Creating figure schematics
I have my first two papers in the pipeline, and one common element is a schematic as a figure (one is detailing an experimental design, and another is a flow diagram of exclusion criteria). Because these figures have image resolution requirements from the journal (usually 300 dpi or more), some sort of image manipulation program is necessary (e.g., GIMP). However, I find working in such programs to be difficult for me, as I do not have a design background. Even selecting a portion of a layer can be difficult. So, here is the workflow I have developed so far:
- Design the schematic in Powerpoint, using whatever font is required by the journal. You can also set the slide size to fit within the figure size requirements.
- Save the file as Powerpoint file (so you can edit later), and also as an image (*.png).
- Open the image file in GIMP.
- Go to the “Image” tab and select “Scale Image…” from the drop down menu.
- Click each “chain” to the right of the measurement windows, as this will prevent GIMP from changing the size of the image when you change the resolution.
- Set the X and Y resolutions to be anything you want (e.g., 300 ppi).
- Save the image.
- Insert the image into your Word document, or import it into a LaTeX file.
This workflow has seemed OK so far, but there are some problems:
- I have some reservations about whether ppi and dpi are the same thing, and I have had trouble finding literature on the internet that clearly explains the difference.
- I am not positive whether setting 300 ppi on both the X and Y axes in GIMP produces a 300 ppi image, but the final image looks really clear and is probably acceptable to journals.
- Powerpoint will sometimes make perpendicular lines that don’t meet perfectly at corners, or lines extending from circles that don’t seem perfectly straight. GIMP is more reliable at this, but also more difficult to use.
Does anyone have any suggestions? What workflow has worked for you? As a biostatistician, I will likely be making schematics like the ones described here in almost every paper, so I need to learn how to make them well.