Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://curl.sai.msu.ru/mass/download/doc/dimm_soft_description.pdf Äàòà èçìåíåíèÿ: Sun Mar 23 19:35:51 2008 Äàòà èíäåêñèðîâàíèÿ: Mon Oct 1 20:17:54 2012 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: ï ï ï ï ð ï ð ï

Turbina-core(D) program Versions 2.3 ­ 2.25
V. Kornilov March 23, 2008

1

Installation

Requirements to the computer and OS
This program works under the Linux OS system with a kernel newer than 2.6.8. This is needed to provide a smooth work with IEEE1394-based digital cameras such as Prosilica EC650, for example. Although the program can b e adapted to practically any IEEE1394 camera, the current version uses such IEEE1394 camera p ossibilities as the supp ort of Format 7 and of ROI (region of interest). For the right choice of the cameras, the information accessible at http://damien.douxchamps.net/ieee1394/cameras/index.php can b e used. However, our studies show that Prosilica EC650 has very good p erformance to use as DIMM image detector. Camera has read-out noise ab out 10e- only and can provide frame rate ab out 200 fps with ROI 100 â 50 and exp osure 4 ms. The computer must provide a sufficient p erformance to process video stream from the camera. We tested the 2GHz Celeron and 3GHz Pentium-IV machines and found that the average CPU loads were 20-25% and 10%, resp ectively. Note, that for our MASS/DIMM (installed near Kislovodsk), we use VIA EPIA computer with VIA C3 800MHz processor. The machine provides effective work DIMM and MASS software and few additional services with average CPU loads less 20%.

Drivers and library installation
To work with cameras via IEEE1394 bus, the drivers included in the standard OS distribution are needed. Usually, the low-level drivers ieee1394 and ohci1394 are loaded by the OS at b oot. The drivers raw1394 and video1394 can b e loaded by the hot-plug service or manually b efore running the program, as well as at the system b oot. Two devices used by program are: /dev/raw1394 for camera control and /dev/video1394 for video stream input. If the devices are absent in system, issue the following commands manually under root: mknod -m 666 /dev/video1394/0 c 171 16 mknod -m 666 /dev/raw1394 c 171 0 Considering that the program uses the latest but still unstable release of raw1394 and dc1394 libraries, these libraries are placed in a non-standard way to avoid any confusion with another software. 1


Directory structure
For the program compilation, the source files can b e placed in the any directory. Source files include the directory iidc containing camera interfaces and the directory extra containing the compiled libraries libdc1394.a and libraw1394.a and the sources for their compilation. Latest versions ( from 2.09 ) contain the directories infras and server, to where shared with other our pro jects modules were moved. During the execution, the program uses absolute paths for its input and output files. We plan early that the Turbina-core(D) program will b e located in the /opt/turbina/bin directory. At the present day, the DIMM software is placed in separate directory /opt/dimm/ . It is recommended to include the path to this directory in the user's .profile or .bashrc files. The sub-directories /opt/dimm/data/out for the output data file, /opt/dimm/data/log for the output log file yymmdd-dimm.log and /opt/dimm/etc for the configuration file must b e created. The output data are written in the file named yymmdd-dimm.stm and located in the sub-directory data/out/. The structure of the output file is describ ed b elow. Additionally, the directory /opt/dimm/data/ contains the data/, images/ and spectra/ sub directories. The images/ serves as a place for output images, generated by the program. The data/ and spectra/ directories contain the files needed for preat routine.

