Some in the technology industry justify the ongoing inaccessibility of their visual verification schemes due to the claimed “expense” of implementing solutions such as audio CAPTCHA. Let’s evaluate a potential solution that could be built entirely on open source or otherwise easily available technologies at a sufficiently low cost to insist upon its implementation.

This solution relies on the presentation of the standard visual CAPTCHA along with a link to an audio playback of the characters to be entered. Though such implementations have been successfully put in place, this option should be appealing due to its overall simplicity and low cost.

  1. Manually record, using the human voice, an MP3 file corresponding to each possible character in the underlying CAPTCHA string.

  2. For each test, generate the standard image along with a M3U Winamp style playlist file pointing to all the MP3’s making up the string.

  3. Present the user with both the conventional graphic along with a link to the M3U file.

  4. The sighted user will complete the visual verification while the blind user will click the link to hear the audio in their media player.

This solution seems to enjoy the following benefits:

  1. Overall low CPU utilization. If pictures can be combined into a CAPTCHA, it won’t take much more to combine individual audio files into a M3U playlist version of the test.

  2. No text to speech software needed on the server.

  3. No complicated streaming solution needed on the server. It is open source, easily available and already implemented in most if not all web server software.

We hope this further helps the technology industry to make CAPTCHA, visual verification and word verification schemes more accessible and further reduces the justification for its continued use in ways that lock out the blind and visually impaired from participation in more and more of the Internet.