Software can convert text to speech, but it cannot convert images to text. We use Alt-text to replace the image for screen readers.
A screenreader will stop on the image and read the Alt-text that you have written. The content of the Alt-text depends on the type/purpose of the image.
For a selection of samples, please visit this page: Samples of Alt text.
Alt-text is done differently for Children's Picture Books and Comics. Please visit the pages Alt-Text for Picture Books and Comic Book/Graphic Novel Description: Introduction for more information.
The W3C defines alternative text as follows:
A text alternative is text that is used in place of non-text content for those who cannot view the non-text content. Non-text content includes such things as pictures, charts, applets, audio files, etc. People who cannot see for example would not be able to see information presented in a picture or chart. A text alternative is therefore provided that allows the user to be able to convert the information (the text) into speech. In the future, having the information in text also makes it possible to translate the information into sign language, into pictures, or into a simpler form of writing.
In order for people with disabilities to be able to use this text - the text must be "programmatically determinable." This means that the text must be able to be read and used by the assistive technologies (and the accessibility features in browsers) that people with disabilities use.
It must also be possible for people using assistive technologies to find these text alternatives when they encounter non-text content that they cannot use. To accomplish this, we say that the text must be "programmatically associated" with the non-text content. This means that the user must be able to use their assistive technology to find the alternative text (that they can use) when they land on the non-text content (that they can't use).
The below images can be removed from the text. This is suggested because none of these images convey meaning (i.e. they're decorative). Similar situations may arise. If unsure, ask the Production Coordinator.
During conversion from ebook to .doc file, the image quality can be effected. Because of this we remove and replace all images in a document.
First you need to extract the images from the original ebook.
Once you have your new images files, simply replace all the files in the .doc with these new images.
If a caption exists in the original ebook, then you must insert it for the corresponding image in the .doc file. See Captions for instructions.
Select the image and then select
clear formatting. If there was a little dot to the left of the image, then that should now be gone.
For large images, resize them so they fit on the page with their caption, and surrounding text. See How to Resize Images in Word for how to do this.
This is the same thing as breaking/removing links.
Maintain the general location of images in relation to surrounding text, but ensure that the image (and caption) are within the text (that is, the text does not flow around the image) and between paragraphs (not breaking up the middle of a sentence or paragraph).
The wrapping style should be set as in-line with text (right click the image, select
Wrap Text >
In Line with Text). Software can only detect and properly read the Alt text associated with an image when it is placed 'in line'.
To add Alt-text to an image:
descriptionfield (leave the title field blank)
The main guidelines are as follows:
Alt-text is generally 120 word maximum. For longer descriptions of complex images, you will have to put the longer description in the Complex Image Description section at the back of the book. See below for more details.
Write Alt-text so it flows with the surrounding text as not to be jarring to reader.
imagein your description. The screen reader will already pick up that it is an image, so writing the word "image" will be repetitive and redundant.
If the image type is consistent throughout a book, provide this information at the beginning, and then don’t repeat in the following images. It is only necessary to provide this information in the alt-text of the first image. For example, the first image alt-text would read: "A black and white photograph of an Indian girl looking out a window on a train." and all following images will only have the image description without the phrase "A black and white photograph" (i.e. "An Indian girl stands on the platform of a train station with her luggage on the ground beside her. She waves with a smile on her face.")
Be clear and concise in your writing.
Read the text around the image, if there is a person, place, or object that is named in the text, then use it in your description. Do not name characters in images before they are introduced in the text.
You may find that the image is already described in the text. You don't have to describe it again in full, but add some details that are not given to describe the image avoiding repetition.
Only write about what you see and what you know. Alt text should always provide a direct representation of the image and what it conveys visually. Try and think about the audience, reading level, and tone. Use appropriate vocabulary for the grade level and subject matter.
When working with a novel, try to warm it up a little bit, because your descriptions need to help drive the narrative forward. Image descriptions are really an art, not a science.
For image description guidelines and examples, refer to:
We create longer, complex descriptions for images of maps, graphs, and charts. We can also create longer, complex descriptions when we need to transcribe an image, such as an image of a letter. If you are not sure if your image requires a complex description, ask in the Q&A section.
For images that require more complex or extended descriptions we create two descriptions:
Sometimes you will come across images that are Tables, Surveys, or Images with Text, or Images of Words or Phrases. It is a big publisher no-no to put any form of text within an image, and there are a few ways we can deal with this as it comes up.
If there is an image of a table, simply recreate the table within Word following the directions in the Tables Section. Then delete the image of the table once you are done.
