I do not concur that this image SHOULD be recreated as an SVG. The "artifacts" in question may well be "facts". It isn't just a pretty picture, it just looks remarkably like one (that's the beauty of it). :)
The "artifacts" may be tiny aberrations in the collected data vs. the theoretically expected smooth curve. I would recommend, if anything, submitting it as a PNG, as that could be made to contain all the information in the original, although it could not be (later) enlarged without increasing the file size and creating aliasing artifacts. A PNG would most likely be a smaller file for the same image quality, as compared with the JPEG in this instance.
Technically, an SVG *could* be created that has the error bars and data points preserved, but one would have to be *very* careful. In fact, one would probably have to be in possession of the original data! 24.152.177.61 (talk) 15:55, 11 September 2010 (UTC)RicoS
- No, they are artefacts. Luckily, the original data is available at NASA's website (http://lambda.gsfc.nasa.gov/product/cobe/firas_monopole_get.cfm), so I quickly created an svg version with GNUPlot. Quantum Doughnut (talk) 10:05, 4 February 2011 (UTC)
