I have a graph with a large image. I want to save/export the graph keeping the axes, labels, and all annotations as vector art, but the image data as a PNG bitmap with a resolution I can specify. The PDF and EMF export formats should be capable of producing mixed bitmap/vector data. However the Igor Pro PDF and EMF export produce only vector output, which makes huge, uneditable files since the images are treated as collections of rectangles instead of bitmaps. If I save everything in PNG format, the axes and labels are not vectorial but the file size is small and manageable. I am using Igor Pro 6.3.
Should this go to the wish list or is there a workaround to produce mixed bitmap/vector output?
As you already tried, you should save the image graph as PDF. Usually, Igor does exactly what you would expect. It will save the axes and so on as vector objects and the image itself as bitmap (with a resolutions equaling 1 data point = 1 pixel as default, independent of the actual graph size, but in a scaled form. This may lead to blurred images, depending on the pdf viewer which may or may not apply smoothing. You can adjust the resolution from the command line). But there are occasions where images are rather saved as individual vector rectangles. One instance I know of is when there are NaNs in the data, i.e., one data point ‘left out’ (funnily, this is not the case for Infs). Then the image is saved as rectangles, leaving out the ‘NaN’ ones.
So in short, search for ‘unusual’ data points in your wave and replace them with something else (maybe infs to make them disappear) before saving. I haven’t found the exact conditions for this in the help and manual, but maybe I’ve looked in the wrong places.
As you already tried, you should save the image graph as PDF. Usually, Igor does exactly what you would expect. It will save the axes and so on as vector objects and the image itself as bitmap (with a resolutions equaling 1 data point = 1 pixel as default, independent of the actual graph size, but in a scaled form. This may lead to blurred images, depending on the pdf viewer which may or may not apply smoothing. You can adjust the resolution from the command line). But there are occasions where images are rather saved as individual vector rectangles. One instance I know of is when there are NaNs in the data, i.e., one data point ‘left out’ (funnily, this is not the case for Infs). Then the image is saved as rectangles, leaving out the ‘NaN’ ones.
So in short, search for ‘unusual’ data points in your wave and replace them with something else (maybe infs to make them disappear) before saving. I haven’t found the exact conditions for this in the help and manual, but maybe I’ve looked in the wrong places.
May 31, 2013 at 08:30 am - Permalink