Acrobat X reads all of ai map

The text (out of Illustrator) is present. Most likey it is not tagged. The text is, maybe, vector?
In a like manner, OCR text (say, Searchable Image (Exact)) is present after OCR of a scanned page.
.
You can use TORU to apply the Figure element to the entire virtual page & give Figure some alternate text.
ROL will still read the OCR text. For this situation you can select all the OCR text on the page.
From the Tags panel Options menu you would create a tag from the selection. (Say the paragraph element (P)- does not really matter as it'll be removed latter.
.
The structure tree will now have this element and the element will have 1 or more child containers holding the text. Select a container, right click, select "Change tag to Artifact". Do so for each container. Delete the element (tag) created earlier.
.
ROL will no 'read' the Figure alternate text only.
.
For text out of Illustrator, that is associated with the graphic/image - try putting it to raster - so, no "text" for ROL.
.

OCR explains a lot A friend said ai placed vector map's figure tag with Alt text reads fine in JAWS screen reader 12. And that my complaint is a "Feature."

Another Q: in tags for figure Alt text doesn't show but I get Tags>Article>Story, Span, Story, Story, Story, Figure, Figure
Below, in both there is a box with text PathPathPathPathPathPathPathPathPathPathPathPathPathPathBut in Order: Alt text for figure shows on page in highlight Figure.

What is and why PathPathPathPathPathPathPathPathPathPathPathPathPathPath?

Thanks for your much appreciated help.

"Path" — The lines (vertical or horizontal) that make the sides of a frame (text or graphic)or an underline, etc.
.
AT ignores.
.
Tags — affirms you are viewing the Tags panel.
.
Document and Article are Block Level Structure Elements (BLSE) for Grouping of other BLSE's - Grouping elements do not directly associate with content.
.
Story (from InD) role maps to PDF element Section (also a BLSE Grouping element).
.
Span is a Grouping element for inline content (text string inline with paragraph's flow but has/may have something special/different about it). Often seen in Tagged PDF that is strongly structured.
.
Figure, like Formula and Form, is an illustrative element. Figure is the element used for graphic content (graphic - graphic can be of "text" but no renderable text)that contributes to the document's semantics (aka "real" content).
.

Getting a lot of Spans with 1 space in them between lines of text. Will this throw the reading off?

Kat,
I'd be guessing. Can you share the PDF? If so, you could open a free acrobat.com account. An uploaded file can be shared via link which you could post here.
.
acrobat.com
.
online Help for upload and share
.

Access_issues.pdf is up at http://www.mcgraphics.us/source/icon.php
I did a full check all the empty spaces I took out were flagged as not in structure. But doc seems to Read out loud ok.
Any suggestion on anything I could do better. Client needs to make decisions on what to do about the number-icons and Alt text for maps and photos.
This is much less scary with your great advice.

