windows 95 systems tested. This is a Windows 95 problem, NOT
a VSSPlot problem - Windows 95 doesn't run DOS software too well.
You will need to "Restart in MS-DOS Mode" to run VSSPlot at all;
it does not run reliably in a DOS window. Additionally, the
Windows 95 Paint accessory does not read PCX files (see 6.2).
Complain to Microsoft :)
2.3 Installation.
VSSP-xxx.EXE is a self-extracting ZIP file. The "xxx" refers
to the version number; "101" = version 1.01, "110" = version 1.10,
etc. This version contains the following files:
VSSPLOT.EXE : The program itself.
VSSPLOT.PIF : PIF file for Windows 3.x operation.
VSSPLOT.ICO : Windows 3.x icon. Can be imported to Quikmenu too!
VSSPLOT.CFG : Configuration file.
VSSPLOT.TXT : You're reading it now....
To install VSSPlot, copy VSSP-xxx.EXE to the same directory as
your VSSOBS.EXE program (or KSOLOREP.EXE, or VSOLJ files), then
run VSSP-xxx.EXE. The VSSPlot files will be extracted, after
which you can run VSSPlot.
To run VSSPlot from DOS, type in VSSPLOT at the DOS prompt.
VSSPlot will run from DOS "menu" programs eg: Automenu, Quikmenu,
PCBoss, etc; and can be run from the task switchers in DR-DOS 6.0
and Novell DOS 7.
Windows 3.x users can install VSSPlot into Program Manager. Select
"File|New|Program Item", click "OK", then use the "Browse" function
to find and select VSSPLOT.PIF. Windows should then fill in the
"Command Line" box for you - leave all other boxes blank, and
click "OK".
The configuration file VSSPLOT.CFG is initially set for a
24-pin printer on LPT1, and a time zone of UT+9.5 hours. This
happens to be my setup, yours is probably different, so press F10
at the VSSPlot main menu BEFORE trying to do anything. Your new
setup will be saved for subsequent sessions.
VSSPlot sometimes uses the following temporary files:
VSSPLOT.DFT : For Discrete Fourier Transform data
VSSPLOT.$$$ : For temporary storage
These should be deleted by VSSPlot when it exits.
3.0 THE LEGAL STUFF
--------------------
3.1 The usual disclaimer.
OK, here it is in plain English: This program has been tested
and is believed to be free of harmful flaws. Because I can't
possibly test every combination of computer, operating systems,
software and accessories out there; be warned that VSSPlot might
do something unexpected that may harm your system. Or produce
silly and meaningless output.
So if you, your computer, your business, your research, your
family/friends/colleagues, or anything else connected with your
existence suffers loss or damage from your use of VSSPlot (or
inability to use VSSPlot) - IT'S YOUR PROBLEM, NOT MINE.
If your local laws prohibit disclaimers like this one, then
DO NOT USE VSSPlot. There, now wasn't that easier to read than
the usual screens full of legal obfuscation? :)
Incidentally, any trademarks and copyrighted names I happen to
mention in this document are the property of their respective
owners. In most instances it should be obvious who they are!
3.2 Distribution.
VSSPlot may be distributed non-commercially for personal or
research use; ie: VSSPlot is freeware. However, I will accept
donations in Australian $ or by International Money Order!!
Anyone wishing to distribute or sell VSSPlot for *commercial*
gain requires my written permission; and will need to pay me a
licencing fee.
I hereby assert copyright and intellectual property rights to
this software, and the use of the name "VSSPlot".
4.0 ASSOCIATED SOFTWARE.
-------------------------
VSSPlot does NOT have a data entry function as such. It reads files
generated by other programs. Therefore, you will need to have one of
these programs (or their output files) to use VSSPlot. Or at least
an ASCII file editor to make VSOLJ-type textfiles...!
