https://sourceforge.net/p/p3d/feature-requests/2011-02-18T13:39:15ZRecent changes to feature-requests2011-02-18T13:39:15Z2011-02-18T13:39:15Zskamannhttps://sourceforge.net/u/skamann/https://sourceforge.neted6ceea86f77a4779fd0e901f649a9fa90bc11fc

Hi Christer, it would be very helpful if p3d could give an estimate of the spectral resolution in an exposure. I think in principle it shouldn't be a big deal to calculate this: If p3d fits the arc-lines with Gaussians when running the "Fit centroids"-function, it would be sufficient to divide the wavelength of the line by the best-fit FWHM of the Gaussian and to store/print the mean value for every line. Many thanks, Sebastian

2011-01-21T14:24:06Z2011-01-21T14:24:06ZAnonymoushttps://sourceforge.net/u/userid-None/https://sourceforge.net4e13fc618c1c79d8ce055a8f65489255864d930f

The wavelength calibration of p3d introduces a wavy pattern in the error cube \(see attached plots for an example\). The cube containing the reduced data is not affected. Changing the extraction method, the number of arclines used or the degree of the polynomial fit does not help. Feature occurs with different instruments \(I've tried PMAS and ARGUS\). Possibly the interpolation/rebinning of the error cube during the wavelength calibration is the cause for this\!? Best regards and thanks, Sebastian \(skamann@aip.de\)

2010-09-02T06:54:37Z2010-09-02T06:54:37ZChrister Sandinhttps://sourceforge.net/u/christersandin/https://sourceforge.net534cdc7e58d0fa3d5a82a629024b00de4d1b7d1a

In the case that the selected lines cover only a portion of the entire range the spline routine may fail horribly and ruin all line positions. This has to be fixed somehow; possibly by setting the rational spline parameter \(sigma in SPLINE\) to, say 0.5 instead of the default 1.0 \(this change is not implemented in the code\). Note, for a vast majority of cases this is not a problem.

2010-03-16T20:37:17Z2010-03-16T20:37:17Zp2whttps://sourceforge.net/u/p2w/https://sourceforge.netd9438b9190f09d142c4aa6e4f4db543556ab5fb9

SVN r111 PMAS LARR 4kx4k When displaying the fiber flat that was just created, using the "road-and-hills" button, the window that appears has a few problems compared to other windows of the same type that are displayed for other data: \- It is much smaller than e.g. the one for trace and bias. \- In the first tab, it displays the raw data, but that is apparently squeezed vertically to fit the small window size. \- It also only displays one of the four PMAS CCDs. \- The statistics field is not working. I don't get any error message when opening it, so I just attach a screenshot.

2010-03-16T01:46:26Z2010-03-16T01:46:26Zp2whttps://sourceforge.net/u/p2w/https://sourceforge.netccbd74c2ef6d73360ab9c11780df4d9b7a14d31c

SVN r111 PMAS 4kx4k LARR data When creating a new master bias, the file selection in the file dialog defaults to \*a.fits;\*a.fits.gz;\*a.fits.Z This causes only the \*a.fits\* files to be displayed. p3d does not seem to search for the files of the other quadrants. As a results I have created quite a few useless one-quadrant master bias files today... Solution would be to display all \*.fits files to allow the user to more easily select all required files or to make p3d automatically select all other raw bias files \(as it already does for other raw data types\).

2010-01-28T15:10:57Z2010-01-28T15:10:57ZOle Streicherhttps://sourceforge.net/u/olestreicher/https://sourceforge.net9c4e4c34a2625b258b70aebed49d165db4f597bf

In the moment, the values are taken as ADU without applying gain in output files and in the display. When the gains of CCD quadrants are different, then different values are written to the file, and after applying the dispersion mask there is no way to find out which gain was used for each pixel. This is the case for the current PMAS 4k4 chip. I would propose to change the unit to electron counts instead of ADU at an early processing step.

2010-01-26T16:30:49Z2010-01-26T16:30:49ZOle Streicherhttps://sourceforge.net/u/olestreicher/https://sourceforge.neta38482ebcbc7c0ec6e7dab3df64cfb79274879de

When selecting "Read lines", only the lines within the currently estimated range are read in -- independent whether they would be part of the final fit.

2010-01-26T16:26:24Z2010-01-26T16:26:24ZOle Streicherhttps://sourceforge.net/u/olestreicher/https://sourceforge.netb58f04f2fa1004174f42cbe473f6949373fcc208

Show the residuals of Centroid matching before the final fit is done. This would be nice to estimate the quality of the fit and adjust parameters without doing all steps again. The residual could be shown in the same manner as in the final fit.

2010-01-26T15:30:24Z2010-01-26T15:30:24ZOle Streicherhttps://sourceforge.net/u/olestreicher/https://sourceforge.nete1d862e548d82008cd069d93bf2db50ea99fd328

Show the residuals of Centroid matching before the final fit is done. This would be nice to estimate the quality of the fit and adjust parameters without doing all steps again. The residual could be shown in the same manner as in the final fit.

2010-01-26T13:52:45Z2010-01-26T13:52:45ZOle Streicherhttps://sourceforge.net/u/olestreicher/https://sourceforge.net90d1b2afaa1df5bcf682ebaa9b44a409c1fec2c8

If one gets lost in the manual dispersion calculation setup \(matching lines in the line list and in the spectrum\)m one should be able to cancel this.