Home | Problem Reporting | Contact Us | College of Forestry Home | Sitemap

> Helpdesk > Training > Graphics Primer

spacer

A Graphics Primer


Another informative Web site:

Resolution

  1. The dots of "dots per inch" can be thought of as the individual elements of the picture. These dots, or picture elements are referred to as pixels.

  2. Higher resolution means more dots per inch, means the dots are smaller.
     
  3. When you scan an image, the scanner samples your image at the resolution you request. A 1-inch square image scanned at 5 dpi would be converted to 25 pixels of information.
    You can see in the example below how lower resolution sampling blurs the detail of an image.


  4. Color prints are only printed at about 200 dpi maximum.
    Negatives and slides are printed at resolutions of about 2000 dpi.

  5. Resolution arithmetic -- The resolution you choose at scanning time determines the number of original/true image pixels you have to work with. Say you're starting with a 5"x4" original.
    Scanned at 75 dpi gives a digital image size of (5 x 75) x (4 x 75) = 375x300 pixels
    Scanned at 150 dpi gives a digital image size of (5 x 150) x (4 x 150) = 750x600 pixels

  6. Some resolution arithmetic examples --
    1. You scan your original 4x4 in. image at 300 dpi. ==> Now you have 1200x1200 pixels.
      If you scale the image to print 8 by 8 inches, you spread the 1200 pixels across 8 inches in each direction. So your printing resolution would be 1200/8 or 150 dpi.
    2. If you scan your 5x4 inch image and print it half size at 2.5x2 inches, and if you want to achieve 150 dpi on printed output, then you scan at half that, or 75 dpi. Why? 5x4 inches scanned at 75 dpi gives an image size of (5 x 75) x (4 x 75) = 375x300 pixels. Scaling to print this at 150 dpi gives 2.5 x 2 inches, because 375/150 = 2.5 inches and 300/150 = 2 inches.

  7. Maybe it's easier to think about working backwards. Start by thinking about how you're going to use the image. How many pixels will you need? Take example (b) above.
    • I want to print a 2.5x2 inch image at 150 dpi.
    • So, I need (2.5x150) x (2x150) pixels -- a 375x300 pixel image.
    • In my original 5x4 inch image I need 375 pixels along 5 inches, 375 divided by 5, or 300 pixels along 4 inches, or 300 divided by 4, or 75 dpi.

  8. Unfortunately, when viewing images onscreen, technical limitations force you not to think in terms of dpi, inches, and image size.
    • Our monitors show a fixed area of pixels, which is usually 640x480 or 800x600 or 1024x768.
    • Pixel size onscreen is determined by the current display settings of the workstation.
    • The pixels of your 800x600 desktop will be bigger than the pixels of your 1024x768 desktop.
    • We think of the pixels of a 150 dpi image as being smaller than the pixels of a 75 dpi image but the pixels of your display desktop are not going to change size to correspond to the dpi of the image you're displaying.
    • And, the number of dots in the image you're viewing will remain constant.
    • So, an image (or text) displayed at 640x480 desktop resolution (large pixels) will cover more screen real estate than the same image displayed at 1024x768 desktop resolution (smaller pixels).
    • Likewise, scanning an image in at a higher resolution will result in more dots in the image. More dots to be displayed onscreen will result in an image that displays larger onscreen because there are more pixels to display.

    The Bottom Line A 3x5 inch image scanned in at 75 dpi resolution contains fewer pixels than the same image scanned in at 150 dpi resolution.
    When displayed onscreen the higher resolution image contains more dots so it will occupy more onscreen real estate.

    For some examples and further explanation, see: http://www.scantips.com/interpol.html

  9. If an image is enlarged/resized after scanning, the graphics program doing the enlarging has to make up information, manufacture pixels based on the original pixels. You could think of it as each individual, original pixel, getting bigger, therefore, resolution is decreased.

  10. Resampling vs. Scaling -- Resampling is a very drastic change. Every single pixel is torn down and rebuilt. Actually, it's replaced with an approximation of others nearby. Scaling is not a change affecting the image pixels at all, rather, it affects the spacing of the original pixels on the printed paper. The original pixels are not otherwise affected, resolution is affected.
    You scale when you print.
    You resample when you resize the image (change the number of pixels) in a graphics program.

