Re: [ferret_users] Odd lines on postscript with fill command

[Stephen Guimond <sguimond@xxxxxxx>]

Hi Ryo,

Thank you for your messages.  I actually discovered the +antialias option, too but I wasn't accounting for higher resolution with -density.  Now when I do...

convert -density 1200 +antialias file.ps file.png

the result is very close to the actual postscript file WITHOUT the white lines.  I would agree with you that this is the best option (as of right now).  Of course, one could then convert the .png to .pdf.  Only problem with this is it requires imagemagick, which not everyone has.  I would suggest Ferret developers make this white line postscript problem a top priority so everyone can get the right plot for publication.

Regarding the fusing of polygons, would interpolating the data to a very high resolution grid in space and time alleviate some of the aliasing issues?  Either that or smooth the variable beforehand?

Thanks again,
Steve

----- Original Message -----
From: Ryo Furue <furue@xxxxxxxxxx>
Date: Friday, September 18, 2009 1:48 pm
Subject: Re: [ferret_users] Odd lines on postscript with fill command
To: guimond@xxxxxxxxxxxxx, sguimond@xxxxxxx
Cc: romantonk@xxxxxxx, ferret_users@xxxxxxxx

> Hi Steve,
> 
> | Next, I tried Roman's suggestion and indeed unchecking the aliasing
> | box in ggv does remove the lines.  However, I am sending these files
> | to a journal office and they will probably still print out with 
> those| lines since the file itself was not changed, just the viewer.
> 
> I've sent journal offices PostScript files produced with Ferret
> and indeed have once gotten white lines in the very final PDF!
> I thought they knew better. . . .
> 
> | Note that I tried converting from plt to ps with Fprint and the same
> | line feature is there.
> | 
> | Any other suggestions?  
> 
> So, I suggest you
> 
> 1) tell them the existence of potential problem; or
> 2) convert your graphics into high resolution pixmaps
>  (See my other message I've just sent).
> 
> I'd be reluctant to use option (2) because I don't know
> how high is high enough.  But, then, you could ask them
> what resolution they want.
> 
> Ryo
> 

=======================================================
Stephen R. Guimond
Graduate Research Assistant
Florida State University
Center for Ocean-Atmospheric Prediction Studies (COAPS)
=======================================================