; CALIBRAT ;--------------------------------------------------------------- ;! describes the process of data calibration in AIPS ;# INFORMATION CALIBRATION ;----------------------------------------------------------------------- ;; Copyright (C) 1995, 1997, 2002 ;; Associated Universities, Inc. Washington DC, USA. ;; ;; This program is free software; you can redistribute it and/or ;; modify it under the terms of the GNU General Public License as ;; published by the Free Software Foundation; either version 2 of ;; the License, or (at your option) any later version. ;; ;; This program is distributed in the hope that it will be useful, ;; but WITHOUT ANY WARRANTY; without even the implied warranty of ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ;; GNU General Public License for more details. ;; ;; You should have received a copy of the GNU General Public ;; License along with this program; if not, write to the Free ;; Software Foundation, Inc., 675 Massachusetts Ave, Cambridge, ;; MA 02139, USA. ;; ;; Correspondence concerning AIPS should be addressed as follows: ;; Internet email: aipsmail@nrao.edu. ;; Postal address: AIPS Project Office ;; National Radio Astronomy Observatory ;; 520 Edgemont Road ;; Charlottesville, VA 22903-2475 USA ;----------------------------------------------------------------------- ;--------------------------------------------------------------- CALIBRAT LLLLLLLLLLLLUUUUUUUUUUUU CCCCCCCCCCCCCCCCCCCCCCCCCCCCC ---------------------------------------------------------------- For a list of adverbs, verbs, and tasks in AIPS related to the calibration of interferometric data, enter ABOUT CALIBRAT To learn about the process of calibration, read on. The AIPS Calibration and Editing System The AIPS calibration and editing system works mainly with multi-source, time ordered, indexed "raw" uv data files and the associated tables that describe the data and also contain calibration and editing information. When the user is satisfied with the calibration and editing (or is simply exhausted) task SPLIT is used to apply the calibration and editing tables and to write single-source uv files for the imaging and deconvolution routines. The more important tables used are summarized below: Code Name Use ==== =========== ============================================== AN Antenna Keep subarray geometric information, time info polarization, etc. BL Baseline Contains baseline dependent "Correlator" corrections. BP Bandpass Contains the bandpass correction tables. CL Calibration Contains calibration information and the model values which has been applied to the data. CH IF Contains frequency offsets for "IF"s. FG Flag Contains flagging information. NX Index Indexes the file for rapid access. SN Solution Contains CALIB solutions. SU Source Contains source specific information The general utility routines for managing AIPS tables include TACOP, TABED, TAMRG, PRTAB, TAPLT, TAFLG and SETJY. All of these tables can be written to FITS files, and thus may be archived. Multi-source UV data files Many operations needed for calibrating and editing data are simplified if the data for all of the sources involved are contained in the same file. For this reason, AIPS allows "multi-source" uv data files. These files differ from single- source files by the inclusion of a SOURCE random parameter (visible via IMHEAD) and by the presence of Source (SU), Calibration (CL) and Index (NX) tables. Note that a file in "multi-source" format may contain the data for only one source. Most AIPS tasks will process uv data files in either multi- or single-source format. The details of the operation may depend on the format of the input file. Most of the single-source routines will process multi-source data but will not distinguish between sources. Since multi-source files must be in time order, some sort-order-sensitive tasks (e.g. UVMAP, MX) will refuse to process multi-source data. This is usually desirable as they cannot distinguish between sources ! A few multi-source tasks, especially APCAL and CLCOR, must have a multi-source input file. There are several ways to obtain a multi-source uv data file: - Read a multi-source FITS file with UVLOD. - Read a VLA Modcomp tape with FILLR or FILLM. - Read an NRAO/SAO format file(s) with VLBIN followed by (possibly DBCON, UVSRT, UVAVG, TAMRG and) INDXR. - Convert one or more single-source files to multi-source format using MULTI then follow with (possibly DBCON and) INDXR. General Method of Calibration The heart of the general calibration is task CALIB, which will solve for antenna based amplitudes, phases, group delays and phase delay rates. The model used may be one of the following: 1) point model specified, 2) "clean" model or 3) point sources at the