@stefan I try to add ALT to most of my images but It's hard to describe a picture and to what details.
Feel like AI tools can be tremendously useful for this use case. There are a few apps out there that help in this regard but need one integrated to this platform.
As for using AI, I think this is an interesting idea, and I think it can work really well if the algorithm can provide a few suggestions, with different focus of the description, so that you can choose the most relevant one.
Yes, if you click on the current text, which simply says: "No description added", you will see an explanation prompting you to "Describe for people who are blind or have low vision".
But you have to click the first link to see it, and I am guessing most people just don't.
@stefan I will usually caption/alt-text images, as I have reminders in place to remember to do it.
The reason it doesn't get done as often as it should, IMO, is that people aren't in the habit of doing it. So making it mandatory in clients would be one possible option, or adding reminders during the posting workflow.
@martinhowitt I don't personally like the idea of making alt text mandatory, my concern would be people filling it up with garbage just to get past this requirement.
There are also valid cases when you really can't add the description right away, for instance, if you're in a protest and things are getting heated as you document the situation, or maybe due to a physical disability. Hence we have the "alt4me" hashtag.
Either way, I appreciate you thinking about how to fix the issue!
@stefan i usually caption but when I don't it's either when the post/ comment is meant for only particular people who are not visually impaired and if the context is already provided by the post. In the second case, I usually will mention that in the alt-text.
@stefan I think there's a combination of reasons for me, but a common reason I "forget" is because ivory's share dialog doesn't allow alt text entry when I'm sharing.
Other reasons are things like the image just doesn't matter at all (e.g., the post would be understood no differently with an image or not)
@stefan
I tend to caption my images, but I voted for 'Not enough time' as sometimes you just want to post something right in that moment.
Also: as stupid and non-inclusive as it is, I think some jokes are so much funnier when there is no 'bottom text' to rely on for an explanation of the joke.
@stefan I am now in the habit of writing alt text for three reasons: 1) it is a courtesy that I learned is helpful and important to other people here in the Fediverse; 2) it can be really fun bc I can write little extra notes to people there; and 3) I subscribe to a bot that will chastise me on my timeline if I do not. That has been the biggest deterrent!
@stefan Yes! And thanks. I really do want folks who use screen readers to be able to enjoy Mastodon as much as people who can see photos do. I really try to do a good description, but I’m also open to feedback if I fall short. I try not to boost photos that don’t have alt text, though I will star them. I always click on alt text in other people’s photos, too, to learn.
@stefan 98% of the time I caption images, if I don't it's either because I don't think/want many people will see it (usually happens in replies), or I forgot, or it was something I didn't know how to describe or explain (like a piece of abstract art or a confusing meme).
@stefan I sometimes don't add alt text because of the somewhat poor UI on the @elk alt text compose screen.
For some reason, I will click "Close" rather than "Apply", deleting all of the text that I wrote. I will then post, thinking I included alt text, and don't realize my mistake for a few minutes. Then I have to go back and edit the post to include the alt text. If I'm just copying and pasting, that's fine. But sometimes I have to rewrite the whole description from scratch. Ugh.
@Stefan Bohacek This advice is unfortunately too much geared towards those who post pictures of a cat or a bird or a flower or something. It doesn't work if the picture doesn't emphasise one specific element or a few but showcases the whole scene. Then there are no "important details", and at the same time, everything is or can be an "important detail".
And then you have no other choice but to re-tell the whole picture in words.
Oh, and if a blind or visually-impaired user is completely unfamiliar with the very setting of the picture, they might be curious about what else is in the picture that does not fall under "important details".
I've been captioning my images for two years now. The reason I didn't for my first year on fedi was that I was uploading art and I didn't want to try to put my own art into words. I felt that if I could say it I wouldn't have to paint it.
It's waaaay easier to caption a snarfed panel from some manga or something, or a frame from a show or movie, which is most of what I upload now. Other people's art. I can easily take my own reaction or interpretation and put that into words.
I'm now really committed to captioning my own art from now on, and have done some with my last couple of dozen paintings, but that also has limited what I paint to things that I can express semantically. I sometimes wish I could get a li'l help with describing things like this: https://idiomdrottning.org/the-holy-one-bends-and-contracts.png
(I don't remember what description I did use for that when I uploaded it.)
Add comment