Running the program
Starting from the earlier versions). on-line work with Program allow h c cfg b a d i name p numb -- -- -- -- -- -- -- version 2.4 of the program, the executable file is named dimm (pro test in The name Turbina is retained for the whole software pro ject enabling the MASS/DIMM instrument and off-line processing of its data. s few options in command line. They are listed b elow short help another configuration file cfg than default turdimm.cfg, put program into background after start (demonize it), automatic initialization on start, debug mode on, Internet machine name differing from default "0:0:0:0", p ort numb er differing from default 16200

The debug mode p ermits to run the program even when the camera is disconnected. In this case, artifical frames will b e generated with known parameters of the images. For example: ./dimm -c new.cfg -d There is the p ossibility to redirect program standard output in the pip e for processing program such as preat: ./dimm -d | ./preat To suppress the wide program output, the redirection is allowed: ./dimm -d > /dev/null In the real usage as part of ASM (Automatic seeing monitor) system, the program must b e run in the background mode: 2


./dimm -b In this case standard and error outputs redirected automatically in /dev/null and program acts daemon role. At the start of the program, the camera(s) is detected automatically and its identification numb er (UID) is displayed. The right numb er must b e entered in the configuration file, otherwise an error message will b e issued. Rememb er that the program is only controlled via socket connection b ecause it is designed for a remote control either by supervisor or another program. To control the program manually, the simplest way is to connect to it via the standard telnet utility. For example, connection from the same machine looks as: telnet localhost 16200 where the 16200 is the program server p ort numb er. Then, the program control is p erformed by sending a Sup ervisor-format command, including 1) command ID numb er, 2) command itself, 3) commands keywords with or without value (see Sect. Control of the program through socket). To terminate the program execution the command quit must b e used.

2

Initial conventions

Coordinate system
It is supp osed that the digital camera is installed on the MASS/DIMM device in correct orientation. The horizontal lines (rows) must b e parallel to the line connecting the DIMM ap ertures, i.e. to the flange of the device. Then the longitudinal differential motion corresp onds to the horizontal (along CCD rows) direction and the transversal differential motion corresp onds to the vertical (columns) direction. In the program the horizontal and vertical coordinates are designed as x­ and y­ coordinates. Pixel coordinates such as the optical center p osition or common images center are always written in the order first x, then y. To acquire more frames during basetime, the vertical size of the working subframe must b e minimized. It is p ossible if two images are placed along horizontal line by correct alignment of the DIMM mirrors in the device. In the program, an image with a smaller x­coordinate is always called "left", the other image is "right". Whether this corresp onds or not to the image orientation on a computer screen, dep ends on the image viewing software. In the output data, the coordinate with fractional part 0.5 corresp onds to the center of a pixel. So, the coordinates 0,0 are the coordinates of the top left corner of the top left pixel on the CCD frame.

Measurement timing
There are several time intervals in the program: exp osure -- one frame exp osure duration in ms, frame rate -- numb er of frames taken during 1 sec, basetime -- a time interval for elementary calculation of the differential motion variance, accumulation time -- the duration of one complete seeing measurement.

3


3

Measurement algorithms

The Turbina-core(D) program is constructed as a server-typ e program. If no activity, it waits for a request from client-typ e programs such a Sup ervisor. After such a request is received and checked, the program executes the requested function and waits for the next request. Any p ossible errors don't terminate the program's work, only the current function terminates.

Figure 1: The structure of the dimm program. Yellow modules form logical chain, pink modules are different calculating procedures, cyan ones -- basic infrastructure. When the command INIT comes, all program's ob jects needed for the measurements are created, the connection with the camera is established and the camera is initiated. The input configuration file is op ened and read. The output files are either op ened or created, dep ending on the current time. This p ermits to run the Turbina-core(D) continuously for many days and to generate separate output files for each night. The command PARK induces the following sequence of actions: all files are closed, the camera b ecomes free and the program's ob jects are destroyed. In the parked state, the program can only resp ond to the commands INIT, GET STATUS, GET IDENT and GET ERROR After the initialization, few measurement modes are supp orted: Test mode, Centering mode and Normal mode. Beginning from the 2.05 version, Raw mode was added. Additionally, latest version supp orts Pictures mode and Estimation mode.

