@smitten Ah yeah, and admittedly it's cause I've never taken the time to write good alt text and often didn't write any. I will now. I was wondering also if the alt text field will replace a screenreader's source even if it's blank. I guess this is really a question about individual screenreader softwares tho now that I think about it.
@smitten hmm cool sure ok. Thank you. I'm also wondering why that's a tolerable effect of the alt text field-in other words, does it disable extant screenreader features in that sense then even tho the alt field was intended to provide perhaps more comprehensive description than the image or text within it or in its caption alone? but people just don't use it properly?
@RedtheBean I'm not sure I understand your question sorry. do you mean that people write alt text that is maybe worse than what an automated process could produce? I have seen that happen, sometimes people will just write "meme image" or something like that which doesn't provide much information. but the best manually written alt text is interesting to read on its own, like how people might describe something on a podcast or the radio. here's a good example I found on the silentsunday hashtag - https://socel.net/@KarenKasparArt/111516496292833774
@RedtheBean the main text of the post should work fine for a screenreader (it can get messed up if the client is written badly but people with screenreaders can find a client that is built more accessibly). I'm just referring to the image alt text which needs to be written manually. when a screen reader comes across an image without alt text it will just say "image" or sometimes it will say the filename which can be super distracting.
there are tools to auto analyze images and try to describe them with AI, and some screenreaders have those tools built in, but it's not very reliable compared to the original poster describing the image.