- From: Pyatt, Elizabeth J <ejp10@psu.edu>
- Date: Fri, 16 Dec 2022 14:42:56 +0000
- To: caroline <woodward.caroline@gmail.com>
- CC: w3c-wai-ig <w3c-wai-ig@w3.org>
- Message-ID: <1860F146-8BE5-4309-B8F8-A3DCFF09E472@psu.edu>
Caroline: You are correct that the context of the document is critical to determining ALT text of the image, so you are actually thinking in the right direction. Some of what I say below is opinion, but some is based on input from actual users. At this point, I would pass on the excellent advise from WebAIM - pretend you are reading the document over the phone. What parts of the image do you describe? If you're instinct is to skip the image, it may be decorative or repetitive. If you're not sure, my advice would be to add an ALT text. They have some great examples of how the same image can have different ALT text.<https://webaim.org/techniques/alttext/> FWIW - I believe you won't need long descriptions of most general audience content, with the exception of those times an entire passage of text is embedded in an image. Screengrabs of warning notices are especially problematic. Academic or research content often requires long descriptions of some sort. Penn State has some information about describing complex images<https://accessibility.psu.edu/images/> and extended descriptions using the HTML SUMMARY/DETAILS tag<https://accessibility.psu.edu/images/imageshtml/#extended>. What I like about these solutions is that everyone, including people who may not be able to perfectly see or understand a graphic can access the information. FWIW - I do agree that images that evoke strong emotions should be described, but for me that's a laugh or a gasp. Maybe not "smiling lady images" per se but I probably would describe a smiling lady winking while holding up bunny ear fingers. Just make sure your ALT text is professionally worded. Hope some of this helps. Elizabeth On Dec 15, 2022, at 1:22 AM, caroline <woodward.caroline@gmail.com<mailto:woodward.caroline@gmail.com>> wrote: Recently, I've wondered whether further clarification about images 'communicating meaning' is needed, and curious if others have wondered about this too. The topic of when something meets the definition of 'decorative' often surfaces when in discussions with other accessibility specialists, which naturally also includes whether something needs alt text or not. 'Essential' or 'meaningful' still seems ambiguous. I personally feel it depends on the context. I've heard folks talk about how illustrations should have enough color contrast for low-vision users to recognize what it is. But the images being discussed weren't communicating anything to the user. They were brand-related illustrations used to help ground the space and make the page look nice. Another felt that people should have access to the emotion being exuded by the image. Hearing this, I'm inclined to think that it would depend on the context, like whether there is value in the person being informed that a person is smiling. There's also considering the audible experience and how random descriptions like 'two hands shaking' sound if it isn't critical to helping the user understand something. These conversations go in circles. My concern is the impact of not being able to agree on when something meets the definition of decorative because of its impact. It affects the people who are trying to learn from us and those we advocate for. This topic isn't limited to where I work, but since it came up again, it's really been on my mind. Thanks for bearing with my long-windedness and in advance for any 2 cents that may bubble up. =-=-=-=-=-=-=-=-=-=-=-= Elizabeth J. Pyatt, Ph.D. Accessibility IT Consultant/Lead Trainer Penn State IT Accessibility https://accessibility.psu.edu accessibility@psu.edu (General accessibility questions) ejp10@psu.edu Ph: 814-865-0805
Received on Friday, 16 December 2022 14:43:13 UTC