Test mode
Test mode (Field Test) works with full-frame CCD images. Its functions are 1. to estimate the field ap erture center p osition with resp ect to absolute CCD coordinates,

4


2. to estimate the field ap erture radius RA needed for sub-frame selection in the centering mode, 3. to calculate roughly the separation S ep b etween sp ots provided by the DIMM mirrors, using the double image of the field ap erture, 4. to calculate main parameters of the CCD: the conversion factor (in electrons/ADU) and the readout noise (in electrons). Before running the mode, the telescop e must b e p ointed on a daytime sky or to a diffusely illuminated ob ject. If the illumination is insufficient, the error "Not enough light" (code 621) will app ear and the measurement will break. Otherwise, the program sets the exp osure time optimal for the actual light level and does the measurement. Some warnings may b e issued to draw attention to the bad alignment of DIMM mirrors or to inappropriate parameters in the configuration file. The configuration file must b e edited to enter the determined field ap erture parameters, so that no more tests will b e needed at each program start. The test mode writes the CCD image to the file "fieldtest.fits" (over-writes the same file) to document or examine the view of the field ap erture (see Fig. 2).

Figure 2: The view of the field ap erture on the frame obtained during Field Test. The red b ox indicates the frame defined for Centering mode, the red cross marks the optical axis p osition

Centering mode
The centering mode uses the p osition and size of the field ap erture sp ecified in the configuration file. In this mode, the subframe fully covers the double image of the ap erture and has dimensions 2RA â (2RA + S ep), i.e. approximately half of the full CCD frame. This mode is intended for the following tasks: 1. to check the telescop e p ointing (if the target star is in the ap erture or not), 2. to determine the p osition of the common center of star images relative to the optical center of the ap erture, 5


3. to determine the exact separation b etween the images. This data can b e used to correct telescop e p ointing and must b e used by the main (normal) mode to define the ROI (stars b ox) p osition. Usually, the centering mode uses the same camera settings (gain and exp osure) as the normal mode, but different settings via configuration file are supp orted. The accumulation time in this mode is ab out few seconds only. Using a frame rate of ab out 60 -- 70 fps, a good averaging of the star's p osition is achieved. The first stage of the algorithm is the calculation of the mean bias and its variance. Although these values can b e obtained from one frame, the values averaged over several frames are used. The threshold for ob ject detection is regulated by the parameter Operations/Centering/ ThresholdFactor. The two brightest detected ob jects are considered as images of the target star if their total signal is greater than Operations/Centering/MinObjectFlux. During the execution, the error "No two star images" (code 620) may app ear. The error breaks the execution and returns ERROR=ERRFAT. Few warning messages describ e the mismatches b etween the measured and desired images configuration. The centering mode (over-)writes the CCD image to the file "centerframe.fits" (see Fig. 3).

Figure 3: The view of the star images on the frame obtained during Centering Mode. The red b ox corresp onds to the CCD frame used in the Normal mode, the central part of the b ox is used for the star images measurements, the left and right parts are intended for bias control

Main mode (Normal mode)
The normal mode is the main mode of the measurements in the DIMM channel of the MASS/DIMM device. The result of this mode is a seeing value (after additional calculations). The mode can b e launched only after successful completion of the centering mode, as far as it uses the results of the centering mode. It is p ossible to run this mode again and again, b ecause the information ab out current images p osition is restored each basetime p eriod. A small fraction of the CCD frame is read-out in this mode. The parameter Operations/Normal/MeasBoxSide defines the vertical size V of this ROI (star b ox). The horizontal size H is equal to the vertical size plus the image separation: H = V + S ep. Two bias b oxes (V /2 â V ) are app ended to the left and right sides of the star b ox. 6


The detection of images in star b ox is done in the same way as for the centering mode. The two brightest sp ots are considered as the star images. The x­ and y­coordinates of each image are used for precise determination of the gravity center. One of the two centering methods can b e used: thresholding or windowing. In addition to the coordinates, many characteristics of the images are calculated: total flux, maximal flux, image radius and ellipticity. Also, the contributions of the readout and photon noise to coordinate errors are estimated. The normal mode checks the p osition of the star b ox. If it is located far from the optical center, the execution breaks with the error "Stars are far from center" (code 623). The algorithm assumes that a numb er of frames may b e empty for various reasons, but if this numb er is greater than Operations/Normal/MaxDropped for one basetime cycle, the error "No two star images" (code 622) is generated. The error "Empty basetime set" informs that all images acquired during the basetime were rejected by the verification process. The normal mode (over-)writes the CCD image to the file "b oxframe.fits", once p er basetime.

