Re: Chromatic font research (returning from Re: UniCharacter (Re: Codes for codes for codes for... (RE: Chromatic font research)))

From: William Overington (
Date: Fri Jun 28 2002 - 06:44:45 EDT

Arnold Winkler wrote as follows.

>Folks, WAIT A BIT.
>This method, as tempting as it is, would make all text "not accessible" for
>people with visual disabilities. And, as you all know, Section 508
>that any electronic information from the government (e.g. web site) must be
>accessible to people with disabilities.
>Here goes a great idea unless we find an accessible way to "display" colors
>for the blind ! Assistive Technologies companies - here is your challenge

An interesting aspect of computing in relation to people with visual
disabilities is that there exist speech synthesiser systems which read out
the words of a web page. I have not had the opportunity to observe one
being demonstrated yet my understanding is that for web pages they read out
what is in the ALT section of graphics.

Readers might, in passing, like to have a look at a web page which I
produced some years ago.

You raise an interesting point which has made me think further about
chromatic font technology.

I wonder if it would be possible for any specification for a chromatic font
technology to allow a font designer to add one or more text strings into the
collection of items that can be accessed for any particular code point, with
the intention that a specialist speech synthesiser could use any such text
to produce sound which a person with a visual disability might be able to

So, for example, if a red flourish were activated, the speech sysnthesiser
could receive the phrase "with a red flourish" from within the font so that
the speech synthesiser could say it out loud. I realize that there are then
various issues which arise as to which language the message should be coded
in and whether each glyph should have a message or whether the font should
have a collection of messages each having a message index value and then the
glyph could have a message index number associated with it if a one bit flag
states that a message is present. This would mean that a font with no
messages in it would have a 0 value for that flag for every glyph, and that
a font with just a few such messages would have a 0 value for that flag for
most glyphs and a 1 value for that flag for a few glyphs. Those glyphs
which had a value of 1 for that flag would then also have an index value to
the list of messages within the font. Normally, most speech synthesiser
usage would just use the text within a font for additional speech, yet the
possibility could be designed in for the speech synthesiser to be set so
that it only uses speech from the font and the font contains speech
information for each glyph. This might be very useful for encoding speech
for minority languages where the speech synthesiser technology might not be
supplied to support that particular language yet the desired result could be
obtained by having the font contain speech expressed using phonetic
characters with the speech synthesiser supporting the use of those phonetic
characters. Thus one speech synthesiser design might provide a part of the
support infrastructure for people with visual disability in many minority
languages, with the font technology providing the remaining part in a
standardized manner.

It is important that this matter has been raised as, if and when chromatic
font technology standards are devised, it may well now be that facilities
for storing such speech information in a font will be included in the
specification. Even if many fonts do not carry any speech information
themselves, the availability of the option may prove highly useful to those
that choose to use it.

William Overington

28 June 2002

This archive was generated by hypermail 2.1.2 : Fri Jun 28 2002 - 05:05:55 EDT