4.1 VSSOBS.EXE.
This is the preferred data entry program for observations
reported to the Variable Star Section, Royal Astronomical Society
of New Zealand (VSS,RASNZ). General enquiries about the
VSS,RASNZ should be addressed to: PO Box 3093, Greerton,
Tauranga, NEW ZEALAND. International phone/fax + 64 75 410216.
VSSOBS generates Foxpro-compatible DBF files; and VSSOBS
can also sort, collate, & extract observations. Latest version
(at time of writing) is 3.3. For more information on VSSOBS,
contact its author (Ranald McIntosh) by email to:
"ranald@ihug.co.nz"; or by writing to: 80 Solomon Avenue,
Redwood, Christchurch, NEW ZEALAND.
Individual VSSOBS DBF files can be merged and prepared for VSSPlot
using the VSSOBS functions "Utilities|Join Database Files",
"Process Data|Filter Database File" & "Process Data|Sort Database
File". See the VSSOBS documentation for details.
VSSPlot will read both formats (pre-Version 2.8, and current) of
VSSOBS *.DBF files.
4.2 KSOLO.EXE & KSOLOREP.EXE.
The American Association of Variable Star Observers (AAVSO)
prefer observers use a pair of their programs, KSOLO and
KSOLOREP, to generate reports. General enquiries about the
AAVSO should be addressed to: 25 Birch St, Cambridge, MA 02138,
USA. International phone + 1 617 354 0484, or email to:
aavso@aavso.org.
KSOLO is used for data entry, then KSOLOREP is used to generate
the final reports as two files. Both programs are available
from the AAVSO's Internet sites: (Web) http://www.aavso.org
or (FTP) ftp.aavso.org/pub/ksolo/. At time of writing they are
distributed within two self-extracting files KPACK-1.EXE and
KPACK-2.EXE. You will need both files.
VSSPlot will read the RAW files generated by KSOLOREP. These
have names of the form: cccMMMYY.RAW; where ccc = observer code,
MMM = month (3 letters), YY = year (last 2 digits). This version
of VSSPlot will NOT read the RAW files output by KSOLO - although
both kinds of RAWs are structured textfiles, their internal formats
are different!
KSOLOREP RAWs can be merged using the COPY or XCOPY commands
in DOS.
4.3 The VSOLJ text format.
The Variable Star Observer's League of Japan (VSOLJ) use a
structured ASCII text file to report observations. Each
observation is contained in a single line of the form:
aaabbbbbbbbb_YYYYMMDD.DDDD_mmmm_ccccccccc
where: aaa = constellation (IAU abbreviation).
bbbbbbbbb = star name (1 to 9 characters).
YYYY = year of observation. First 2 digits optional.
MM = month of observation (1 to 12).
DD.DDDD = day & decimal day (in UT) of observation.
mmmm = magnitude to 0.1 mag, decimal point omitted.
May have a '<' before the number.
ccccccccc = Not always used - observer code & comments
_ = space. There may be more than 1 space between data.
For example, to report an observation of RY Sagittarii at 10.5
on 1996 April 17.62 UT, the line would read: SGRRY 960417.62 105
or SGRRY 19960417.62 105. To report an observation of V854
Centauri at <11.3 on 1995 October 3.01, with a comment about
haze, the line might read: CENV854 951003.01 <113 HAZE.
The VSOLJ normally use the first 3 characters of the "ccccccccc"
field for the observer code. VSSPlot will assume the first 3
non-blank characters in this field ARE the observer code. VSSPlot
also assumes ALL VSOLJ magnitudes are numeric, not "lettered" or
"step" magnitudes.
The file MUST be a plain ASCII text file; the lines must contain
only observations - not headings, comments, notices, etc. Many
wordprocessor programs will insert *special codes* into their
usual output files. Look for a "save as ASCII" or "save as text"
or "print to textfile" command in your wordprocessor, to force
a file to be plain ASCII. Most text editor utilities and many
shareware wordprocessors use plain ASCII files by default.
VSSPlot will read a *.TXT file in the VSOLJ format. The DOS
commands COPY or XCOPY (or a text editor) can be used to merge
VSOLJ text files.
