The very first guideline of the Web Content Accessibility Guidelines 1.0 (WCAG 1.0) states that you must provide equivalent alternatives to auditory and visual content. This is Priority 1 requirement and so MUST be done in order to make your web content accessible:

How do you do this?

• Optimise your multimedia for the web
• Provide a text equivalent for every non-text element W3C Techniques for text equivalents, and update the text equivalent whenever the non-text object changes (WCAG Checkpoint 1.1)
• Provide redundant text links for each active region of image maps
• W3C Techniques for server side image maps (WCAG Checkpoint 1.2)
• W3C Techniques for client side image maps (WCAG Checkpoint 1.5)
• Provide an auditory description of the important information of the visual track of a multimedia presentation (WCAG Checkpoint 1.3)
• W3C Techniques for auditory descriptions
• Adding captions to Quicktime (WebAIM)
• Adding captions to Windows Media (WebAIM)
• RealMedia captions (WebAIM)
• Using MAGpie (WebAIM)
• For any time-based multimedia presentation (e.g., a movie or animation), synchronize equivalent alternatives (e.g., captions or auditory descriptions of the visual track) with the presentation (WCAG Checkpoint 1.4)
• W3C Techniques for synchronising equivalent alternatives to multimedia
• Using SMIL from W3C (Synchronised Multimedia Integration Language)

Why? Text equivalents are necessary for users who are deaf, or hard of hearing, or any user who cannot hear sound on their machine (perhaps because of background noise) cannot perceive the information presented through speech, sound effects, music, etc. An audio description of video information is necessary for people who cannot see (or look at) visual content.

Refer to:
W3C Web Content Accessibility Checkpoints 1.3-1.5