Sound not present in context

[glenrobson]

From IIIF Slack discussion: https://iiif.slack.com/archives/C0AL6BU76/p1736882496943739

From: Shaun Ellis:
I'm working on a viewer component and noticed that the Presentation API 3.0 has a list of recommendations for motivation painting type. If these are simply recommendations, how does a client know for sure whether to present a player or a viewer? For example, I'm seeing some manifests that use "Audio" and others that use "Sound". It seems odd to me that "Sound" is recommended since the mime type is Audio and the HTML tag is .

Would it be better to have this be a controlled vocabulary list that MUST be used rather than a list of recommendations? (Apologies if this is already being discussed/addressed in 4.0.)

---

From Stephen Fraser:

So the classes from the Presentation 3 spec are from this:
https://www.w3.org/TR/annotation-model/#h-classes

"Dataset": "dctypes:Dataset",
"Image": "dctypes:StillImage",
"Video": "dctypes:MovingImage",
"Audio": "dctypes:Sound",
"Text": "dctypes:Text",

Confusingly though, the annotation context doesn't have a Sound class, only an Audio one.
https://www.w3.org/ns/anno.jsonld

The complete example:

https://www.w3.org/TR/annotation-model/#h-complete-example

Also uses Audio .

So it seems that Sound may be an error in both the W3C annotation model and Presentation 3 specification.

"Dataset": "dctypes:Dataset",
"Image": "dctypes:StillImage",
"Video": "dctypes:MovingImage",
"Audio": "dctypes:Sound",
"Text": "dctypes:Text",
"Service": "schema:WebAPI",
"Agent": "dcterms:Agent",

Sound is not the in Presentation 3 JSON-LD context either.

[azaroth42]

The DCMI types class is "Sound": https://www.dublincore.org/specifications/dublin-core/dcmi-terms/dcmitype/Sound/

The Presentation API list is derived from the Annotation list, which is where the bug is introduced in the Annotation context by mapping "Audio" to dct:Sound instead of "Sound". I'll file an erratum and we can add the mapping to the Prezi context ASAP, as contexts are not semantically versioned, the documentation is.

So in short, the answer is to use "Sound" and that it will shortly work if you were to expand to triples.