Administered by:
I saw a post a while ago (can’t remember the source, sorry) that stated that screen readers don’t say CH4 and CH₄ the same way; that it’s better to type CH4. My computer (running macOS) reads them both out correctly. Is there a website that can read submitted text to me , so I can check my work? I’d really like to type chemical formulas with proper subscripts, but not at the cost of accessibility…
(she/her) @agvbergin@mastodon.me.uk@MichaelPorter So I notice in this post you have used a Unicode subscript four character, and I think this is what causes a problem for some screen readers?
On social media where we don't have formatting available, I think it's best to just type "CH4".
Where you do have formatting available (like Word, or a markup language like LaTeX or HTML) then use the tooling or markup in that software to do the formatting rather than using a special Unicode character.
@agvbergin I’m mainly interested in stuff I post on Mastodon - trying to do well by folks who need accessibility features. The only way to get a subscript in this forum is through unicode, I would think.
If the macOS software can find the character in the character viewer using a search for “4” and can read it as a four when speech is enabled, one would think screen reader software could, too. But I’ll bet such software gets updated slowly, and minimally.
@MichaelPorter No, that's not how it works. Using Unicode to imitate formatting is what causes the accessibility problems.
@agvbergin I am *completely* ignorant of how the screen reader programs work, for sure (hence the original request for some kind of web simulator for me to get the experience).
It just doesn’t feel like it would be insurmountable to add the capability to interpret “U+2084” and read out “four” – After all, as I said previously, the system software on my computer manages it just fine. And the software reads out emojis, no?
I’m sure that the vast majority of vision impaired folks don’t need/care for this, but the coders who produce the screen-reading software are no doubt aware of unicode… Maybe it’s just not a priority.
@MichaelPorter @agvbergin
You could write it the way it's done in LaTeX, precede the subscript part with an underscore, CH_4. It's a plain text convention that's become known beyond those who know LaTeX (caret ^ for superscript is even better known).
Most screen readers by default will say "c h underline 4" but one popular screen reader, NVDA, will just pause between the H and the 4. However, if you provide context that indicates there's chemical notation, any screen reader user can navigate it character by character and get every one.
@cwilcox808 @agvbergin My computer and Natural Reader both read it as “C H 4”, no underline, which makes sense if it’s interpreting it as a space. And, as you say, follows established conventions. But just putting the numbers in would work (eg CH3COOH) as long as you don’t have to worry about charge. Natural Reader sounds out “caret” while my Mac doesn’t.