Kat,

Span tags with no "text" —
PDF page 3.
— Table's header row, cell with "Ages 13 to ..." - It appears that you've entered a non-breaking space to put the "Share these..." text string on a seperate line. The [Span] tag identifies this. It'd be ok to leave this.
— Table's "Activity" row, second cell, "... computers) to" is followed by, perhaps a non-breaking space, to keep "1-2" together. The [Span] element identifies this. It'd be ok to leave this.
— 6 others located in the table's cells. May be associated with an inserted non-breaking space or a space character. I'd leave it as-is.
.
Next is on PDF page 4. Line with "Look it up by facility number." It appears that a non-breaking space has been placed after the word "each" to push "facility has..." to the next line (To avoid hyphenation? If so, just turn off auto-hyphenation - which is nice when only producing paper - for "eDocs" I'd turn it off. It is something that lingers from the days of typewritters and paper - trust me, I lived that so it is not hyperbole.)
.
Next is on PDF page 4. Looks to be a space character "in" the circle (having no fill). Likely associated with how you positioned associate line of text.
.
Next is on PDF page 6. Between the list label and "Office appointments:" and again at end of "Office appointments:" - First looks to be associatedwith the space or tab used in InD's placement of the list label (the "dot") from the paragraph tag's auto-number feature. Second looks to be associated with the use of a non-breaking space to keep "M-F" starting at its own line. Third follows second after "12:30 p.m." Looks to be associated with a non-breaking space used back in InD to push the text that follows to its own line. Next is after list item #2's label ("dot"). Same cause as first. Additional occurrences are present with each of the remaining list entries. Similar causes to what is discussed above. Effectively, a consequence of how content was mastered; but, workable by AT.
.
Some observations.
.
--| 1.
Language - Set in advanced tab of document properties. However, PDF page 2 contains content having Spanish as the language. You must identifiy this at the [Body_-_Front_Matter] tag which is the first child of the second [Story] tag. Go into the structure tree of the Tags panel. Select the tag and right click. Select Properties. The dialog opens to the Tag tab. Provided the appropriate value for "Language" there.
.
Now, restore the Language back to English in the next tag ([Figure]). While the content shown is chinese it is an image of the text; so you could set the [Figure] tag's Language to English. Actually, I'd prefer to reset Langauge to English for the content of the third [Story] tage under tag [Article] FM.
.
As is, you have effectively precluded end-users whose language is Chinese and who use assistive technology (AT) from making use of the Chinese text (as it is only an image & there is no Chinese language Alternative Text. So, while the text is "seen" it is not present / available for use by Chinese speaking/reading users of AT.
.
Note that if actual Chinese text were available it is read right to left in vertical columns.
.
Consequently, for the as-is PDF, returning to Language = English at the first [Figure] tag is "workable".
.
.
.
--| 2.
Alt Text for [Figures] —
PDF page 1, Figure: "3 kids jumping" — Just an observation, looks like the visual means to show Mother - Father - Daughter leaping from a run while on an excursion in the great outdoors. Implies healthy, happy, family activity. Alternate Text for a Figure tag is meant to convey the intrinsic "meaning" of the image/graphic. If not worth the word smithing then make the "Figure" content an Artifact.
.
.
PDF page 2, First Figure: "Chinese text col 1" conveys little. In English; so, what of the Spanish & Chinese speaking end-users using AT?
PDF Page 2, Second Figure: "Chinese text col 2" - same as above.
.
.
PDF page 5, Figure: "Area Map of Central California" — Alt text provides no discussion of "why" the map is presented. A user of AT has the same need as those without a vision impairment. The map provides the those without vision impairment information for a specific locationg (central California covers a lot of turf, no?). The map depicts, by the graphic "markers", specific facility locations. The PDF page's light blue banner line above the map is unmarked content and not identified as an Artifact (which it should be). The map contains renderable text (the numbers in each of the "markers"). Any screen reader is going to render these in a chaotic stream. Better to create a stand alone graphic (JPEG or PNG) that can be imported into InDesign (with FrameMaker I'd import it by reference into an appropriatedly configured Anchored Frame which I'd provide appropriately descriptive Alt Text - e.g., Identify we've a map and denote a specific geographic locality. Provide a phone number for contact to speak with an person to learn what facilities are where and so forth. After all, we are providing those with no vision impairment this fundamental information visually. Yes,
with time and effort, the map's content could be provided such that it provides usable information to a user of AT. However, a designated "help" phone number is,perhaps, better.
.
.
PDF page 6, Figure: "Campus Map of Fresno Medical Center" — chock-a-block with unmarked content that is not designated as Artifact. Better to have this content in a graphics format (as above). No renderable text in such. Use Figure element's Alt Text to provide an appropriate description and provide a "help" contact number.
BTW, many FedGov agencies use the "help" call approach for providing information about complex content that does not lend itself to effective accessibility rendering.
.
.
.
--| 3.
Use of "Headings"
.
