Some Exercises with Figures

Read my brief introduction to image files of issues about figures and review graphics101 found at CNET Builder.com. Then do these exercises.

Here is a 72 dpi image (shift click on the link to initiate a local save rather than viewing it in Netscape) ok for on screen use (desktop viewing or projection as in a seminar). Your task is to use Adobe Photoshop 5.0 to edit this file for use as a black and white figure in a thesis that will be printed at 600 dpi.

  1. open the document in photoshop
  2. what is the current size and resolution of the image?
  3. what mode is it in?
  4. what is the final desired size of the picture
  5. what is the final desired resolution of the picture
  6. change the size and resolution to what you desire
  7. is it usable at this point?
  8. what has to be done to make it usable?
  9. unfortunately you have to work on this blasted pc, but your thesis will be printed from a mac
    what format should you save the file in to have it flawlessly transported to a mac?
    what method should you use in ftping the file to the mac?

Here is a pdf file with a figure (2 B, a stereo pair; this one is wall-eye, the standard for journals; check here to learn about the difference between wall-eye and cross-eye; note that if you use the wrong method, right handed helixes are left handed, and things appear inside out) that you want to reproduce, modify and reference for inclusion in the introduction of your thesis something like this, but suitable for printing).

  1. minimize programs so you can see the desktop
  2. note the screen settings (write them down)
  3. change the screen resolution to the maximum possible
  4. zoom in on the figure
  5. place the current screen on the clipboard (print screen function key; alt-print screen function key only saves the active window)
  6. open a new file in photoshop (it automatically detects the object on the clipboard and by default sets the size of the file to that of the object)
  7. paste the image into the new file (ctrl-v)
  8. crop the image
  9. increase the canvas size to that needed for adding text to the image; note that there are different places to put the new space in the image when increasing canvas size)
  10. is the resolution right? if not, fix it
  11. add desired text (remember, to avoid plagerism you must cite the source of this image!)
  12. in real life you would save this image right now to a file; I prefer .tif format since it does not discard any information (it is a 'true image format')
    but, for this lesson lets pretend we only want to put it into a powerpoint slide show. what resolution should we change it to before saving?
  13. change the screen resolution (on the desktop) back to what it was before you started this exercise

Here is a .gel file from the phosphorimager (this one you have to shift-click on, to force the http server to use binary mode to simply send you the file as is. The server has not been told about this MIME type, so it does not know what type of file this is. It is, in fact, a .TIFF format file, simply named something else by Molecular Dynamics, the maker of the equipment that generated the file. If you don't shift click on it, then Netscape tries to read it as ascii text, which it isn't). Check the figure requirements for this journal, and make a figure that looks like this suitable for submission to the journal.

Making Slides of your Images

Note: this discussion is centered on using Adobe Photoshop and the LaserGraphics unit in 6 Althouse building here at Penn State, because these services are to members of the BMB Department. Some of the discussion  may be relevant to  you if you use other software and slide making facilities.

The  LaserGraphics LFR Plus slide maker can expose up to 2048x1365 or 4096x2731 pixels per slide. A 35 mm slide area is 35 mm wide by 24 mm tall. While it might make sense to open an image in Photoshop that is 3.5 mm wide and 2.4 mm tall at 4680 dots per inch resolution, I am told that such an image will print a small dot at the top left of the slide. This suggests to me that there is a scaling process that is part of the printing process. So what I do is open an image that is 4096 pixels long, and 2731 pixels tall, in greyscale or RGB color mode (~30 MB), and create my image. Then, FTP the file saved in Photoshop's native format, psd (~4 MB), to the slidemaker machine. This will print "as is" using the software provided by LaserGraphics. If I want to print the image to a color laser printer, I simply resize the image to an acceptable resolution and print away.

I am told that an alternative is to consider your page space to be 12 inches wide and 8 inches tall. You are encouraged to leave a 0.5 inch margin on all sides. So, you can open a new image in Photoshop that is grayscale, 11.5 inches wide, 7.5 inches tall, at 300 dpi. Use this space to create your image. I am told that 150 dpi works pretty well too, which will shrink the size of your file from >7 MB to 1.85 MB grayscale, or 5.55 MB RGB color.

A third alternative is to use PowerPoint, which can print via a driver to the LaserGraphics format bll, which is then dumped to the slide maker. This is an easy way to make slides, and the resolution does not appear to have to be higher than 150 dpi before you place the images into powerpoint. Note that bitmap images transfer well from mac to pc, but only postscript ones do for vector based images (pict format, the workhorse for macs, does not exist on pc's; I believe 'pic' is a PC bitmap image format, not a PC version of the Mac pict format).

Tips

The slide maker is connected to a Mac, and a PC. Check out information posted by Dr. Song Tan regarding general and Mac related issues, and me regarding PowerPoint. Until recently, only the PC was running PowerPoint; that may be changing as I write.

PostScript

Here is a description of simple postscript commands that illustrate how one can dramatically and quickly modify vector based images. They are in the context of a program called SEQSCAN that I posted for scanning DNA sequence using weighted scoring matrices. It uses cgi scripts (PERLE), about which we'll learn next.