Display Techniques for Pre-recorded Audio

[wareid]

I am working through the ONIX Techniques for displaying metadata to determine how they translate to our systems and what we receive from publishers and authors. I have some particular concerns and questions about the "Pre-recorded Audio" section of the metadata, as I believe it's not entirely reflective of how audiobook content is handled or produced in market. I'll try to explain the issues in order of how they appear in the doc.

Reading this section overall, I'm unsure of the intention behind it. Pre-recorded Audio content (audiobooks, audio tracks for synchronised media overlays, audio/video content within an EPUB) seems well represented by most of the variables, but I do feel it is missing key information about the accessibility of the pre-recorded audio content. We explain the intention of "Pre-recorded audio" really well in the Principles document as:

Indicates the presence of pre-recorded audio and specifies if this audio is standalone (an audiobook), accompanies text (embedded audio and video clips), or represents an alternative to the text (synchronized text-audio playback).

I would expect a similar structure in the ONIX and EPUB techniques documents. I do believe the EPUB techniques document achieves this.

In Understanding the Variables, the variable all_content_audio is based on ONIX code 39 of list 196 (Supplementary material to an audiobook is accessible). While I see we explain the reasoning for this in that section, I think this variable will be confusing for publishers and inaccurate to how audiobooks and their supplemental content are currently presented to the user. Most supplemental content in market is provided separately from the audiobook, some audiobook platforms (mine included) do not bundle them together, and most audiobook platforms do not even have the ability to present supplemental content alongside the audiobook content. I think it would be more accurate for this variable to reflect the intention of the ONIX metadata field (ex. something like supplemental_content_accessible).

In Understanding the Variables, the variable non_textual_content_audio_in_video includes the ONIX codes 06 (Video), 24 (Animated / interactive illustrations), and 25 (Partial performance – video) from list 81. While its very likely content of this kind will contain audio, it's not guaranteed and we don't seem to offer a way for the content creator to specify that it does not. The intention with this variable seems unclear.

In Instructions step 1, it currently says:

IF all_content_audio AND NOT synchronised_pre_recorded_audio: THEN display "Audio only".

I believe this should include audiobook content (identified by ONIX code 01 of list 81), as currently in market most commercial audiobook content is audio-only. So instead I would expect this:

IF all_content_audio AND NOT synchronised_pre_recorded_audio, OR audiobook: THEN display "Audio only".

Similarly, step 2 in instructions says:

ELSE IF audiobook OR non_textual_content_audio OR non_textual_content_audio_in_video AND NOT all_content_pre_recorded: THEN display "Complementary audio and text"

I don't believe audiobook belongs in this instruction, as most (if not all) audiobooks in market do not have text at all. If step 1 is changed, removing audiobook from this step clears this up.

Instructions also appears to be incomplete, missing instructions for how to handle variables like non_textual_content_audio and non_textual_content_audio_in_video on their own or in relation to EPUB/textual content.

EDIT: I realized upon further reading that what might be missing or necessary is a clearer definition or segmentation of audiobook content vs EPUB content. They are two different formats, and while it is true that audiobooks are pre-recorded audio content, they follow different practices and workflows in the industry. Since ONIX does have metadata in list 196 to address audiobook accessibility, I do think it's worth us specifying that in these documents, but I do think we need to clarify audiobook vs EPUB.

[chrisONIX]

Small comment on this sentence: “While its very likely content of this kind will contain audio, it's not guaranteed and we don't seem to offer a way for the content creator to specify that it does not. The intention with this variable seems unclear.”

ONIX list 81 does attempt to provide for this - code 50 - Video without audio and code 24 - Animated / interactive illustrations - are both meant for moving images without audio necessary for understanding the content.

[chrisONIX]

I agree that it is a good idea to make clear the different approach for audiobook content vs EPUB content (vs other content?) - this will make the display techniques easier for data receivers

[wareid]

ONIX list 81 does attempt to provide for this - code 50 - Video without audio and code 24 - Animated / interactive illustrations - are both meant for moving images without audio necessary for understanding the content.

Code 24 is in the list currently, so maybe it should not be?

[chrisONIX]

I propose removing code 24 from the pre-recorded audio section. At EDItEUR we will add to the note of list 81 - code 24 to make it clear that these animations have no sound or audio necessary for understanding the content

[wareid]

@gautierchomel I see the commit removing code 24 from list 81, but that wasn't the only comment from this issue. I don't have concrete proposals for my other comments but I think they warrant some attention, unless I've missed that they are addressed in other issues.

[gautierchomel]

@wareid my apologises for that unintentional shortcut. Those are good points to document and discuss.

I agree with your first comment edit that this is a discussion related to Format. Our intention is to be as Format agnostic as possible for this guide, specifically in the case of the ONIX techniques. the EPUB technique is clearer because it applies to one only format.

Format filters and information are typically provided independently of accessibility metadata, ensuring that finding an Audiobook, as defined by today's provider usage, remains feasible.

This additional accessibility information is provided to cater to readers' needs, providing accurate information about whether a book can be read with ears without any loss of information.

• 81-01 says that primary content is audio, it is not sufficient by itself to state that all content is available as audio.
• 196-39 is a new code that corresponds to accessModeSufficient: Auditory. It allows to specify that all content is available as Audio. Being new, it has to be appropriated.

Confusion between audio and accessible may generate misunderstandings of the "complementary audio and text" mention displayed for audiobooks, but as ONIX can be updated independently of the file, I guess that quick changes will happen, and "full" audiobooks will get 196-39 added quickly. The others are not fully readable in audio, and the information is useful for those who choose a book.

Another point in your comment that I see not addressed is:

Instructions also appears to be incomplete, missing instructions for how to handle variables like non_textual_content_audio and non_textual_content_audio_in_video on their own or in relation to EPUB/textual content.

I interpret that as a need of instruction to content creators (i.e. when do I have to add that metadata) which is not covered by this Guide (see issue #234).

I suggest opening dedicated issues for further discussion on one point or the other (and any more that I may be missing).

[gautierchomel]

@wareid, are there still points in this issue that need to be discussed?