PDF provides H and H1 through H6 for Headings. Application "headings" either role map to these our directly map to them. When H1 - H6 are used the *must* be present in sequence. An H1 first. Others, in sequence, follow. No "skipping". Fraggled PDF Heading tags can/do gobber AT's use of the Tagged PDF. This can/does make the PDF's usability rather problematic. In a weakly structured tagged PDF (e.g., little or no element nesting) is needful that headings proceed from 1 to 2 to 3, etc.
Of course, at a given heading level you can repeat the heading (H2 & H2 as sibling elements).
.
Start in the authoring environment. Establish the appropriate document parapgraph tag / character tag hierarcy for the document content/semantics. Then use it. Remember, FrameMaker - InDesign - MS Word and thier associated Tag management facilities "understand" that the *Built-in* Headings paragraph tags / paragraph styles are go become PDF H1 - H6 when they produce the Tagged PDF. For "headings" - always rely on the built-in headings.
.
Look at PDF page 6. We start with [H3] then [H1]. Then [H3] and [H3]. The string "Fresno Medical Center..." (above the map) is and [H3]; no role mapping which is indicative of the InD paragraph tag used being a Built-in "heading" tag for a third level heading. The same thing for the string "Admittng" and the string "Adult Medicine". The [H1] (Fresno Medical Center) also stems from an InD Built-in heading paragraph tag (the level 1 heading).
.
In FrameMaker the paragraph tag "TableTiltle" becomes the PDF tag [Caption] directly. There may be a similar built-in paragraph tag for InD for use as [Caption]. The string "Fresno Medical Center (hospital and medical offices)" is, actually, a "caption" to the map held in the [Figure] element (tag); so, [Caption] rather than [H3] is appropriate.
.
Note - At a word's end we use the space key to break text flow between words. At a line end/wrap this can leave a space character associated with a Span tag. Keep those as it promotes smoother parsing of the content by AT.
.
.
.
--| 4.
.
PDF page 2 —
The first word "consideren" - uppercase the first letter?
A useful post-processing activity is to use the Find Element feature. Use each of the Find choices. If working with a living document you eventual correct issues in the authoring file so that "it is right upfront". For page 2, Find Unmarked Content identifies that there are two hits associated with the Spanish text flow. After "... la semana." and "...Interpretacion Indirecta". Neither is an Artifact. It appears that the second of the two space characters used between the adjacent words is unmarked. (FrameMaker has "Smart Text" which precludes a second space character between words - perhaps InD has this as well?)
.
.
PDF page 3 —
Page holds a table. The table's caption (Preventive care guidelines...) uses InD paragraph tag {Table_Big_Hed} which role maps to a PDF [H2] element. The document's logical hierarchy (and usability by AT) is better served is the text string was content of the [Caption] element.
The table lacks a Table Summary. The left most column's cells are all tagged with the [TH] element as are the cells in the first row (the header row). Because of this it is needful that cell ID (element identifier) and Headers attribute. Lacking these the usability of the content in the [Table] element is compromised. I find use of ID and Headers an interesting endeavor. However, most times for most tables a "zoom out" look-see shows that the table content can be master in a manner that keeps a simple table structure while maintaining integrity of the table content semantics ("kiss" trumps "wiz-bang").
.
Because, back in InD, you have spanned some table cells you need to V and V that this has ported into the Tagged PDF correctly. Use TORU, click upper left number associated with the table, in TORU - click Table Editor, select appropriate cell, right click, click Table Cell Properties, do V&V. Is Row Span and Column Span correct? [TH] cell's "Scope" is not set. You must do this in the Table Cell Properties dialog for each [TH] of a Table (for all Tables in a Tagged PDF).
.
Use of color as cell background/fill — Poses an issue for anyone who needs to use High Contrast to facilitate use of the PDF. If one needs white text on black background then the content of the cells in the first row "go away". If the client calls for an Accessible PDF then they (& the provider of the PDF) most take something of
a jaundice view of "eye candy" in the content. Certainly, the content can be presented in a pleasiing manner; just have to invest time in trials of different approaches to validate the visually pleasing is still usable to those who rely on AT. Without this due diligence the result can be a PDF that is unusable to a segment of the target audience.
.
.
About PDF page 4 —
The InD paragraph tags {Divider_Headline} and {Headline1} role map the the PDF tag [H1]. The PDF page's logical hierarchy might be better expressed to AT (and its users) by "Facility Directory" being a level 1 heading, the "About this section" being a level 2 heading and the "To find a specific facility" being a level 3 heading.
.
The 6 "marker" graphics on the page. If the Tagged PDF is meant to be accessible it must follow the underlying principles described in the PDF ISO Standard (ISO 32000-1). This provides a Tagged PDF that is, at least, "workable" by AT and thus usable by AT end-users in a user acceptable manner. Use of Read Out Loud as the "green light" is a fundamental mistake. There are a number of post-processing activites called for to geat the PDF prepared. Then, if possible, have someone use NVDA to check the PDF (other AT possible - NVDA is available at no cost).
.
Back to the 6 "markers" — If nothing else, each can be tagged as a [Figure] element with TORU. Provide each with Alt Text. Example (for "Medical centers" marker) —