  11. Recapping points #9 and #10 -- Within a graphics manipulation program, like IrfanView, if you resize an image, the program will really be changing the number of pixels in the image. If you make the image smaller, the program "throws away" the original pixels and makes up a new pixel to replace many pixels. If you make the image larger, the program looks at the existing pixels and makes up new pixels that are added to fit in with the original pixels.
    So, when you resize an image, the program doesn't change the image dpi. It changes the number of pixels you have to work with. You determine dpi when you use the image/pixels at the dimensions you need, determining how much real estate you spread your pixels over.

  12. Connecting scanning resolution to printer resolution - for full details see: http://fla.esf.edu/research/CAVLab/reference/printers.htm
    The bottom line -- Printer "dpi" is a different thing from image dpi (except in Line art mode). For inkjet or laser printers, here are good scaling guidelines for Color or Gray Scale photographs:
    300 or 360 dpi printer - use 60 to 120 dpi image.
    600 or 720 dpi printer - use 120 to 240 dpi image.

    This statement is made for Grayscale and Color modes. Note the exception for line art mode, which is 1-bit 2-color (B or W, like ClipArt or fax), which contains no gray, no halftones, no dithering. In line art mode, the printer CAN use its full resolution, making use of the full resolution of the scanner.

  13. So, what resolution should I scan at? There is no single answer that will cover all situations. Your answer will come from considering the points below.
    • Total pixel count is what's important. The total pixels you have to work with will determine image resolution depending on the image size you use. 
    • Because the final resolution of your image is tied to the output device and image size that you are going to use, when you're scanning you need to ask yourself, "What am I going to use this image for and what could I possibly use this image for?"
    • Remember that you can never gain back pixels of an image by increasing the image resolution in image processing software. Once the pixels are gone, they are gone. When you increase the image resolution in a software package, the computer only looks at the existing pixels and makes its best guess for the pixels it adds to the image.
    • Based on the points above, one recommendation would be to scan your image at multiple resolutions (one as high as reasonable, one for print use, and one for onscreen use) and archive these images as TIF. You can then work from these originals, making your edits, and saving in other file formats as needed.

File Types -- File | Save As...

There are many graphic file types in use. One brief overview is at: http://fla.esf.edu/research/CAVLab/Reference/files.htm

Three of the most commonly used formats are discussed in more detail below.

  1. TIF -
    • No loss of information (which is important for edits)
    • Huge file size
    • Use this format when doing a lot of editing and re-saving of scanned images using PhotoShop, IrfanView, or some other image editing program. When done editing, save as a .JPG file but you may want to keep the .TIF around as an original.
  2. JPG -
    • Compressed format is lossy
          Original -- 55565566665555
          Compressed -- 5(14)
    • With some information lost in compression, you may want to save your original in another format or at least do all your edits from the original rather than a saved JPG.
    • Most efficient, resulting in the smallest file sizes
    • Stores 24-bit color (16,777,216 color) images
      [doesn't support reduced color depth]
    • Designed for compressing either full-color or gray-scale images of natural, real-world scenes. Works well on photographs, naturalistic artwork, and similar material.
    • JPG supports over 16 million colors, but slightly "distorts" the image to compress the file size. For photos, the human eye cannot tell the subtle changes in color, but along straight edges and in pictures with large solid colors, distortion becomes very apparent.  (Note in the sample to the right how JPG leaves "ghosting" and distortion around the letters.)
    • Does not allow transparency.
    • Does not allow interlacing or interleaving.
    • Use this format as the final, post-edit version of the image that you will use in PowerPoint, Word, e-mail, and Web pages. Use a .TIF for the original to edit from and then save as .JPG.
  3. GIF -
    • Compressed graphics format developed in 1987 by CompuServe.
    • Lossless (no image information is lost in compression )
          Original -- 55565566665555
          Compressed -- 5(3), 6, 5(2), 6(4), 5(4)
    • GIF format reduces the color depth to a maximum of 256 colors to compress the size of the file. For most logos and text, 256 colors is more than enough, and will give you the best results.
    • Usually best for compressing low-color cartoon-like images, line drawings, and logos. (Note in the sample above how GIF's limited color depth affects a photo.)
    • Allows transparency. You can choose one color in the color palette that maps transparently when displayed. Gives the effect of seeing through an image or of an image floating on a background.
    • Allows interlacing or interleaving. Allows browsers to display a low-resolution image that appears immediately and continually redraws until it reaches its highest resolution.
    • Can create animated GIF images (as a single GIF file).
  4. EPS -
    • EPS files are often preferred and requested for publication by journals because EPS graphics are vector (line) graphics rather than raster (pixel) graphics and, therefore, can be resized without distortion.
    • This URL describes the difference between vector (scalable) and raster (pixels): http://www.cof.orst.edu/net/software/present/graphics/vector.htm
    • The FAQ below describes the process for setting up the postscript printer and software that you need and then how you create an EPS file by printing to a postscript file from any application and then converting the PS file to EPS.
      FAQ: http://wwwdata.forestry.oregonstate.edu/helpdesk/kb/faq.php?id=3040
       