Raw data mode
To have a p ossibility of some expanded analysis, the mode Raw data was implemented in the program. This mode works in the same way as Normal mode, but the results of the processing of the frame (the p osition of the left and the right image sp ots and fluxes) are stored in the output file. During the accumulation time, the star b ox p osition is not corrected and the calculations after each basetime are suppressed in order to to provide temp oral data sampling as regular as p ossible. Unfortunately, it is not p ossible to get the exact time when each frame was really exp osed; instead, the time of the frame extraction from the kernel memory is recorded in the output file. In the raw data mode, the configuration parameters for the normal mode were used b efore Version 2.10. Now, the configuration file contains the resp ective subsection. The mode is useful for testing purp oses and for some research tasks.

Estimation mode
To have a p ossibility of quick estimation of the current seeing data, the Estimation mode was added in the program. This mode works just the same same way as Normal mode, but finishes after one basetime. The mode uses parameters from Operations/Normal/ section of the configuration file. This mode provides more correct measurements than Centering mode b ecause develops more image frames within the same time. Also, this mode (just like Normal mode) calculates the image fluxes with resp ect to background level rather than to threshold level. Since the Estimation mode uses image data within star b ox window, its p osition on the CCD frame must b e known b efore.

Pictures mode
The Pictures mode provides grabbing of the star b ox images and storing their sequence as one output file "b oxrecord.fits" in /opt/dimm/data/images/ sub directory. The mode uses own subsection in configuration file to set the record length, frame exp osure and frame rate. Output

7


file in FITS format has 2 dimensions only: x the same as the side of star b ox and y is equals of the side of star b ox multiplied by numb ers of the frames.

4

Configuration file

The parameters needed for correct work are given in the file turdimm.cfg. It is organized into sections and sub-sections. The meaning of each section and parameter are obvious from the example of the configuration file provided b elow. Section "General" SubSection "Site" DeviceNumber SiteName Longitude Latitude EndSubSection SubSection "DIMM" ApertureBase ApertureSize EndSubSection EndSection Section "Operations" SubSection "Normal" Exposure FrameRate BaseTime AccumTime MeasBoxSide StarRadius ThresholdFactor CGMethod MaxDropped EndSubSection SubSection "Centering" Exposure FrameRate AccumTime ThresholdFactor MinObjectFlux EndSubSection SubSection "FieldTest" Exposure FrameRate AccumTime ThresholdFactor EndSubSection

MD K 2 50 43 44

0 G 4 1

9 O 0 2

; ; ; ;

d o l l

e b o a

v s n t

i e g i

c r i t

e v t u

serial atory or ude of t de of th

number site na he site: hh mm ss e site: dd mm ss

20 ;aperture separation, cm 9.3 ;aperture diameter, cm

4.0 ; 150 ; 2.0 ; 60 ; 60 ; 7; 3; threshold ; 10 ;

f w b a m i n o m

r o a c e m u r a

ame exposure in ms (>=1) rking framerate, Hz se time in sec. cumulation time in sec. asuring subframe side, pxs age radius for CG calculation, px mber of sigma's window - CG computation method ximal number of bad frames

4.0 ;f 30 ;w 2.0 ;f 4 ;n 300 ;m

r o u u i

a r l m n

me exposure in ms (>=1) king framerate l time of the centering phase, s ber of sigma's total flux for target images, ADU

100.0 0 25 0.1

; ; ; ;

f w f f

r o u r

a r l a

m k l c

e exposure in ms (>=1) ing framerate, 0 - freerun mode time of the centering phase, s tion of maximal signal

8


SubSection "RawData Exposure FrameRate AccumTime EndSubSection SubSection "Picture Exposure FrameRate AccumTime EndSubSection EndSection Section "Camera" SubSection "Geometr FieldAperture Scale FocalCoeff Separation OpticalCenter EndSubSection SubSection "Type" Model "Digital Identification Digitization EndSubSection SubSection "Paramet Format Gain ReadOutNoise BiasControl EndSubSection EndSection