This one is a bit of a judgement call. It depends on the context of the image.
If it’s critical to understanding the book that readers know exactly what was asked in the survey then we want to transcribe the entire survey and then add a Producer's Note at the beginning of the book to explain what we have done.
For blank lines you can replace them with
If it doesn’t really matter to the reading experience what the details of the survey are then Alt-Text will do. There are some great examples on how to write Alt-Text here.
Some Images will have text as part of that image. In these cases there are two solutions:
Again, this is a judgement call. Both options can work, but it also depends on what the text is. When in doubt just ask!
A very big publisher no-no is having Single Words or Phrases as Images. This is one of the biggest accessibility sins. When this happens with must transcribe the image inline with the text, and then remove the image itself. If the image is written in another alphabet you may have to either use Unicode or set up your keyboard to type that alphabet. Please see the Languages section for more information.
In rare cases, you may be describing the image without a copy of the original image in your document. This is often the case when digitizing from print.
You still need to describe the image, but also denote that the text is not part of the regular paragraph text.
Typically, you may have something similar to the following:
BEGIN IMAGE DESCRIPTION Sally picks a completely black kettle off the floor. END IMAGE DESCRIPTION
The specific words you use will differ depending on the language used in the book (e.g. figure, illustration) and whether it is a description (you wrote) or a caption.
Especially in children's books, you may want to have the images as a background with the text on top.
In Line with Textunder
Wrap Text. (You can always Undo to change it back.)
Check out these excellent tutorials on writing alternative text descriptions:
Q: Hello! I have a question about an image in the book I'm working on, "Seeking Allah, Finding Jesus". In the Prologue, every time the author writes "Muhammad", he inserts a small image. The footnote explains that the image is a symbol representing an Arabic phrase: "peace and blessings of Allah be upon him". I'm thinking of handling this by deleting the image of the symbol and replacing it with the phrase in brackets. What do you think? If I do that, should I insert a prod note explaining what I did?
A: In this case you would treat it as a regular image and insert the Alt-Text: symbol representing an Arabic phrase: "peace and blessings of Allah be upon him". You would then put a Producers Note at the beginning of the book to explain why you did this. For more info see here.
Q: I'm wondering if there is any way for me to check whether the images I've inserted into a book are the correct format (eg. JPG or PNG). They all look the same to me and I can't find any way to identify the file type within the Word doc. I follow the procedure for inserting images outlined in the wiki so it usually doesn't matter but in very long books, I sometimes miss a few but I can't double check if I missed any because I can't tell what file format they are.
A: Unfortunately, there is no easy way to find the image types within the word document. You can save your edited word document as a webpage, which will create a folder with all the extracted images. This file will show you all the image format types within your document, but it wont necessarily show you where they are. You would have to go through the images in the folder and try to match the images to the document, which can be time consuming.
Another tip is to look at your workflow and try to see where you are making an error in your workflow to try and avoid it from happening again.
There are two issues that can be causing this to occur, and they both have slightly different solutions: one has to do with workflow of extracting images, and one has to do with the workflow of replacing images.
Extracting Images: The only method of extracting images that could cause an incorrect file format to occur is with screenshots. This is easily fixed by just double checking with each screenshot you take that you are saving the file in the correct format (gif, jpeg, png, or svg.) Again, this is only if you are using screenshots, if you are not using screenshots then you should address your workflow.
Replacing Images: You mentioned this only happens with larger books, which is understandable since there are more images and more room for errors. I do not know what your workflow is for replacing images in the text, but it is a good idea to evaluate your current workflow and see if there is any way you can improve it to help avoid future errors. If you want more tips on how to do this just ask.
I would not stress over this; if it is only happening with large books, and after evaluating your workflow you find that you are still having some images slip through that is okay. The best way to identify exactly where they are only occurs after the Production Coordinator converts the book at tests the new file.
Q: I'm trying to keep in mind the aim for clarity and focus on relevant information, but I'm struggling with this. One of the first maps in this book is a map of the North-West area of Canada with a lot of lakes, rivers, and important villages labeled. Throughout the book, there are a few other maps which "zoom in" on specific areas which have already been described in this larger map, perhaps with some added details. Do I need to re-describe the area exactly? Or how should I proceed with these cases? For example, I've included the large map of The North-West and a map of Red River. Lots of the info in the Red River map are already present/described in the North-West map.