phase center with flux densities given by the source (SU) table. Method 3 is the only one allowed if solutions for more than one calibrator are being determined in the same run of CALIB. CALIB may however be run multiple times, once for each calibrator. In that case, each run of CALIB makes a separate solution (SN) table, and the solution tables are merged as they are interpolated into the calibration (CL) table by CLCAL. If old SN tables are to be ignored, they should be deleted using the verb EXTDEST. If CPARM(3 and/or 4) is specified then CALIB will list all baselines in each solution interval which disagree with the model by more than the specified amount. These "closure" error listings are useful in the identification of bad data. Data may be edited and have a previous calibration applied before determining the solutions. Flux densities for the calibrator sources may be entered in the SU table using the task SETJY. The flux densities of secondary flux calibrator sources may be determined from the solution tables using GETJY. If CALIB has been run on both primary and secondary calibrators, GETJY will determine the source flux densities, put them into the source table and correct the amplitudes in the SN tables to what they would have been had the flux densities been in the SU table prior to running CALIB. Solution tables may be concatenated and smoothed using task CLCAL. CLCAL may also interpolate the smoothed, concatenated solutions to the times of a calibration (CL) table and either update the CL table or write a new table. Up to 46655 CL tables are allowed. It is a good idea to create a new CL table each time, as this allows you to return to earlier stages of the calibration if anything goes wrong. CLCOR may be run several times writing to the same CL table using different groups of sources and calibrators. It is strongly recommended that CL table 1 never be modified, as the calibration process can then be restarted by returning to CL table 1. (CLCAL and some other tasks will actually refuse to modify CL table 1, for this reason). Various combinations of sources and calibrators to be used may be specified in CLCAL. It is the responsibility of the user to ensure that an appropriate set of SN tables is applied to the appropriate CL table at each step of calibration. Incorrect application of SN tables will probably result in miscalibrated data. Various corrections (such as phase jumps) or external calibration (e.g. weather related corrections) may entered directly into the CL table by tasks CLCOR, TABED, or APCAL. SN and CL tables may be listed using LISTR with OPTYPE='GAIN'. A given calibration table can optionally be applied in many of the calibration and editing related tasks. Calibrated single-source files may be generated using task SPLIT. SPLIT will also perform several useful operations on single-source files such as applying a solution (SN) table or averaging spectral channels. General Data Editing and Examination The task LISTR will print selected portions of the data in several forms with calibration and editing optionally applied. LISTR can also be used to display information from solution or calibration tables and to give a summary of the observations in a multi-source uv data file. POSSM will average spectra given a set of data selection criteria and produce a plot file. This plot file may then be displayed on any of the plot output devices. UV data files may be gridded into an image using UVIMG for examination on the TV display. Time-baseline displays should be particularly useful for spotting bad data. Task UVFLG can be used to specify data to be ignored. For single-source files, UVFLG modifies the data file but, for multi-source files, entries are made in the flag (FG) table. UVFLG may also be used to unflag data. Multi-source data may be flagged/unflagged for a specified reason which makes it easy to undo selected sets of flags. Most of the calibration and editing routines let you specify which flag table (if any) to apply to multi-source files. A preliminary version of the generalized, interactive TV-based data editor (TVFLG) is now available. General polarization calibration The calibration of polarization sensitive visibility data involves two distinct operations: 1) determination of the effective feed responses and the correction of the data to remove the effects of imperfect feeds and 2) removal of any systematic phase offsets between the two systems of orthogonal polarization. These two components of polarization calibration will be considered serarately. The effective feed response can be parameterized in a number of ways; the most general is in terms of its polarization ellipticity and the orientation of the major axis of the ellipse. Another