" 4.0 ;frame exposure in ms (>=1) 0 ;working framerate, 0 - freerun mode 10 ;full time of the mode s" 4.0 ;frame exposure in ms (>=1) 0 ;working framerate, 0 - freerun mode 10 ;full time of the record

y" 184 0.634 200 40 320 240 ; ; ; ; ; r i c p o a m o r p d a n e t i g n f i u e e e c s of the field diaphragm, px scale, arcsec/px ction to focal changeing rable images separation al center position on the frame

Camera (DCAM 1.31)" ;camera model 0x00f31000000755c ;camera serial number (UID) 12 ;camera digitization, bit/px - not used ers" 658 4 0 11 2 9 . . 9 2 0 5 5 ; ; ; ; c w R R a o O e m r N l era kin in ati format, px g gain, dB electrons ve number; not used in Prosilica

Some of the configuration parameters are not used and probably will b e removed in the future. Note, that unused in the current program version configuration parameters doesn't lead to any problems. In contrary, absence of needed parameter string may induce early termination of the program with corresp onding error message.

5

Structure of the output file

The output file yymmdd-dimm.stm is an ASCI I file. Each line b egins with a prefix indicating the typ e of its contents. In the examples given b elow long lines are artificially split by the symb ol >>>. Parameter lines b egin with the prefix P and contain the copy of the configuration parameters. They are written at the start of the program. Each time an INIT command is issued, the configuration file is re-read. However, only the parameters that are changed are written again in the output file. The parameter lines look like 9


P 2006-11-19 02:04:26 General/DIMM/ApertureBase = 20

Centering mode output
Each of the program modes is rep orted in the output file with the M-line (prefix M), p ossibly accompanied by additional data strings. The mode Center provides only an M-line. The example of such line: M 2006-11-18 01:42:47 Centering: X=9.7 Y=6.7 dX=42.2 dY=-4.5 FLUX_L=46761 >>> FLUX_R=49528 BS=15.3 RMS=5.5 The fields of an M-line contain: 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. Prefix M Date YYY-MM-DD UT time hh:mm:ss Mode name "Centering:" Offset X (in pixels) of the image pair common center with resp ect to the optical center of the device as defined in the configuration file field Camera/Geometry/OpticalCenter Offset in Y Horizontal separation dX, px Vertical separation dY, px FLUX L ­ the integral flux in the sp ot 1 (left), ADU FLUX R ­ the integral flux in the sp ot, ADU BS -- bias (background) level, ADU RMS -- rms of the background noise, ADU

Normal mode output
The start of this mode is marked by the M-line containing date/time stamp and the name of the mode. The data lines follow then. New data lines are output after each basetime and contain the raw results of the seeing measurement. They b egin by the prefix d and look like d 2006-11-19 02:33:59 294 18068 18568 0.232 0.240 1932 2001 42.89 -2.23 >>> 1.286 0.840 1.501 0.594 0.020 0.019 -17.4 16.3 0.62 0.58 3.33 >>> -0.02 3.34 0.08 15.40 5.47 The data are mostly given for each sp ot separately. Sp ot 1 is the leftmost on each CCD line. The contents of each d-line is: 1. 2. 3. 4. 5. 6. Prefix d Date YYY-MM-DD UT time hh:mm:ss of the measurement Numb er of frames averaged Flux in the sp ot 1, ADU Flux in the sp ot 2, ADU 10


7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28.

