An extraneous line appears in the printed version of a Mathematica-image in a pdf-document

Question

My system/workflow

• Mathematica version 11
• MikTeX 2.9: tex->dvi->ps->pdf->paper

I have run into the following difficulty. I cannot tell whether the problem is caused by Mathematica, MikTeX, Adobe, or even the local printer. Asking here first, because the chain of production begins with Mathematica.

The problem:

1. I produced a piece of 2D-graphics needed in my lecture notes. A copy of the Folium of Descartes together with a little extras to demonstrate the implicit function theorem. One of the produced EPS-files is here http://users.utu.fi/lahtonen/FoilX.eps
2. I then included into a LaTeX-document. The appearance of that image is exactly what I hoped in the TeXWorks preview as well as when viewing the final PDF-document.
3. I use four variants of the same image (with varying extra components) in the lecture notes. When printed, each and every one of them has two extra lines: a ray from the origin to Northwest, and an shorter vertical line at the left border. A sample scanned version is here http://users.utu.fi/lahtonen/3316_001.pdf

Internet searches give the following hits

• GeoNet. Does not feel relevant.
• Something possibly related in an Adobe forum.
• A more promising hit at an AutoCAD site (currently not responding to me)
• Something similar on SE

Can anyone suggest something, diagnose the problem, or anything else? One of the hits mentions that the problem may be caused by an image containing transparencies. My EPS looked just fine with GhostView (or something named like that). Are there more sophisticated tools freely available?

Answer 1

The problem seems to have been solved!

I think that it was due to clipping. You see, when parametrizing the Foil $$ x=\frac{3t}{1+t^3},\qquad y=\frac{3t^2}{1+t^3} $$ I had allowed the range of the parameter $t$ to range over an interval that extended to the left from the region in the final image. While Mathematica and MikTeX (and the software I used to view my document) handled the unseen parts just as I had hoped, some residues about the clipped part of the curve had remained in the file, causing the printer to add those two extra lines.

After reproducing variants of the images, where $t$ was constrained in a way that $x$ never wandered below $-2$, the next print looks just fine.