5.0 MAIN FUNCTIONS
-------------------
VSSPlot is run from a menu using the function keys:
F1 - tells you a little bit about VSSPlot.
F2 - reads a DBF, RAW or TXT file.
F3 - displays a lightcurve, if numerical magnitudes for the star
were found in the file.
F4 - displays a Phase plot and mean lightcurve, if numerical
magnitudes for the star were found in the file.
F5 - does period analysis using a Discrete Fourier Transform, if
at least 20 numerical magnitudes were found in the file.
Assumes the DBF, RAW or TXT file has been sorted by time.
F9 - a Julian Day <--> UT calculator.
F10 - VSSPlot configuration & printer setup.
Q - quit VSSPlot.
The lowest 2 lines display your current printer setup, the name
of the next PCX file VSSPlot will make (see 6.2 below), the
current date & time, and the current Julian Day (based on your
Time Zone and the date/time). VSSPlot will understand both whole
and fractional-hour Time Zones.
5.1 Reading a DBF, RAW or TXT file.
This function scans a file for observations of a specified star
within a specified Julian Day or UT date range. Files will be
assumed to be in the same directory as VSSPlot, unless you include
the drive and path in the file name. The file extension describes
the type of file VSSPlot will attempt to read: .DBF for VSSOBS
files, .RAW for KSOLOREP files, and .TXT for VSOLJ files. If no
extension is specified, VSSPlot will assume a DBF file structure.
To avoid ambiguity, VSSPlot first asks for the constellation, THEN
the variable name. The constellation should be given as the
IAU-standard 3 letter abbreviation (eg: "CEN", "PAV", "TRA"). The
variable name is the usual letter, V-number, Bayer/Flamsteed, NSV,
or catalogue number (eg: "S", "RY", "V854", "KAPPA", "NSV12495",
"NOVA1996", "GSC8979.00902"). The name needs to be spelled exactly
as it appears in the file - it does not matter if you enter the
name in capitals or lowercase characters.
A running display of records scanned and matches found is
shown. If no matching observations are found, the file read
aborts with an error message. A match is defined as a record
containing an identical star/constellation name AND a time
between your specified JD limits. The UT date/times in VSOLJ-type
textfiles are converted to the equivalent JD by VSSPlot during the
scanning process.
Matching observations are stored in conventional RAM ( <=640kB ).
This version of VSSPlot ignores any extended (XMS) or expanded
(EMS) memory; ie, only conventional DOS memory is used.
I know some of you would like to load in EVERY observation ever
made of VW Hyi or Omicron Cet. It's not possible with this
version - but I'm working on it!
VSSPlot continues to store matching observations until it reaches
the end of the file, or available memory is insufficient. In
the latter event, scanning is aborted, but observations read so
far remain in your computer's memory. If your computer has about
600kB free memory before starting VSSPlot, then it can store
about 14000 observations.
If matching observations are found, the following info is
displayed:
Number of matching observations.
Earliest Julian Day (& UT date) found in matching observations.
Latest Julian Day (& UT date) found in matching observations.
Number of numerical estimates - this includes "fainter than
{value}" observations.
Brightest and faintest magnitudes found, if numerical
observations were found. "Fainter than" observations
are regarded as valid magnitudes by VSSPlot.
You then have the option of resetting the JD or UT limits to ignore
unobserved extremes; eg: if you were silly enough to select JD
0 as the earliest limit for *this* year's results!
Magnitude limits can be manually defined or you can just let
VSSPlot define the magnitude scale automatically.
"Lettered comparison stars" estimates (eg: "B(2)v(1)C" or "v=g")
cannot be plotted on a numerical lightcurve or phase plot. Such
estimates will be ignored by VSSPlot's processing. It is assumed
that a VSOLJ-type text file will contain ONLY numeric magnitudes.
This version of VSSPlot also ignores VSSOBS's "sky" & "class"
numbers, KSOLOREP's "comment codes", and anything after the first
three non-blank characters in the VSOLJ "comments".
