IDL PS errors

---

email: gsg@debye.mit.edu
I have noticed that the imaging routines in IDL when using "set_plot,'PS'" do not display certain rows and columns when using xv. i
the following gif images are all 2000x500 images that should show continuous lines of 50 raster segment that go from right to left every 50 rasters. A 'r' in the name implies that the array was rotated by 90 degrees before being plotted. In particular the 'gif' images were mad using 'xv'. the pdf images were made from the Ida postscript using ps2pdf. While there are still some errors in the pdf files, they are nowhere as bad as the 'gif' images. In the particular cases I show, the gaps are obvious, some times only occasional, some times being more than the visual lines.
500_2000_.gif 500_2000_r.gif Compare with "500_2000_.pdf" "500_2000_r.pdf"
In each set, the two images are just rotations of each other. In each case the lines were generated going from left to right in the left image. The right image I just rotated the array before plotting it.

Compare with "2000_500_.pdf" "2000_500_r.pdf"
I think that this might why the spectrograms I have produced using IDL have not shown detail I expected. I am questioning how to proceed to make and print spectrograms without GIPS. I will at least come into MIT and try different devices in IDL. Clearly 'PS' is not reliable, but not as bad as I had feared. What devices have other people tried? For your info, the IDL procedure I used to generate these images is "test_im.pro"
To make things even weirder, I increased the x axis by 1 to 501 (This allows an extra segment on each scan) and see what I got.
501_2000_.gif 501_2000_r.gif
Compare with "501_2000_.pdf" "501_2000_r.pdf"

Conclusions

Obviously the 'pdf' files produces by 'ps2pdf' are much better than the 'gif' files produces by 'xv', they are not error free. The variations in intensity of certain line segment is quite common. There are still a few missed or moved line segments. I have not tried printing these plots which I will do soon.