File Size

  1. Larger files longer printing times and longer download times
  2. TIF - huge file size (lossless), JPG - smallest file size (lossy)
  3. Higher resolution  more pixels more information bigger file size
  4. Larger dimensions (inches) more pixels more information bigger file size
  5. More colors, greater color depth more information bigger file size
    (BUT, you can't save a JPG image at a reduced color depth.)
    For more information about color depth, see: http://fla.esf.edu/research/CAVLab/Reference/candbd.htm

Scanning and Editing Tools

  1. Flatbed scanner

    - Start | Programs | HP ScanJet Software | HP PrecisionScan Pro

    - In the Menu Bar, click Scan | Preview to get a preview scan of your slide

    - If the box (with dropdown) for resolution is not displayed in the toolbar, use Tools | Output Resolution to display the resolution selection box.

    - Select the Output Type True Color, Grayscale, Black & White (vector, raster), 256 color, Text
    This URL describes the difference between vector (scalable) and raster (pixels): http://www.cof.orst.edu/net/software/present/graphics/vector.htm

    - Click and drag with the crosshair cursor to select the area you want to scan.

    - Save (and scan) your image: Scan | Save As | and set the Save as type: in the dialog box


    To save text (OCR - Optical Character Recognition) --
    Click and drag to select the text area to scan -- Check in Output Type menu that Text is selected -- then Scan | Save As | and set the Save as type: to TXT
          OR
    Use Scan | Save as Whole Page


  2. Basic image editor

    IrfanView
    It's free and all you really need is the i_view32.exe file with no messy installation.

    If the program is already installed, you'll find it in Start | Programs | Graphics
    To install on your machine - Start | Run | \\forestry\apps\win32\install\graphics\iview.exe

    • The blue left and right arrows on the toolbar are useful to page you through the different images in the same directory as the File you have Opened.
    • The zoom icons, , are useful for images that may not fit within a fullscreen window.
      • And note the many options in View | Display Options
    • The crop function is in the Edit menu, Edit | Crop
    • All other image editing functions you're likely to want are in the Image menu.
      • Enhance Colors | Gamma Correction is useful for over/under exposure
      • Other interesting options -- Create Panorama Image, Effects, and Swap Colors 
    • File | Batch Conversion/Rename will let you convert a number of files all at once (like from TIF to JPG). You can even use the Advanced Options and also apply any of the edits in the Image menu at the same time (the same edits to all images, like resizing to thumbnails).
    • File | Thumbnails is a quick way to work with multiple images.
      Select one or multiple images (using the <Ctrl> key combined with mouse clicking) and then right-click a selected image to get options such as creating an HTML page or contact sheet image from the selected thumbnails.

  3. You can create animated GIF images using Animation Shop which is installed as another tool when you install Paint Shop Pro. If the program is already installed, you'll find it in Start | Programs | Graphics
    To install on your machine - Start | Run | \\forestry\apps\win32\install\graphics\pspro.exe
    The easiest way to create an animation is with File | Animation Wizard...
    View your animation with View | Animation
    Use Edit | Animation Properties, Edit | Frame Properties, and Effects | Image Transitions for basic modifications to your animation.


  4. Slide scanner available for checkout at the Helpdesk - Start | Programs | SprintScan35 | PolaColor Insight

    You'll see the following screen when you open the program. Click Preview to preview the slide scan.




    Go to the Scan tab to set the desired resolution. 

    You're also given some scaling options, by size or a percentage. 
    Don't get too caught up in having the scaling happen when you scan. Remember that you can scale in another program later and you don't have to do any scaling here. What's important now is, "How many pixels do you want to have to work with?" Set the Resolution accordingly. 