5.2 Displaying a Lightcurve
By default, VSSPlot displays lightcurves as two 'panels' per
'page'; and initially plots all numerical estimates for the JD
interval represented by 'page one'.
'Page one' begins on that JD divisible by 100 which is equal to
or precedes your specified earliest JD. If you used UT, your
earliest UT is converted to a JD and rounded-off in the same way.
The total number of 'pages' is calculated according to the length
of your specified JD or UT interval, and the 'page magnification'
(see below).
Each 'panel' has a JD scale along its base. The UT dates of the
start & end of each 'panel' are written vertically on the 'panel'
sides. Note that the JD is used to derive the UT date, so the
actual UT time at that point is 12h00m.
The magnitude range is always set to be equal to (faintest mag -
brightest mag + 2). VSSPlot will auto-scale the magnitude range
if you do not specify magnitude limits.
Function keys available include:
PGUP - go to previous 'page' (inactive on page 1).
PGDN - go to next 'page' (inactive on last page).
HOME - go to first 'page'.
END - go to last 'page'.
INS - magnify 'page' 10x. Magnifications available are
1x (1000 days/page), 10x (100 days/page) & 100x (10
days/page). This should cater for everything from
Miras down to fast Cepheids I hope!
DEL - decrease magnification by 10x (inactive at 1x).
F9 - print the screen (see 6.1 below).
F10 - save the screen to a PCX file (see 6.2 below).
ESC - quit graphics mode.
Positive magnitude estimates are plotted as yellow "+", while
fainter-than estimates (optional display) appear as green "v".
VSSPlot searches all observations in memory to display plottable
estimates - on a slow computer and/or very large dataset, this
may take a while.
While VSSPlot is busy, the colour of the program title (lower
left on screen) changes from green to red. When plotting is
complete, it reverts to green again.
The current date & time is shown to the right of the program title,
and will be included on printouts & PCX images.
5.3 Displaying a Phase plot & Mean Lightcurve
You will be asked for an Epoch (Julian Day) and a Period (days).
VSSPlot then takes each observation, computes the absolute value
of the fractional part of: (JD - Epoch)/Period, and uses this to
plot the magnitude according to its Phase. If the period is
constant during your specified JD interval AND your Period is
close to the truth, then the phase plot will form a narrow strip
of points. If your period is wildly wrong, the Phase plot
appears as a broad band with no discernable maximum or minimum.
If your period was right for PART of the JD interval, you may see
both kinds of plot (best appreciated on a slower computer,
because you can watch VSSPlot work).
The mean lightcurve displayed below the Phase plot is based on
the arithmetical average magnitude for each 0.05 phase interval.
NB: a Phase plot is meaningless for non-periodic stars.
Function keys work as for Displaying a Lightcurve. Because
observations are retained in RAM, you can exit the Phase display,
then re-select it using a different Epoch and Period on the same
data. This should allow you to 'tune in' on a true epoch and
period fairly quickly.
While VSSPlot is busy, the colour of the program title (lower
left on screen) changes from green to red. When plotting is
complete, it reverts to green again.
One interesting fact to emerge from a Phase plot is the "working
hours" of the contributing observers. Try Period = 1 day.
5.4 Period Analysis
This function is still under construction, but it should give you
a hint at what VSSPlot is trying to do...!
Period analysis is done using Discrete Fourier Transforms (DFT);
a slow but effective method that can cope with the gaps and
irregularities endemic to variable star observations. The DFT,
like all Fourier Transforms, attempts to define the signal
(lightcurve) as the sum of simple periodic variations.
This version of VSSPlot assumes the DBF, RAW or TXT file is already
sorted into *time* order. VSSOBS has a function "Utilities|Sort
Database" which will sort into star order, with observations of
each star sorted into JD order. KSOLOREP does the same thing when
it processes KSOLO's RAW files. If you are using a VSOLJ-type
text file, you will need to sort it yourself - it does not matter
if the star names are out of order, as long as the times for each
star are in order earliest<-->latest within the file.