"This figure depicts a square having a blue fill" and represents medical center facilities."
.
Also, the "4" in the blue box. Use a fully graphic representation (JPG, PNG,...)
.
.
PDF page 5 —
As mentioned previously, as-is all the numbers are rendered. Confusing. A graphic file holding the content can serve the user with no vision impairment. For those with vision impairment, the content is in a [Figure] element having appropriate Alt Text.
.
The string "Maps and locations - Centeral California" use InD paragraph tag {Heading_1} which role maps to PDF element [H1]. The document's logical hierarcy would be better support and provided to AT by the text string being the content of a PDF [Caption] element.
.
.
.
PDF page 6 —
As mentioned, heading elements not ordered appropriately. "Map not to scale" is identified as an Artifact. As it is provided for viewers not having visual impairment it is "real" text that supports this page's semantics. Consequently, it needs to be tagged and have tag correctly placed in the structure tree.
.
The "4" in the square with blue fill. The [Figure] element is applied to the background square and the "box" the 4 is in (box in a box here) lines that delineate the "boxes". The "4" character is identified as an Artifact. Problem stems from how mastered in InD. You want to use a "4 in a box" that is solely a graphic image - no text. The InD configuration of the InD paragraph tag for string "Fresno Medical Center" muddies it a bit as well. A leading space (or tab) for the InD paragraph tag is in the same PDF page "space" as the filled background box and the number.
.
Mentioned before - the "map" holds 36 content items that are inaccessible. After mastering the "map" in the authoring application process it out to a graphics file format (JPG, PNG, whatever). Use this in the maste InD document. Pre-stage as "Figure" with appropriate Atl Txt. Output Tagged PDF of this will not be an issue. As-is, to fix you'd want to slick all tags on the page and rebuild structure tree manual via appropriate manual tagging. Kinda neat but time consuming. Easier to deal with back in "authoring space".
.
.
Print the PDF accessible text and do a close comparison of that content to the PDF's structure tree and PDF page content's flow. Use Acrobat's Accessibility Full Check (both Adobe PDF and Section 508 checks). If Adobe PDF check identifies any issues they need to be fixed.
Typically, the Section 508 check tells you (for item "c") to have a person check for "color" - Accessibe means "color" is not used to convey meaning - if it is you must provide an alternative for users who have color preception related vision issues. (Mentioned use of High Contrast on the PDF - some content goes away here because of the use of white text and use of color backgrounds. What's used makes use of High Contrast problematic/unsucessful.
.
.
If the delivered PDF is to be accessible some re-work in InD is indicated. It is doable. Old unspoken assumptions of "proper" content mastering are often part of the problem. They are predicated on production/delivery of pieces of paper. eDocs such as accessbile PDF are not "paper" and require a broader/deeper perspective.
.
.

This is great. A lot to learn and digest. We do the production for 12 100-page books. The design comes from corporate, so we can't stuck with. But there are a lot of what you've suggested that we can do better.

1. quit taking with an end of line space out of structure tree in pdf
2. set up H1, H2, H3 structure correctly
3. Change map captions to captions. i just read in the ID manual on how to set Map captions in layout.
4. Read your post about a dozen times more.

Your taking the time to help is a true gift. I'll let you know how it's going. And undoubtedly will have more questions as I move forward.
Thanks so much for your kindness.