parameterization which is adequate when all feeds involved are relatively good and of similar design is a linear approximation in which the response to a given polarization is assumed corrupted by a small complex gain times the orthogonal polarization. Both of these methods are to be implemented in the task PCAL; the selection is made by the adverb SOLTYPE. In general, the polarization of the calibrator(s) to be used to determine the feed parameters will not be known a priori and must be determined simultaneously with the feed parameters. Observations of a given source over a wide range of parallactic angles is necessary to separate calibrator polarization from the feed parameters. Task LISTR may be used to determine the parallactic angles at which a given source has been observed. Several calibrators sources may be included. Accurate calibration of data to be used to estimate the feed parameters is essential. In particular, the phase calibration of any calibrator to be used should be determined from the calibrator itself (selfcalibration). The values of the feed parameters determined depend on any systematic phase offsets between the orthogonal polarization systems. If the phase offsets are time variable then they must be corrected before determining the feed parameters; if the phase offset is constant then it may be removed after determining the feed parameters. In this latter case, the derived feed parameters must be corrected for the change in the phase offset. The normal phase calibration technique treats parallel hand visibilities in the two orthogonal polarizations independently. Thus, there will be a systematic phase difference between the two polarizations systems. This difference may be due to differences in instrumental phase delay for the two systems or an effect of the propagation medium (Faraday rotation) or both. Faraday rotation effects are particularly bothersome as they may be rapidly time variable and increase rapidly towards lower frequency. The phase offsets at a given time may be determined from data from a source with a known angle of linear polarization which have had the effects of imperfect feeds removed. The phase of the right-left correlations or the conjugate of the left right correlations indicates the phase difference between the two systems. Polarization calibration in AIPS follows amplitude and phase calibration and proceeds as follows: 1) any time variable component of the phase difference between the right and left hand systems must be removed by suitable modifications of the appropriate CL table or direct correction of the data for single source files. The CL tables may be modified using CLCOR. 2) PCAL is run to determine the polarization of a set of calibrators and the feed parameters. The derived calibrator polarizations are placed in the Source (SU) table and the feed parameters are placed in the AN table. AIPS tasks which apply the standard calibration (e.g. LISTR, SPLIT) can then be instructed to apply polarization corrections using DOPOL=true. 3) A constant phase difference between the right and left hand circular systems may be determined for each IF using a source of known polarization angle which has been phase calibrated on itself with the task 'LISTR'. Use STOKES='POLC'; DOPOL=TRUE; DOCAL=TRUE; GAINUSE=(appropriate CL table) OPTYPE='MATX'; DPARM(1)=1. The expected RL phase is twice the polarization angle. 4) Corrections to the RL phase difference (1 per IF) may be made to the CL table and the feed parameters in the AN tables using task CLCOR. AIPS tasks which apply the standard calibration (e.g. LISTR, SPLIT) can then be instructed to apply polarization corrections using DOPOL=true. General bandpass calibration The calibration of spectral line data is very similar to that of continuum data with the exception that the frequency response of the antenna gains have to be determined and corrected for in addition to the normal gain solutions. This complex bandpass response function is usually determined by observing a strong continuum source at the same frequency as the line observation. The task BPASS is designed to take visibilty data from a specified calibrator(s) and determine the antenna-based complex bandpass functions. It does this in a manner analagous to self-cal in that the data are divided by a source model and then the antenna gains are determined as a function of frequency. These are written to a BandPass (BP) table and can then be applied to the data by a variety of programs. The BP tables are applied to the data by setting the adverb DOBAND > 0 and selecting the relevant BP table with the adverb BPVER. There are five modes of bandpass application. The first (DOBAND = 1) will average all bandpasses for each antenna within