This version of VSSPlot uses JDs "as reported" (ie: geocentric),
because the extra calculation for Heliocentric Julian Days isn't
really worthwhile unless you are searching for periods of less
than one day in very long datasets.
DFT analysis requires "normalized" data. To improve calculating
speed, VSSPlot will normalize and overwrite the data already in
memory; thus making it useless for lightcurves, etc. You will
be asked if you want DFT to proceed (just in case you pressed
F5 by accident!). After a DFT, you should re-load the original
observations (or another star) from the DBF/RAW/TXT file.
VSSPlot needs to know 3 parameters for DFT analysis: the longest
period and shortest period (both in days) to test, and the number
of intervals to test (in the interval longest <--> shortest,
inclusive). You cannot test for a period longer than the timespan
of the observations in memory. VSSPlot will not attempt a DFT if
less than 20 numerical magnitudes are available.
During the analysis, the entire dataset is methodically
"scanned" for EACH period, and a relative 'signal strength' is
calculated. This process is very trigonometry-intensive and will
take some time on older computers without math coprocessors (see
7.1 below). Progress is indicated with a "% completed" display.
VSSPlot will also make use of a temporary file on disk at this
time.
Fourier Transforms can be presented as a FREQUENCY spectrum or a
TIME spectrum. Frequency spectra are good for identifying
'harmonics' and 'beat frequencies'; whereas time spectra are more
easily comprehended by non-mathematicians. VSSPlot produces
time spectra; plotting the 'relative signal strength' for each
period (in days) between longest & shortest. Closely spaced
intervals usually define a "spike" centered on the actual period.
A signal strength of 1 or less corresponds to "noise"; in general,
a signal is not considered significant unless it has a strength
of more than 10. VSSPlot will auto-scale the spectrum to suit the
signal of maximum strength.
As with the other graphics displays, you can print out the time
spectrum or save it to a PCX file.
DFT's have one disadvantage: not all of the "spikes" may be real
periods. They are ALIASES, caused by the time spacing of the
observations. Common aliases in variable star work correspond to
periods of one day, one week, one lunar month, or one year. These
could be caused (respectively) by all observers being at similar
longitude, observing only on a certain weeknight, avoiding Full
Moons, or from the star being unobserved/unobservable for the same
part of each year. Another common effect (when one or two observers
dominate the observations) is the splitting of each "spike" into
a cluster - each component separated by the average time interval
between that observer's estimates of that star.
The only cure for aliases is more observations, particularly at
odd hours or very early/late in a particular star's observing
season; or at least at non-regular intervals if you happen to be
the only observer!
Periodic stars (eg: Miras) should show one well-defined "spike"
corresponding to their primary period. Additional "spike(s)"
may be present - for example, R Centauri has a secondary period
half as long as its main period. Semiregulars will exhibit
several periods.
Cepheids may show a distinct "primary - 1/3 - 1/5" period
pattern. The biggest "spike" is the primary period, with other
periods at 1/3, 1/5, etc of the primary period. Eclipsing binaries
produce complex DFT spectra, depending on the actual shape of the
lightcurve, and the true period may not be immediately evident.
In fact, inspection of the lightcurve may reveal the period more
easily than the DFT spectrum! Quasi-regular variables (eg: dwarf
novae) also produce complex DFT spectra with numerous 'periods';
while irregular stars show little more than "noise".
The DFT is a powerful method for uncovering periodicy in a
lightcurve, but its results should not be used blindly.
5.5 Julian Day calculator
Provided for your convenience. Internally, VSSPlot works
exclusively in Julian Days. This function lets you convert UT to
JD or vice versa; very useful if you're like me and can't remember
where you put your JD calendar when you need it quickly!
5.6 Configure VSSPlot
Lets you set up your printer, and the default observation
file format. Printer may be a 9 or 24-pin dot matrix (Epson
compatible), or a laser/inkjet (HP LaserJet Plus compatible).
