[CCD-CR] Vydany testovaci verze Munipacku

["Lukas Kral" <krall@troja....cvut.cz>]

Vazeni kolegove,

nejsem si jisty, zda se o tom v teto diskusi uz nepsalo, ale pokud se k Vam
tato zprava jeste nahodou nedostala, vezte, ze byly pred nedavnem oficialne
vydany nove testovaci verze fotometrickeho baliku Munipack - pro Linux i pro
DOS. Linuxovou verzi vyviji Filip Hroch a Rudolf Novak, DOSovskou verzi nyni
dale vylepsuji ja. Oboji viz
http://munipack.astronomy.cz/

O Linuxove verzi by mohl napsat spis Filip nebo Rudolf, ja o ni zatim jeste
mnoho detailu nevim.

Co se verze pro DOS tyce, zustaly v ni bohuzel puvodni starsi verze programu
provadejicich samotnou fotometrii (muniphot.exe) a identifikaci hvezd
(munimat.exe). Je to z toho duvodu, ze se nedarilo sehnat prekladac Fortranu
90 pro DOS, aby mohly byt prelozeny nove verze. Nicmene tyto programy jsou
funkcni a na druhou stranu jsou provereny lety praxe, byt napr. v rychlosti
se jiste nemohou merit s novymi linuxovymi verzemi (a nejspis i v dalsich
parametrech). Zustava zrejme jen (zde uz kdysi probirany) problem, ze nekdy
nejsou nalezeny velmi jasne hvezdy v poli (nad 32767 ADU). Chyba je zrejme
prave v puvodni stare verzi muniphotu a vyresi se az s prekladem nove verze.

Jeste par nejdulezitejsich zmen DOSovske verze (vse ostatni najdete v
prilozenem textu, prip. v kontextualni napovede primo v programu - F1):
- odstranen problem nastavajici na rychlych procesorech (Runtime Error 200)
- opraveny chyby programu munial (vyber hvezd a vypsani dat), ktere nekdy
zpusobovaly jeho pady
- provedeny upravy tak, aby bylo mozne zpracovavat uz primo FITSy i z jinych
kamer nez SBIG
- moznost volby velikosti aperturni clonky pro fotometrii
- moznost volby vystupu - diferencialni/instrumentalni magnitudy
- mnohe drobnejsi upravy zprijemnujici interface, hlavne pamatovani si
zadanych udaju

Zajimavy je podle me i novy (polo)automaticky rezim DOSovske verze, ktery
umoznuje prubezne zpracovani snimku uz behem pozorovani (v tomto rezimu se
zpracovavaji pouze nove pribyle snimky). Ja nyni pozoruji tim stylem, ze
disk pocitace v kopuli, kam se prubezne ukladaji nove snimky z kamery, si po
mistni hvezdarenske siti nasdilim jako dosovsky disk na pocitaci dole v
kancelari (volba Windows Pripojit sitovou jednotku), na tomto PC spustim
Munipack a snimky z kopule zpracovavam v automatickem rezimu takto pres sit.
Je to docela zazitek, kdyz vam svetelna krivka roste pred ocima v realnem
case...!

Pokud budu mit na podzim vice casu, chtel bych konecne vytvorit
plnohodnotnou verzi Munipacku primo pro Windows, mam uz i prekladac Fortranu
90. Zatim prosim testujte tuto DOSovskou (a samozrejme i Linuxovou) verzi -
pokud se vyskytne nejaky problem, napiste mi (DOS) nebo Filipovi a Rudolfovi
(Linux).

Prijemne pouzivani preje
Lukas



What Is New in Munipack for DOS 2.2 (Beta 2, Release 1)
=======================================================
Content:
1. Corrected Errors and Removed Problems
2. New Features
3. Warnings
4. Known Problems
5. Info for Users of Non-SBIG Cameras



1. Corrected Errors and Removed Problems
----------------------------------------

* runtime error 200 on fast processors (error of Borland CRT unit)

* bad algorithm in Munial, that calculated radii of circles when plotting
  stars - it caused errors (runtime error 201) with some fields

* Y2K problem with FITS files in Munial - was already solved in 2.1 
  version, but now also the new standard format of date in FITS files is 
  supported (YYYY-MM-DD). If the old format (DD/MM/YY) is found, dates later 
  than 80/01/01 are expected to be from 20th century (1900 is added), 
  earlier dates are read as 2000+DD.

* problems with reading FITS from other programmes in Munial - the 
  procedures for reading parameters (such as date, time, exp. time) from 
  FITS were completely rewritten, so now they are much more flexible. See 
  also part 3.

* a tiny error in Munigraph - when there were many bad values (99.999) in 
  the input data file, less good points were plotted than should

* Y2K problem of some older versions of CCDOPS camera controlling software
  (by SBIG), which produce .STx files with bad date (101 instead of 01).
  The year 101 is now read as 2001 by the konve program

Also many other small problems, that could cause instability of the program,
were solved (both in executables and batch files), but they do not affect 
the normal behaviour of the program, so it would be useless to write about 
them all here.