A: Maps are challenging for anyone. If you haven't already checked it out, you should go look at the Alt-text samples we have on the wiki (there are two map examples that could help.) To answer your questions: I would not repeat what has already been described. You can start the zoomed in descriptions with something along the lines of 'This map is a closer image of the Red River area…' you could include some sweeping description like "it includes the areas between these rivers" or "it is bordered by these rivers" etc. Then you can add more detail into. Remember that starting big and going small can help in these cases. I also recommend opening a new document to work on the descriptions, so you can easily edit and rewrite as you go. It is good to remember what the purpose of the map is, and to base your description on that purpose. In this case it looks like the map is meant to show the layout of the land, so you should describe it in a way the listener can get a sense of the layout. Where are the rivers and important points? Where to do they lead to. You can start big (this is a map of this area that features rivers and this lake) then start at one point and work your way through the map like you are traveling through it. Reading it aloud will also help.
Q: One more question from His Needs, Her Needs. At the very end of the book, there are 5 full-page images that are advertisements for other books by the author, for his website, and for his publisher. They aren't decorative but I'm not sure how important it is to keep them (and therefore add alt-text to describe their content). Would you recommend deleting them or keeping them?
A: You can delete these images.
Q: In Appendix B of His Needs, Her Needs, there is a 10 page questionnaire but each page is an image file that contains text, not actually text. Am I right that I should transcribe the text from each image and then delete the image files?
A: In this instance the context is critical to understanding the book, and should be transcribed, the images should be removed, and a Producer's Note should be places at the beginning of the book explaining this has been done. I took a peek at some other parts of the book, and there seems to be other surveys as well. When it comes to these sorts of images it depends again on context, and is a judgement call. If the details of the survey is critical to the understanding of the book, then we do the transcription and Prod Not as mentioned above; if the survey is not critical to the understanding the book then simple Alt-Text is enough (remember to keep the Alt-text brief and concise, and you can refer to our Alt text examples for some inspiration if you need it.) There is a Question and Answer a bit further down on this page that addresses this as well for your reference. You can also see the section on Image and Surveys for more information.
Q: Another caption question! The book I'm working on has a painting at the beginning of each story that goes with the theme of the story. I have extracted the images of the paintings from the epub file but they each have the painting's title and year of creation incorporated into the image as a caption. I know that a caption in the image isn't good enough because it won't be read by TTS software and will be hard to see for people with limited vision. But is it okay that there will be a second caption when I add mine or do I need to figure out how to crop the image to get rid of the built-in caption?
A: Great question! There are two options for this; (1) you could simply transcribe the text from the image at the end of the caption that already exits, or (2) you could transcribe the text from the image into the Alt-text.
Q: What do I do when a single caption refers to and describes two different images? Is it okay to give both images the same caption? ie. copy and paste the caption and apply it once to the first image and then again to the second image?
A: You only have to apply the caption to one of the photos. With the Alt-text and the placing of the images the reader will be able to figure out what is going on.
A: It depends on the context of the image. If it’s critical to understanding the book that readers know exactly what was asked in the survey then we might want to translate the entire survey content in a prodnote; otherwise, if it doesn’t really matter to the reading experience what the details of the survey are then alt-text will do. It’s a bit of a judgement call.
Q: If there is a photograph of a letter, should we transcribe it into the alt-text?
A: It depends on two factors that should be taken into consideration when writing alternative text (requires some judgement):
Q: For images that contain a lot of information (such as an Army record that contains a lot of text, finger prints, tables, etc.), what is recommended in terms of Alt Text?
A: Very good question! Image descriptions can be tricky. The first question I like to consider is the purpose of the image: is it decorative? for visual interest only? provide essential information? We also consider if it's adequately described in the surrounding content already.
Let's take "Shaw's US Army Record" image from Rich man, poor man as an example (below). There's a lot of text embedded within this image.
What's the purpose? This image is located at the end of the book (in the biography section) and a brief caption is provided. I haven't read the book, but I think the general purpose of the image is to give the reader a sense of what his Army record looked like, so the alternative text could describe some essential things that we can glean from the record, such as: this is his "Military Record and Report of Separation, Certificate of Service" and it contains both personal data and information about his military history.
What do I see? How much information we want to provide about his personal and military history depends on: 1) what we can actually see! (the image quality is poor, so most likely the author doesn't think it's very important that we see or know the details), and 2) what information might readers want to know that's here and isn't already in the book.
To be honest, I can't make out much information here, so I think it would be perfectly appropriate to leave out specific personal and military history details from the Alternative Text description.
For further exploration: Diagram Center is the authority on image descriptions and they have great guidelines and training material: Image Description Training