Caution
This document is an English translation of the “freee Accessibility Guidelines.” The normative version of this document is in Japanese, and the English version is informational. The English translation is incomplete, and any links with their link texts left in Japanese are untranslated. Please be aware that there may be inaccuracies in the translation or parts that are outdated.
Screen reader stops reading long text in the middle
Last Updated: February 14, 2025
Tags: Screen Reader
Question / Problem
When reading text in NVDA’s browse mode, the screen reader may stop reading in the middle of text that is expected to be read all at once, such as a single line of text or alternative text of an image. Should this behavior be suppressed by implementation?
Answer / Conclusion
No action is required as this is a behavior determined by NVDA settings.
Explanation
In NVDA’s browse mode, when reading text one line at a time using the down arrow key or up arrow key, text longer than a certain number of characters is treated as multi-line text and read out accordingly. Therefore, even text that is expected to be read all at once, such as a single line of text or alternative text of an image, will stop reading in the middle if it exceeds this character limit, and the rest will not be read until the down arrow key is pressed. This is because it may be difficult to understand extremely long text if it is read all at once. Even if it were possible to suppress this behavior by implementation, it should not be done considering the intention of this behavior.
However, for alternative text of images, it is desirable to keep it concise as it is difficult to understand long text due to the lack of structure in the text.
The default value of this setting is 100 characters, and it can be changed in the “Maximum number of characters on one line” setting in the “Browse mode” section of the NVDA settings.
Note that when dividing long text, half-width alphanumeric characters and whitespace seem to be used to determine the division point, so Japanese text containing English words may be divided at a much shorter length than the setting’s character count.