the time range requested thus generating a global solution for each antenna; the second mode (DOBAND=2) will use the antenna bandpasses nearest in time to the data point being calibrated; the third, and most cpu intensive, mode (DOBAND=3) is to interpolate in time between the antenna bandpasses and generate the correction from the interpolated data. Modes 4 and 5 are like modes 2 and 3 (resp.) but ignore the bandpass solution weights which are used in modes 2 and 3. Calibration of VLA data (continuum) The AIPS calibration and editing facilities have a basic design similar to the calibration and editing system on the VLA Dec-10. The general correspondences between Dec-10 tasks and tables and their AIPS counterparts are summarized in the following table: Dec-10 AIPS Function ====== ===== ============================================= ANTSOL CALIB Determine antenna based calibration and write SN solution table. " GETJY Determine source fluxes from SN table. DBCON DBCON Concatenate uv data files. FLAGER UVFLG Flag bad uv data. FILLER FILLR Load data from a Modcomp tape before Jan 88. FILLM Load data from a Modcomp tape after Jan 88. GTBCAL CLCAL Apply solutions from (ANTSOL/CALIB) to gain/calibration table. GTBCOR CLCOR Apply corrections to gain/calibration table. LISTER LISTR Print selected portions of a uv data file. PASSUM POSSM Average spectra in uv plane. POLCAL PCAL Determine polarization calibration parameters. SETJY SETJY Enter source info into source table. ====== ===== ============================================== .CAL SN Table containing antenna-based solutions for interpolation into gain (calibration) table .GAI CL Table containing gain corrections to be applied to data. .NDX NX Index table for rapid access to selected portions of data. The following procedure is recommended for calibrating VLA data: 1) Load data from tape. Data observed before January 1988: Read data from a VLA Modcomp tape with FILLR. The CL table increment DPARM(8) should be appropriate for the configuration, frequency and weather at the time of observation. Two (2) minutes should be appropriate for many circumstances. The AC correlator data will be written as AIPS IF 1; the BD data will be AIPS IF 2 and will be kept in the same uv data file. These may be selected by BIF and EIF in many of the calibration/editing tasks. Data observed after January 1988: Use task PRTTP to obtain an index of the tape or tapes to be read. Data may be loaded 1 band at a time using FILLM. If the number of output visibilities specified (NPOINTS) is sufficiently large (see PRTTP listings), then data from multiple tapes can be written into the same AIPS file by specifying OUTNAME, OUTSEQ and leaving NPOINTS the same. Tapes should be loaded in the correct time order. Multiple output files may be written if necessary, e.g. separate files for "channel 0" and line data are written. A subset of the spectral channels may be selected using BCHAN and ECHAN. The CL table increment DPARM(8) should be appropriate for the configuration, frequency and weather at the time of observation. Two to five minutes should be appropriate for many circumstances. The AC correlator data will be written as AIPS IF 1; the BD data will be AIPS IF 2 and will be kept in the same uv data file. These may be selected by BIF and EIF in many of the calibration/editing tasks. 2) Enter the flux density of the primary flux density calibrator (usually 3C286, or 3C48) in the SU table using SETJY. Use ZEROSP = flux,0 if only the I flux density is known. One run of SETJY is required for each source and each IF, despite what you might think from the presence of a SOURCES input, as ZEROSP can handle only one (I,Q,U,V) flux set at a time. Setting the CALCODE allows selection by this parameter later. The flux densities of 3C286 and 3C48 on the scale of Baars et al. (1977, Astron. Astrophys. 61,99) are given in the VLA Calibrator Manual (Sep. 1986) as: 3C286: Log S = 1.480 + 0.292 * Log v -0.124 * (Log v)**2 3C48: Log S = 2.345 + 0.071 * Log v -0.138 * (Log v)**2 where S = flux density in Jy and v is Frequency in MHz. These values at a few selected frequencies are: Frequency (MHz) S 3C286 (Jy) S 3C48 (Jy) =============== ============ ============ 1465 14.51 15.37 1680 13.55 13.76 4885 7.41 5.36 14765 3.48 1.75 15035 3.44 1.71 22485 2.53 1.09 3) Use LISTR with DOCRT=FALSE to print the scan-averaged raw data for all primary and secondary calibrators in matrix format. Use OPTY='MATX', DOCALIB=FALSE, DPARM=3,1,0 (for amplitude, r.m.s., ampscalar averaging) or DPARM=5,1,0 (for amplitude, phase, ampscalar averaging). Separate runs must be made for the two IFs, using BIF=1 and BIF=2. 