W3C home > Mailing lists > Public > w3c-wai-ua@w3.org > April to June 1999

Re: image action?

From: <meijer@natlab.research.philips.com>
Date: Thu, 27 May 1999 12:07:59 +0200 (METDST)
Message-Id: <199905271008.MAA29692@lark.natlab.research.philips.com.>
To: w3c-wai-ua@w3.org

Hi Jon and others,

I think what David meant is to use existing tags like IMG but
add the user option in browsers to select and "tap" the regular 
image output (as normally visible on-screen) and send the bitmap
to an application that can turn it on-the-fly into corresponding
image sounds. Software for the latter is already available from

   http://ourworld.compuserve.com/homepages/Peter_Meijer/winvoice.htm
   
but browsers would have to be modified to support the use of 
the alternative image rendering in sound. So we are not dealing
with wav files but with audio generated directly from gif's,
jpg's, and so on. Using the current plug-in technology would
not work because that requires changes to the web page like
adding specialized embed or object tags. The idea is not to 
change or extend the markup (perhaps that can be useful too), 
but rather to change the interpretation and processing of
existing markup, thus allowing for sound rendering of all 
images on all existing web pages, thereby avoiding any burden 
on website developers as well as avoiding the need for extra
storage space and Internet bandwidth that the use of wav files
would have involved: everything is done at the client side by 
his/her browser in cooperation with a powerful sonification 
add-on.

For instance, the McDonalds website 

   http://www.mcdonalds.com
   
shows amongst other images the well-known McDonalds logo with
the yellow golden arches making a big "M". After selecting this 
image in the envisioned modified browser and sending the bitmap
to The vOICe sonification software, this would result in the 
"logo sound" (here an 88K wav file for demonstration purposes)

   http://ourworld.compuserve.com/homepages/Peter_Meijer/mclogo.wav

This uses a left-to-right sound scan with height mapped to 
pitch. Blind website visitors could thus access and analyze 
the content of images on an HTML page. However, the link to 
popular browsers is currently still missing. Could there be
guidelines for this?

Sighted list members can check the spectrographic reconstruction
of the above sound sample in

   http://ourworld.compuserve.com/homepages/Peter_Meijer/mcvoice.gif

which was done by The vOICe Java application as available from

   http://ourworld.compuserve.com/homepages/Peter_Meijer/javoice.htm

Best wishes,

Peter Meijer


Soundscapes from The vOICe - Seeing with your Ears!
http://ourworld.compuserve.com/homepages/Peter_Meijer/

===

Jon Gunderson replied

> What type of markup are you using to associate the WAV file with the image?

to the message from David Poehlman

At 04:14 PM 5/20/99 -0400, you wrote:
>There is now software available that will provide an auditory
>rendering of images and this can be done on the fly from the web
>provided that the browser sends a construct to activate the software
>when an image is downloaded.  The software currently employed to
>handle the image to sound conversion in the form of a wav file from a
>bmp is called the vOICe which some of you may have heard of.  Peter
>meeger and I and others have veen discussing this on the ig list and
>It was my thought that we could ask him to work with this if it is
>felt to be viable?  This would have to be a user choice and would of
>course not be renderable by something like lynx, possibly pwweb speak
>and hpr as well, but could provide a level of functionality regarding
>images that is currently lacking.  The user would also have to undergo
>some training but we are working on that part of it to comprehend the
>audio.
Received on Thursday, 27 May 1999 06:08:04 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 27 October 2009 06:48:57 GMT