RMS flux fluctuation in the sp ot 1 divided RMS flux fluctuation in the sp ot 2 divided Maximum intensity of the sp ot 1, ADU Maximum intensity of the sp ot 2, ADU Average sp ot separation in x, px Average sp ot separation in y, px RMS of the sp ot separation in x, px RMS of the sp ot separation in y, px Covariance in x with lag 1, px2 Covariance in y with lag 1, px2 Estimated rms noise in x, px Estimated rms noise in y, px Average p osition of the two-sp ot center in Average p osition of the two-sp ot center in RMS of the average x-p osition, px RMS of the average y-p osition, px Average FWHM of sp ot 1, px Average ellipticity of sp ot 1 Average FWHM of sp ot 2, px Average ellipticity of sp ot 2 Average background, ADU RMS background fluctuation, ADU

by the flux by the flux

x, px y, px

At the end of each accumulation time, a D-line is written. It contains the same time stamp as the last d-line, the numb er of individual measurements (d-lines), average p ositions in x and y and the rms disp ersions in x in y (all in pixels). These rms values take into account the lowfrequency comp onent of the differential image motion (time scales longer than basetime) and is b etter suited for the calculation of seeing. The D-line looks exactly like d-line: D 2006-11-19 02:33:59 27 18068 18568 0.232 0.240 1932 2001 42.89 -2.23 >>> 1.286 0.840 1.501 0.594 0.020 0.019 -17.4 16.3 0.62 0.58 3.33 >>> -0.02 3.34 0.08 15.40 5.47

Test mode output
The mode FieldTest generates one M-line and several L-lines. The M-line contains the date/time stamp, the name of the mode and, in the following data fields, X0 and Y0 -- the center of the field ap erture in pixels with resp ect to full CCD detector plane; dX and dY ­ the estimates of the horizontal and vertical image separations; R1 and R2 ­ the estimates of the ap erture radius in pixels obtained by two different methods; BS ­ the flux level outside the ap erture; FLUX A ­ the mean flux inside the ap erture. The example of this line is: M 2006-11-18 01:42:47 FieldTest: X0=323 Y0=237 dX=36 dY=-1 >>> R1=179 R2=181 BS=20 FLUX_A=1688 11


The L-lines represent the measurements of the variance of the difference D of two consecutive frames with the same exp osure. After the date/time stamp, the exp osure in ms follows, then 2 the mean flux inside the ap erture in ADU, the variance D (not rms) of the difference in ADU2 , and the flatness of the ap erture illumination. The flatness characterizes the relative systematic 2 2 variation of the brightness in the ap erture, it is calculated as (S - D )1/2 /S , where S -- the 2 is the variance of this sum. sum of two these frames and S Finally, the conversion factor (GAI N in e /AD U ) and readout noise (RON in e ) are calculated and written as a separate M-line for NoiseTest: M 2007-01-12 22:31:21 NoiseTest: GAIN=2.01 RON=11.13

Raw data mode output
As usual, the M-line declares the mode start: M 2007-01-07 21:00:43 RawData Then, the block of the data lines follows. Each data line b egins with the prefix r and contains high-precision time mark, the frame sequential numb er, the x- and y-coordinates for the left and right star images and the total fluxes for b oth images. r 75643.7503 0 80.521 20.485 40.114 20.500 9481 9498 The raw data output is finished by one d-string which summarizes the measurements.

6

Control of the program through so cket

The program is controlled by commands issued through the socket connection (p ort 16200). The first group of symb ols in each command is normally its numb er. However, it is not interpreted and can b e any symb ol(s) except space. The commands are not case-sensitive. The accepted commands are: 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. INIT ­ initialize, read configuration RUN or RUN NORMAL ­ start measuring seeing for one accumtime. RUN CENTER ­ start measuring the sp ot centers, full frame. RUN TEST ­ start measuring the illuminated ap erture, full frame. RUN SCENARIO [="..."] ­ run scenario of measurements. The actual scenario can b e either set directly in this command or defined earlier by the SET SCENARIO command. RUN RAW ­ start measuring seeing for one accumtime with extended output. RUN ESTIMATION ­ start measuring seeing for one basetime only. RUN PICTURES ­ start recording of the frames during one accumtime. GET STATUS can have the following answers: OK STATUS=READY, OK STATUS=BUSY or OK STATUS=PARKED GET IDENT returns IDENT="Turbina-core(D) Vers.nnnn" GER ERROR returns the description of the last error GET OFFSET returns offset of the two-sp ot center relative to the ap erture center [px] 12


13. 14. 15. 16. 17. 18. 19. 20. 21. 22.