4) Use UVFLG to specify any data to be flagged. Use FLAGVER=1. TVFLG may be used as an alternative to steps 3 & 4. One warning, most users should set FLAGVER to 1 in the flagging step, this ensures that all the flagging information is written into the same FG table and is not spread between several tables as can happen if multiple editing passes are made. 5) Determine calibrator solutions using CALIB. Remain calm! CALIB isn't really as bad as it looks. Most of the inputs can be left at null (default) values. Use RESTORE 0 to set defaults, or set DOCALIB=FALSE, FLAGVER=1, SMODEL=0, DODELAY=FALSE, SOLINT=0, APARM=0, SOLTYPE='', SOLMODE='', CPARM=0, ANTWT=0 and CLR2NAME. The only parameters you must then specify are the input uv file (use GETNAME), CALSOUR (set to the names of your calibrators), UVRANGE or ANTENNAS (if restricting the solution for your calibrators by uv distance or antenna number) and REFANT (if possible, choose a stable antenna near the center of the array that was present throughout your observations). Separate runs of CALIB must be made if different calibrators need different values of UVRANGE or ANTENNAS (usually the case) or if a source model other than a point model is to be used. Multiple calibrators may be processed in a given run of CALIB if the same UVRANGE and ANTENNAS is appropriate. Each run of CALIB will produce a separate SN table. IMPORTANT: If a run of CALIB is redone, the previous SN table corresponding to that run should be deleted using EXTDEST. LISTR with OPTYPE='GAIN' may be used to determine which calibrators are present in a given SN table. The closure error listing option in CALIB may be useful in the identification of bad data. Use CPARM(3)=5, CPARM(4)=5 to obtain listings of all calibrator data differing from the model by more than 5% in amplitude and 5 degrees of phase. PRTMSG may be used to obtain a hardcopy of this information. Both 3C286 and 3C48 are resolved by the VLA in some configurations and frequencies. Point models for these sources are therefore only accurate over a limited range of baseline length. The range of baseline length used can be controlled by the adverb UVRANGE. If there are too few baselines to a given antenna, accurate solutions may not be possible; therefore, it is frequently necessary to limit the antennas used to the inner antennas on each arm. (The antenna pad numbers which include the order number from the array center on each arm can be determined by running PRTAN). The VLA Calibrator Manual suggests the following sets of UVRANGE (in kilolamda) and inner number of antennas. 3C48: Band UVRANGE Array No. ant. per arm Notes ==== =========== ===== ================ ================= 20cm 0-35 A 7 " B,C,D All 6cm 0-50 A 3 " B,C,D All 2cm 0-60 A 1 Not recommended " B 3 " C,D All 1.3cm 0-68 A 1 Not recommended " B 3 " C,D All 3C286: Band UVRANGE Array No. ant. per arm Notes ==== =========== ===== ================ ================= 20cm 0-18 A 4 " B,C,D All 90-180 A All Reduce flux 6% 6cm 0-25 A 1 Not recommended " B 4 " C,D All 150-300 A All Reduce flux 2% 2cm 0-150 A 3 " B,C,D All 1.3cm 0-180 A 2 " B 7 " C,D All The values of UVRANGE for each secondary calibrator may be determined from the VLA Calibrator manual or by using UVPLT to plot the amplitudes as a function of baseline length. 6) Determine the flux densities of the secondary calibrators using GETJY. GETJY reads the SU and SN tables to determine the flux densities of the secondary calibrators, enters the values in the SU table and corrects the SN tables. The fitted values are listed on the message terminal and written to the message file. To run GETJY set INNAME etc. using GETNAME, SOURCES='list of secondary calibrators', CALSOUR='primary flux density calibrator(s)', QUAL=-1, CALCODE='', TIMERANGE=0, SUBARRAY=0 and ANTENNAS=0. 7) Apply SN table(s) from CALIB to CL table 1 writing CL table 2 using CLCAL. Use SOURCES=(list of program source names), CALSOUR=(list of phase calibrators), OPCODE='CALI', GAINVER=1, GAINUSE=2, REFANT=reference antenna desired. If there are several groups of sources and calibrators, each group should be processed in a separate run of CLCAL, all writing to CL table 2. If all calibrators are to be used and the calibrator codes are given in the SU table set CALSOUR='', CALCODE='*'. 8) Examine calibrated data using LISTR. Use as in 3) above but use DOCALIB=TRUE, GAINUSE=2. Alternately, UVIMG can be used to grid the uv data into a form that it can be displayed on a television device. Use UVFLG to flag data if necessary. Also, TVFLG may be used to interactively edit the data set. 9) Use SPLIT to calibrate and edit the data into single-source uv data files. Use SOURCE='' (or specify sources individually), STOKES='',DOCAL=TRUE, GAINUSE=2, APARM=0. AC and BD data may be separated, if desired, using BIF and EIF. Polarization Calibration of VLA data Polarization calibration may be performed on amplitude and phase calibrated VLA data using the following procedure: 1) Run PCAL on one or more phase calibrator sources observed with a wide range of parallactic angles. Set INNAME etc. to the input multisource file, CALSOUR=list of calibrators to use, TIMERANG=0, ANTENNAS=0, UVRANGE=(appropriate range), DOCALIB=TRUE, GAINUSE=CL table to use, CLR2N (to clear IN2NAME etc.), PMODEL=0, SOLINT=2, SOLTYPE='APPR', PRTLEV=0, REFANT=reference antenna, CPARM=0. 