2. New Features
---------------

* a lot of "cosmetic changes" in the interface to be more user-friendly, 
  such as remembering of some settings, protection against owerwriting of 
  the data file with results etc. - I hope you will like it :-)

* new version of KONVE - program for conversion from STx to FITS - by Filip
  Hroch was used, now also PixCel255 format is supported and some errors
  were removed

* aperture size selection is now implemented. In the previous versions, the
  resulting magnitudes were measured in the smallest available aperture
  (a circle with radius of 2 pixels), which was too small for some 
  less-quality telescopes that produce diffuse stars, and therefore caused
  bigger data dispersion than was necessary.

  But note that choosing larger aperture can cause a small systematic shift 
  of the resulting magnitudes. I think the reason is that not all the stars
  has the same PSF, when using some telescopes, especially refractors).
  On my images, red stars are more diffuse than the white ones (I use a 
  photo-lens as the objective) and when I use a bigger aperture, the red
  stars seem to brighten, when comparing them with the whiter ones.

* "Automatic processing" - for users that are familiar with Munipack and 
  expect their data to cause no problems. Does STx->FITS conversion, dark 
  and flat correction, photometry and matching by one click

* "New images only" processing - very similar to the previous procedure, but
  takes only images, that were not processed yet. It can be useful, when you 
  are observing, want to preview the light curve almost in real-time and
  need to choose different stars than before.

* "New images only +" - the same as above, but does also choosing stars and
  plotting automatically (chooses data for previously selected stars). Use
  it when observing and want to preview THE ACTUAL LIGHT CURVE BY ONE CLICK!

* "Delete all files" in the "Utilities" menu - removes all temporary files 
  created by Munipack (files in the TEMP and FITS directories)

* "View FITS" (F3) now works - I wrote a simple FITS viewer (primitive, 
  slow, but it works). In future version it should be much better :-)

* Data filtering procedure in the Utility menu was added. See the interactive
  help (F1) for details

* Plotting the light curve is now interactive - you can adjust offset and
  range of the Y axis by pressing arrow keys. You can also draw horizontal
  grid lines by pressing "g".

* You can choose, if you want to get instrumental od differential magnitudes
  as the output

* when choosing variable and comparisons, a file with the same number as the 
  reference file is plotted, and you see which file it is

Use the interactive help (F1) when running Munipack to get further info on
each function.



3. Warnings
-----------
* Run Munipack ONLY from its home directory, otherwise it will
  have problems with paths!

* Use AS SHORT PATHS AS YOU CAN (I mean the path where you install Munipack
  and the paths to data, dark frame and flat field). If you use very long
  paths, there will be problem with the limitation of a DOS command length
  (some commands will not work at all).

* In "Setup", write the paths without backslash at the end - f.e. C:\DATA
  (and NOT C:\DATA\ !) 



4. Known Problems
-----------------
* In several cases, brightest star in field was not found during the
  photometry. The conversion to FITS was updated and works well, but the
  photometric routine (muniphot) stayed the old and it os probably the
  problem. I hope I will be able to compile Filip Hroch's new version of
  MUNIPHOT also for DOS soon.

* On one PC I noticed some graphics malfunction using the FITS viewer, but
  on all others it works well

* If you have too many DOS system variables defined in your autoexec.bat
  (such as paths etc.), there can be few space for system variables defined
  by Munipack, which can cause various crashes. To avoid this, remove all
  useless definitions from your autoexec.bat (sometimes they are added
  automatically when installing other programs...)



5. Info for Users of Non-SBIG Cameras
-------------------------------------
I have some great news for you - now you can use Munipack for DOS and it 
really works! :) The only condition is, that your software must allow you 
to save your images to FITS format (Flexible Image Transport System).

For the following explanation, let us expect that you have installed 
Munipack to C:\MUNIPACK.

a) Save images from your camera to FITS format, including the dark frame
   and flat field images.
b) Copy your FITS files to the C:\MUNIPACK\FITS directory, except the 
   dark and flat.
c) Copy your flat field to C:\MUNIPACK\TEMP and rename it to flat.fts
d) Copy your dark frame to C:\MUNIPACK\TEMP and rename it to dark.fts

Now you can begin with the photometry process. Start Munipack, skip the
"Setup", "List data" and "Convert data" commands and begin with "List FITS 
files".

Then you can use the following commands in the same way as SBIG users -
except the "Automatic" menu, which expects the input images in STx format.
You cannot use the functions in this menu at all, maybe in some future 
version...

Look to the manual of your camera to get some technical parameters such
as read noise or gain, which you must set by "Edit mphot.opt" before doing
the photometry. The preset values are for SBIG ST-7 camera.

NOTE: some camera-controlling software may create FITS files with the date 
or other parameters written in some non-standard format, so at least for the
first time, check if the JD in the resulting data file is correct!!!
It should be calculated as 
TS + 1/2 EXPT,
where TS means time, when the exposure was started, and EXPT is exposure 
time (so the JD means the mid-exposure time).

If you find a problem, contact me.


------------------------------------------------------------------------
Document written by Lukas Kral <krall@troja....cvut.cz>
Last Update: 11th July 2001