GET SEPARATION returns the sp ot separation in x,y [px] GET FLUX returns the total and maximal fluxes in b oth sp ots [ADU] GET DATA returns the full data string of current data (like d-typ e in the output file) GET MODE returns the current mode or last finished mode SET SCENARIO="..." sets the current scenario SET OBJECT="..." ­ sets ob ject ID and writes it to the output file STOP ­ stop measurements after finishing the current mode STOP NOW ­ stop immediately PARK ­ disconnect the camera, stop all actions, wait for INIT QUIT ­ park and exit

The program answers are similar to describ ed early in Turbina 2.04 User Guide: OK WAIT=tt ­ correct command is received, needed execution time is tt sec OK STATUS=READY ­ command execution is successfully finished OK STATUS=BUSY ­ command execution isn't p ossible due to current work OK STATUS=PARKED ­ PARK command is successfully done ERROR STATUS=PARKED ­ the command cannot b e executed b ecause DIMM is parked 6. ERROR STATUS=ERSYN ­ the command has wrong syntax 7. ERROR STATUS=ERFAT ­ an error has occurred during execution The scenario syntax is the same as in the Turbina vers. 2.04. To denote Center mode, Normal mode, and Field test the symb ols c, n, t are used. The scenario is case insensitive, as well as the commands. For example, to run Center mode once and Normal mode three times, the scenario "c+3*n" can b e used. 1. 2. 3. 4. 5.

7

Changes in the version 2.4
1. All the coordinates output data are in pixels only. Now the configuration parameter scale is used for further processing. An external request GET OFFSET returns the star distance from the optical axis in arcsec. 2. The problem with the free-run camera mode (when the camera defines itself the maximal frame rate for a given exp osure) is fixed. To initiate this mode, set the configuration parameter framerate equal to 0. 3. The relative dimensions of the Normal mode subframe are changed to provide more freedom for star images motion in the horizontal direction. 4. The program was renamed to dimm. 5. Two options -i and -p are added for running the program in the command-line mode. The first key must b e used with PC Internet name, the second -- with the program's server p ort numb er. The default values are: "0.0.0.0" for the machine name and 16200 for the p ort numb er. 6. In the versions b efore 2.4, the star image diameter was output, instead of radius. This is fixed. Also, the definition of the image elongation is changed. 13


7. The calculation of the conversion factor (Gain in e /AD U ) and the readout noise (RON in e ) was added in Test mode. 8. The D-line format is extended to the d-line format.

8

Changes in the version 2.5
1. To provide more unified MASS and DIMM data structure the extended file names are introduced. The dimm log file name is yymmdd-dimm.log instead of an imp ersonal yymmdd.log, the output file name is yymmdd-dimm.stm. The name yymmdd.dimm is reserved for processed data file. 2. New command RUN RAW is added. In this mode, the individual data strings are output in the file during accumulation time. This mode is useful for testing purp oses and for some scientific tasks.

9

Changes in the version 2.6
1. Two options -b and -a are added for command line usage. The first key forces program execution as background job. This option is useful when program must b e started without terminal connection ( by cron, for example). The second key is used when automatic initialization is desired.

10

Changes in the version 2.9

1. New command RUN PICTURES is added. 2. New application-sp ecific command GET FOCUS is added for automatic focus control in the case of telescop e RCX400.

11

Changes in the version 2.20

1. The er r in the output of measured radiuses is corrected. Correct radius is greater by ro factor 2. The values are not used in the program algorithm, so no recalculations are needed. 2. The maximal fluxes in left and right images are added in return of the command GET FLUX.

12

Changes in the version 2.25

1. The algorithm of the star images detection in the frame was modified to provide faster and more robust work. The algorithm have fixed image fragmentation which arisen during some seeing condition. 2. Integral and maximal fluxes calculation in the thresholding mode was changed to calculation ab ove background not threshold. It provide Schtrel numb er undep ending from target star brightness. Also, these fluxes may b e used for photometric analysis without any correction. 14