2) Use LISTR to determine to apparent RL phase angle of the polarization angle calibrator source (e.g. 3C286). Use OPTYPE='MATX', DOPOL=TRUE, STOKES='POLC', SOURCE='(polarization angle calibrator)', DOCAL=TRUE, GAINUSE=(appropriate CL table, the source should be phase calibrated on itself only), UVRANGE=(appropriate range for source/frequency/array), ANTENNAS=(appropriate list for array/frequency), FLAGVER=1, DPARM=1,0. LISTR needs to be run once per IF specifying BIF. The matrix averaged RL phase angle will be printed after the matrix of phases. Check that none of the phases differ from the mean by more than a few degrees. If any do then use UVFLG to edit this data and go back to step 1. Special care will be needed if the average is near +/- 180 degrees as the average value given may not be reliable. 3) Use TACOP to copy the CL table with the results of the amplitude and phase calibration. (Note: this step is not essential but reduces the magnitude of the disaster if the the next step is done incorrectly.) Set INNAME, OUTNAME etc. to point to the relevant file, INEXT='CL'; INVER=(result of amplitude and phase calibration), OUTVER=(current highest + 1). 4) The RL phase offset corrections can be made using task CLCOR. The phase offset correction is the expected value (66 deg for 3C286) minus the observed phase from step 2. Set INNAME etc. to point to the file whose CL and AN tables are to be corrected, SOURCE='', STOKES='L', BIF=1, EIF=2, TIMERANG=0, ANTENNAS=0, SUBARRAY=1, GAINVER=(CL table just created by TACOP), OPCODE='POLR', BPARM=(RL phase corrections, 1 per IF in order). Note: if more than one CL table needs to be corrected, use the OPCODE='POLR' option only once; other CL tables should be corrected using OPCODE='PHAS' and correcting 1 IF at a time. Correction applied by CLCOR are cumulative; multiple runs with OPCODE='POLR' may result in invalid feed parameters in the AN table or incorrect RL phase offsets in the CL table. 5) Use DOPOL=TRUE in SPLIT or LISTR to apply polarization corrections before copying the data to a single source file or listing the data. Calibration of VLA line data The calibration of VLA line data follows closely the recipe laid out in the section on VLA continuum calibration. However there are are number of steps which are different so below is listed the recipe recommended for the calibration of line data. 1) Reading the data. If your data are on a VLA Modcomp tape then they should be read into AIPS using FILLM, as descibed in C(1) (C(1) means the continuum calibration description above, section (1)). FILLM will fill a typical line observation into two files, a large one containing the line data only, and a smaller file containing the "Channel 0" data. Most of the calibration and editing on performed on channel 0 and the results copied over to the line database. 2) What if your channel 0 data are meaningless? For instance if you have observed a maser source which has no associated continuum emission. In this case you have to follow two steps. You should still use the channel 0 data for your calibrator sources to generate the appropriate calibration tables, but in the editing step you will have to edit the channel 0 continuum data, and then in a second pass edit the line data set with examination of the relevant channels. 3) Editing. You should follow the same steps as outlined in the continuum section, i.e. steps C(3) - C(4), with the caveat mentioned above. However if all your editing has been done on the channel 0 case - which is the most common method, the FG table generated should be copied to the line file. Use TACOP with INNAME etc set to the channel 0 data, OUTNAME to the line data, INEXT = 'FG', NCOUNT = 1, INVER and OUTVER should be set to whatever is appropriate. The KEYVAL, KEYWORD and KEYSTRNG adverbs should be set to their null values. There is one problem in this step which will be fixed in the future, it is that since the flagging was done on the channel 0 data the channel numbers in the FG table are inappropriate for the line data. This can be fixed using TABED. Set INNAME etc to the line data, INEXT = 'FG', INVER = 1, OUTVER = 1, CLRONAME, BCOUNT = 0, ECOUNT = 0 to edit the whole table, OPTYPE = 'REPL', APARM = 6, 2, 2 to edit column 6, subscript 2 (i.e. the last channel number), KEYVAL = 127,0 for a 127 channel database. 