    When you click the Scan button you'll be prompted for file location and name and what file type to use.


Using Your Images

  1. Windows Explorer -- As you deal with multiple image files in multiple folders within Windows, you may want to make use of the Windows Explorer display options to View | Thumbnails.

  2. Screen Capture -- There are many times when it's useful to capture an image of your current screen display. (The Excel tip below describes one example.) 
    <Alt><PrntScrn> captures the currently active window to the clipboard
    <Ctrl><PrntScrn> captures the entire desktop to the clipboard
    Of course, once you've captured the image to the clipboard, you can paste the image into any application, likely a graphics editing program to crop the image.
    Remember -- Because you're capturing a screen image from 1024x768 pixels or so, the resolution will be fairly low, around 75 dpi.
    And on any Web page -- right-click any image in your browser window and Save Picture/Image As...

  3. When you insert an image into any Office application (Word, PowerPoint, Excel) the Picture toolbar will be displayed or you can use View | Toolbars | Picture. One of the Picture toolbar tools, lets you select a color in your image to be transparent. Note that the tool will be grayed out, not available, if the image you inserted was a GIF.

  4. Excel -- Need a high resolution image of your Excel chart for a publication? The normal Copy | Paste sort of techniques won't work as a way to capture your image and move it to a graphics editing program. Try this tip: http://wwwdata.forestry.oregonstate.edu/helpdesk/kb/faq.php?id=2893
    Another alternative is that you can make a registry change so that slides saved out of PowerPoint are saved at 300 dpi. See: http://support.microsoft.com/kb/827745/en-us

  5. MS Word -- You have a number of options for how you insert your images into Word. For the definitive page of information from Microsoft, including links to many more pages of information and descriptions of strange behaviors where, "Microsoft has confirmed this to be a problem," go to: http://support.microsoft.com/support/kb/articles/Q167/7/38.ASP

  6. Powerpoint --
    1. If you want PowerPoint to size a large number of photos, one per slide, use New Presentation - Photo album...  to browse PowerPoint to the images you want to include.
    2. BE SURE TO COMPRESS your pictures in PowerPoint or you'll have unnecessarily large images included which bloats your file size and drives up printing times.
      1. Click on a picture on one of your slides
      2. Locate the Picture Toolbar
      3. The options below are probably what you want to use.
    3. Insert | Picture | From File will let you place an image file on the current slide. If you want the image to cover the entire slide, you'll have to resize and crop the image to dimensions proportional to the slide dimensions. You can crop and resize the image in an image editor like IrfanView before inserting the image or you can use Powerpoint tools to resize and crop.

      Once the image has been inserted on the slide in Powerpoint, one technique would be to
      1. Click and drag the image to move it into position in one corner of the slide.
      2. Then resize the image so that the shortest dimension matches the slide and the other dimension slops beyond the edge of the slide. Be careful while resizing the picture not to destroy the aspect ratio and distort the image. (Click and drag a handle at one corner of the picture. Holding the [Shift] key down as you click and drag will preserve aspect ratio for all types of resizing in Powerpoint.)
      3. Now you can crop the part of the image that extends beyond the slide.
        To crop the picture from within Powerpoint you'll need the Picture toolbar --
        View | Toolbars | Picture
      4. When the floating Picture toolbar opens, click the cropping tool, Seventh from the left.
      5. Once you've clicked the cropping tool, position your mouse over a selection handle at the border of the image, and click and drag to crop using that handle.

    If you're in a situation where you can't crop anything from your image to make it proportional to the slide dimensions, then you could set the slide size to the picture size in File | Page Setup. Of course this affects all the slides in the slide show. You could use branching (one show links to another, ...) to get around this problem. With the image open in IrfanView you can see the images dimensions in the window title bar.
          PIXELS divided by RESOLUTION (dpi) equals INCHES


| About The College | Teaching | Research | Extended Education | Students | Faculty/Staff | Alumni, Donors, & Friends |

| News & Events | Publications & Products | Site Map | Search | Contact Us |

Forestry Computing Helpdesk, Peavy Hall 222
Oregon State University, Corvallis, OR 97331 | (541) 737-2152
Contact us with your comments and questions | About this site
Copyright © 2003 Oregon State University | Disclaimer