From gtaylor@cv3.cv.nrao.edu Fri Aug 11 09:48:59 2000 Date: Fri, 11 Aug 2000 08:52:01 -0600 (MDT) From: gtaylor To: Greg Taylor Cc: Tim Cornwell , dshepher@zia.aoc.NRAO.EDU, julvesta@zia.aoc.NRAO.EDU, smyers@zia.aoc.NRAO.EDU, gmoellen@zia.aoc.NRAO.EDU Subject: aips++ testing log for August 10th Testing AIPS++ Detailed Log Aug 10, 2000, session: note: TTT = trouble BBB = bug QQQ = question Resumed testing after a bug in stable was tracked down by George and subsequently fixed so that I could access my measurement set there. Checking up on d-term calibration done last time. Plotted up visibility data RL imaginary vs real on 3C286. Results don't look good. Made the same plot in AIPS which looks reasonable but very, very different. Looks like aips++ has tried to zero out Q and U on the polarization calibrator (3C286 was used in this experiment). Tried to start up an imager tool but it crashed aips++. No idea why. Tried again and this time it came up ok. Now try looking at I, Q, U and V on an unpolarized source. Had a little trouble with the viewer window functionality in that the animation control window appears on all screens and can't be "hidden". Submitted a bug report on this. Also discovered that using the "go away X" in a viewer window crashes aips++ -- use "Done". Anyway, Stokes I and V look ok but Q and U have strong (10%) instrumental signal suggesting that the d-term calibration is in error. Went and talked to Athol about it. He said that if I sent him the script I used he would send me a modified script that would do the d-term calibration the way I want. Now waiting on said script. --- end of testing. Frustration meter reads: |****** | low high