[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

From |
cnguyen@stata.com |

To |
statalist@hsphsun2.harvard.edu |

Subject |
Re: st: Graph output: wmf to gif? |

Date |
Wed, 02 Apr 2003 15:38:47 -0600 |

"David LB Schwappach" <dlb@h-ero.de> wrote: > my original query has inspired an interesting discussion on the list. I > agree with Lee Sieswerda and others, that eps is best for paper > publication of graphs, such as journal submissions. However, as initial > format for gif conversion for web publication, my experiences are not as > good as reported by Lee. Through my eyes, the result are never as sharp > and precise as wmf to gif converted graphs (this may be an issue Actually, the opposite is true but you are correct about the sharpness. EPS is more precise than WMF and therein lies the problem. EPS uses points as a unit of measurement (which is 1/72 of an inch) and you can have fractional points. WMF and PICT use pixels as a unit of measurement with 72 pixels per inch (in the ideal resolution) and of course pixel sizes are integers. When you convert an EPS file to a bitmap, the EPS converter tries to maintain the size of the line relative to the size of the image. Because a pixel is the smallest screen unit, it tries to fool the eye into thinking a line is thin using antialasing and drawing it "lighter". A solid black one pixel line next to a light gray one pixel line appears "thicker" even though they're the same pixel thickness. The minimum line thickness in WMF is 1 pixel so smaller images appear "sharper". I do agree with David that his examples seem to show that WMF to GIF appears to give better results. However, the EPS image appears to be slightly smaller, the font used seems to be wider (and thinner in stroke) giving less room to the actual plot area. You typically don't get the benefit of antialiasing when using WMF (I pasted a graph into Photoshop and curves and circles remained jagged). I'd also like to point out that the last bar actually shows up in the EPS graph whereas the WMF graph shows nothing. Anyway, I still contend that EPS to gif is the better way to go as the results at http://www.stata.com/stata8/graphics.html show but it requires the image be large enough. Any graph that contains more than just horizontal and vertical lines as well as thick lines should look better. Take a close look at http://www.stata.com/stata8/g/twoarea.gif and http://www.stata.com/stata8/g/g_linee.gif. btw, you should be able to print a PDF file to a non-PostScript printer using Acrobat Reader as a kind of a cheapy RIP (Raster Image Processor). I used to do that with that with a lot of our color PostScript files that I needed to print on a color non-PostScript printer. You can avoid paying the often expensive PostScript option on many printers doing this. -Chinh Nguyen cnguyen@stata.com * * For searches and help try: * http://www.stata.com/support/faqs/res/findit.html * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/

- Prev by Date:
**st: predict or adjust after logistic?** - Next by Date:
**st: March activity on SSC** - Previous by thread:
**Re: st: graphs in pdf format** - Next by thread:
**Re: st: Graph output: wmf to gif?** - Index(es):

© Copyright 1996–2016 StataCorp LP | Terms of use | Privacy | Contact us | What's new | Site index |