VSSPlot supports dedicated or networked printers attached to
LPT1 or LPT2.
Most old printers can emulate a 9-pin Epson, often by setting
DIP switches on the printer (if it is not an Epson). Virtually
all 24-pin printers (also called "LQ" or "letter quality") will
recognise Epson printer command codes. The HP LaserJet Plus is
emulated by nearly all laser, inkjet & bubblejet printers.
VSSPlot sends a "page eject" code after printing to a networked
printer. In most installations this should force the page to be
printed immediately it reaches the head of the print queue.
6.0 PROGRAM OUTPUT
-------------------
Printing and saving to PCX only work on the current graphics display.
If you want to print/save several pages, you must display each in turn
before printing/saving.
6.1 Printing the display
Press the F9 key while a graphic display is onscreen; and VSSPlot
will read the VGA video memory, format it to suit your printer,
and send it to your printer.
While busy printing, VSSPlot will change its graphics title
colour from green to red. When the title reverts to green, the
print operation is finished. If the printer is offline or not
ready (eg: out of paper), the print function will abort.
This version of VSSPlot prints at 120dpi to all dot-matrix
printers and at 150dpi to laser/inkjet printers. Each screen is
printed to fit on an A4-sized page in landscape mode. The
"function keys" display is removed from the printout.
6.2 Saving the display to PCX file
Press the F10 key while a graphics display is onscreen, and
VSSPlot will read the VGA video memory, convert it to a 640x480
16 colour PCX file, and store it on disk.
While VSSPlot is doing this, the graphics screen title colour
changes from green to red. When the title reverts to green, the
PCX file is saved. The "function keys" display is removed from
the image during the PCX conversion process.
VSSPlot simply numbers its PCX files sequentially (1.PCX, 2.PCX,
3.PCX,... etc) and saves them in the VSSPlot directory. When it
reaches 50000.PCX the sequence repeats, overwriting any pre-existing
1.PCX, 2.PCX, etc. VSSPlot stores its latest PCX file name in
VSSPLOT.CFG so that images from your previous session are not
accidentally overwritten. Renaming PCX files to something more
meaningful is up to you.
VSSPlot's PCX files will read successfully into NeoPaint, the
Windows 3.1/3.11 Paintbrush accessory, PaintShop Plus, VPic,
CShow, and many other graphics & DTP programs. My main reason
for choosing the PCX format was because it *is* so widely
supported - and free of licensing restrictions.
7.0 ERRORS
-----------
7.1 "But it's not DOING anything...??!
Well, I could have programmed some entertaining but useless display
to distract you while VSSPlot slaves away; but it would have slowed
things down even more. Seriously, some VSSPlot functions take time
to finish. I have deliberately coded some of them for maximum
compatibility rather than speed; personally I prefer a program that
works slowly on my computer to one that doesn't work at all!
The worst time hog is the period analysis routine. VSSPlot has to
execute a block of several trigonometric instructions many times
- in fact, (observations * number of test periods) times - to get
a result. If you want to try 10,000 periods on 20,000 observations
on your IBM-XT, run it overnight....
To speed this routine up, try: (1) fewer periods, (2) a faster
computer, (3) a math coprocessor, or (4) don't do it! :)
7.2 Messages from VSSPlot.
"Configuration file not found" - VSSPlot uses this to define your
printer and some internal parameters. Unfortunately this error will
crash VSSPlot under Windows 95. Avoid this by exiting Windows 95
completely and running VSSPlot from DOS.
"FILE NOT FOUND" - the file you want to read wasn't found. The
program assumes the file is in the same directory as VSSPlot. If it
is in another drive or directory, specify the full path and file name
eg: "d:\stars\mystuff\june1996.dbf".
"NO matching observations" - Two possibilities: (1) the star wasn't
found in the file (did you spell it exactly as filed?), (2) there were
no observations of the star in the specified time interval.
