VisiBone Palette Survey

You can measure how many users have small color palettes by inserting the following HTML and JavaScript on a web page, as is done at the bottom of the VisiBone Home Page:

     <p>Your screen is in
     <a href="http://www.visibone.com/palettesurvey.html"><script><!--
          // VisiBone Palette Survey, Version 1.1, 21-Jan-2000
          // © 2000 VisiBone, patent pending
          var depth;
          if (window.screen == null 
           || window.screen.colorDepth == null) {
               depth='unknown';
          }
          else {
               depth=window.screen.colorDepth;
          }
          if (depth == 'unknown' || depth == 0) {
               if (window.screen != null
                && window.screen.pixelDepth != null) {
                    depth=window.screen.pixelDepth
               }
          }
          document.write('\<img src="'+depth+'.gif" '
                              +'alt="VisiBone Palette Survey"'
                              +'border=0\>');
          // --></script><noscript><img src="unknown.gif"
          alt="VisiBone Palette Survey" border=0></noscript></a>
     bit color mode.</p>

Then use your site statistics to measure the hits on 1.gif, 2.gif, 4.gif, 8.gif, 15.gif, 18.gif, 24.gif, 32.gif, unknown.gif. You're welcome to use the following image files, though they're designed for the home page colors (#CCCC99 on #330000):

Kindly link whatever images you use back to this page. Results for the VisiBone Home Page to be published here. Report back after running this a week or so and I'll add your results too. Thanks to Beth Santoro for the pivotal clue of the existence of screen.colorDepth.

Intial results I had reported here were in error. Smack me upside the head but I was counting hits from this page! Since all the number.gif files appear above, every hit to this page counted as a hit to every palette size. Duh.

Now things are much better. Not only am I (1) counting hits via the home page exclusively, (2) I'm weeding out 304 status codes, which are duplicate visits, and (3) I also removed my own personal hits to my home page.

Results from part of April 2000:

     color depth         color depth in
hits in bits      Pct    number of colors
---- -------      ---    ----------------
 986   16         38%    65,536
 831   32         32%    16,777,216
 291   24         11%    16,777,216
 243 unknown       9%    ?
 140    8          5%    256 (web-safe benefits)
  42   18          2%    262,144
  10    1          -     2 (monochrome)
  10    4          -     16 (VGA)
   8   15          -     32,768
   9    2          -     4
----             ----
2570             100%

The total number of hits at 256 colors is 5% of the total. This is lower than reported by www.thecounter.com (8%) which makes sense: webmasters and designers use this site and tend to have bigger palettes. (I'm frankly surprised any have 8-bit palettes.) The unknown figure at thecounter is over 21% though. Still don't understand that. By the way, the 16-color VGA and monochrome computers get no advantage to web-safe colors.

Here are results for July 2000:

     color depth         color depth in
hits in bits      Pct    number of colors
---- -------      ---    ----------------
1916   16         42%    65,536
1571   32         34%    16,777,216
 580   24         13%    16,777,216
 242 unknown/-1    5%    ?
 216    8          5%    256 (web-safe benefits)
  73   18          2%    262,144
   6    1          -     2 (monochrome)
   3    4          -     16 (VGA)
   3   15          -     32,768
   1    2          -     4
----             ----
4611             100%

From the object model:

COLOR	HTML	LAB SWATCH CODES	HTML TAGS CHAR	REGEXP CC POP	faq bulk press about home stein@visibone.com color wheel color blindness links articles tnar contact
PRODUCTS		— FREE COLOR —	— FREE HTML —	— FREE ETC —

© 2002 VisiBone