4) Calibration. The basic calibration steps described in C(5) - C(8) should then be performed on the channel 0 data. When you are satisfied with your results you should copy the relevant CL table over to the line database. Use TACOP with INNAME etc set to the channel 0 data, OUTNAME to the line data, INEXT = 'CL', NCOUNT = 1, INVER and OUTVER should be set to whatever is appropriate. The KEYVAL, KEYWORD and KEYSTRNG adverbs should be set to their null values. 5) Bandpass correction. This should be performed on calibrated data; you should ensure that when CLCAL was run in the calibration stage that your bandpass calibrator was include in the list of SOURCES. Determine antenna based bandpass response functions using BPASS. SOURCES=(list of bandpass calibrators), BCHAN and ECHAN should be left as zero to select the full range, or can be set if for some reason you only wish to determine the bandpasses over a small channel range. DOCALIB = 1, GAINUSE = whichever appropriate, FLAGVER = 1, SOLINT = 0, REFANT=reference antenna desired, BPVER = 1. All other adverbs should be at their default values. BPASS can be used to do more than this simple-minded bandpass determination, see the HELP/EXPLAIN file for details. 6) Use SPLIT to apply the gain and bandpass calibration and edit the data into the single-source files necessary for the imaging tasks. Use SOURCE=(line sources), DOCALIB=1, GAINUSE=whatever, DOBAND=1, BPVER=1, FLAGVER=1, APARM=0. If you wish to average subsets of the spectrum into a pseudo- continuum file you can select the channels using BPARM and setting APARM=1,0. Calibration of VLBI data (continuum) The details of calibration of VLBI data depend strongly on the type of data, sources and calibrators (if any) used and personal preference. The following will suggest a number of steps which may be needed: 1) Reading the data into an AIPS multi-source file. If the data are in the NRAO/SAO format, AIPS task VLBIN will read them directly into a multi-source file. FITS multi-source files may be read directly by UVLOD. For other sources of data, task MULTI followed by INDXR may be used to convert single-source to multi-source files. 2) Consolidating the data. If the data for a given experiment are contained in several files, they may be concatenated using DBCON. If the data contain multiple correlations of the same data then UVSRT and UVAVG can be used to get rid of the redundant data. The combined CL table can be compressed using TAMRG. For TAMRG use INEXT='CL',INVER=1, OUTVER=1, APARM=4,1,4,0,1,1,1,0 BPARM=1,4 CPARM=1.157E-5,0.2 DPARM=0. (Re)index the source file using INDXR. 3) Amplitude calibration. ANTAB reads Cal Tech-like text files containing antenna gain and system temperature information and places the information in a TY and GC table. These are then used by APCAL to create an SN table, which is then applied using CLCAL. 4) Fringe fitting. Use CALIB to determine delay and rate residuals. If calibrators were observed, then solutions may be determined for them which can then be applied to the program source(s) before fringe fitting the program source. In either case, use DOCAL=TRUE and GAINUSE=(the appropriate CL table). If a CLEAN model is to be used, fill in IN2NAME etc.; otherwise provide values for SMODEL, or enter source flux densities in the SU table using SETJY. Other parameters are CALSOUR=(source name or names to fringe fit), FLAGVER=1, DODELAY=TRUE, REFANT=desired reference antenna, SOLINT=(suitable solution time, 0=> scan), APARM=0, The values of DPARM depend strongly on the source being processed, and on the expected delay and rate errors. For program sources with calibrator source solutions applied, the search windows can be small (remember 0=>Nyquist range). The values in ANTWT depend on the prior calibration. Use of APCAL before CALIB may result in data getting appropriate weights. 5) Apply solutions to the calibration table. The results of CALIB may be examined with LISTR with OPTYPE='GAIN'. If the results are satisfactory then they may be applied to the CL table applied in CALIB. CLCAL will optionally smooth the SN table and apply it to the specified CL table. It is best to write a new output CL table. 6) Examining/editing the data. Use LISTR with OPTYPE='MATX' or OPTYPE='LIST' to examine the data. Use DOCALIB=TRUE, GAINUSE=0. Bad data may be flagged using UVFLG. TVFLG may also be useful. 7) Calibrating and editing data into a single-source file. SPLIT will apply the calibration and write a single-source file. Note: CALIB can be used to self calibrate single-source files. ----------------------------------------------------------------