"NO numerical magnitudes" - Several VSSPlot routines will issue this
message. The program ignores any estimates defined in terms of
lettered comparison stars; ie: it may have found observations of your
star but there are no "V = number" -type estimates. VSSPlot needs
numeric magnitudes for plotting and analysis.
"Graphics startup error" - VSSPlot can't run your computer display
in 640x480 16 colour VGA mode. Older displays (text,CGA,EGA,
Hercules monochrome) will NOT WORK with VSSPlot. Newer displays
should not have a problem.
"DFT data has overwritten observations in memory" - You have done a
period analysis, and the normalized observations for this have
replaced the original observations in memory. You can continue
with period analyses on this data, or start afresh by selecting F2.
7.3 - Other errors
Any other scarey messages that appear (eg: "runtime error xxx at
yyyy:zzzz") should be reported to me. Include as much information
as possible - the exact wording of the message(s), what kind of
computer and operating system you have, etc; and what was VSSPlot
attempting to do? I need to be able to reproduce your problem
exactly in order to fix it in future versions.
8.0 WISH LIST for later versions (so far) & program history
------------------------------------------------------------
Perhaps the first wish should be a totally bug-free VSSPlot that can do
everything for you except actually OBSERVE the stars.... :)
Seriously, feel free to add to the list below. This version only
contains the features *I* want, and the features that other users
insist should be included! The wish list has yet to be programmed....
* Faster plotting, especially for multi-page lightcurves.
* Real "Letter Quality" printing on 24-pin printers. I've got such a
printer myself, and I want *nice* looking graphs!
* Use of EMS/XMS memory on computers with > 1MB RAM.
* Identifying individual observers' estimates. Who *does* make those
estimates that are consistently 3 mags out anyway??
* Daily/5-day/10-day mean magnitudes on lightcurves.
* Measuring elapsed time between any two points on a lightcurve.
* Period analysis by Phase Dispersion Method as well as DFT.
* User-specified PCX file names.
* A built-in PCX file viewer.
* Saving screens to files other than PCX format. NB: due to licencing
restrictions I will NOT be supporting the GIF format.
Versions:
1.10 - Fixed (I hope!) arithmetic bug in Period Analysis routine.
Added support for AAVSO and VSOLJ observation files.
Added UT dates to lightcurves. Added UT time inputs.
1.01 - First general release. Added support for laser (& compatible)
printers.
1.00 - Limited test release. Added Windows 3.x PIF & icon. Fixed RAM
overrun problem on huge DBF files. Improved the speed of Period
Analysis routine. FINALLY wrote this TXT file!
0.95 - Added "save to PCX file" option. Fixed divide by zero error in
Period Analysis routine. Upgraded DBF file reader to cope with
both known DBF file structures.
0.90 - Added 'pageing' and 'magnify' functions to displays. Added
Period Analysis routine. Improved lightcurve plotting.
0.8 - Added Phase Plot routine. Added dotmatrix printer output.
<0.8 - development versions.
9.0 CREDITS AND ACKNOWLEDGEMENTS
---------------------------------
* Ranald McIntosh, the creator of VSSOBS, for encouragement and slabs
of real data from the VSS,RASNZ variable star database.
* Justin Tilbrook, Peter Nelson, and others of my testing team who have
requested anonymity (modest lot aren't they?).
* Joe Grida, for letting VSSPlot loose on a variety of computers in
his shop for testing purposes. Fortunately they are still in
saleable condition! :)
* The team at AAVSO headquarters, for helping me out with the AAVSO
KSOLO program.
* Taichi Kato (Kyoto Uni), for help on the VSOLJ reporting format.
* Bob Beauchaine (wherever he is) for placing his "egaprtsc" laser
printing unit in the public domain. Assimilated and tweaked for
VSSPlot's use.
-------- end of VSSPLOT.TXT ---------------------------------------------
Return to Daisaku Nogami
vsnet-adm@kusastro.kyoto-u.ac.jp
VSNET Home Page