Skip to main content


As a screen reader serving the blind & vision impaired, our users rely on audio rather than visuals, therefore this week we bring you an audio testimonial from Jacques in South Africa. Jacques is a technical specialist for a financial organisation.

“When dealing with inaccessible software, NVDA’s object navigation often saves the day…NV Access & add on contributors around the globe, your efforts are appreciated” - Jacques

Thank you Jacques for sharing how NVDA helps you!

#NVDA #Testimonial

in reply to NV Access

For anyone who uses captions, on the YouTube version of this video, we tried out YouTube's auto-captions today, rather than writing them out manually. We'd be interested in how you find them and any feedback either way (It's certainly faster, although I suspect not quite as accurate?) YouTube link: youtu.be/fPbZ1sA69oo
in reply to NV Access

I don't know how good they are but in the deaf community they've been called "autocraptions"
in reply to Kara Goldfinch

@KaraLG84 That doesn't sound promising! :D Ok I'll see if we get any other feedback. Might need to go back to manually writing them up next week. I just thought I'd try it since I used them for the Spanish one last time (Had to for that as I don't speak Spanish myself)
in reply to NV Access

@KaraLG84 It is still my experience that manual captions are still far better than anything automatically generated.
in reply to NV Access

Have you tried Whisper? It’s super accurate, especially with the larger models.
in reply to Mikołaj Hołysz

@miki Previously we'd written the captions manually. Since we needed to use the auto captions last time (because I don't speak Spanish), I thought we'd try it on an English language video this week and see what people thought.
in reply to NV Access

A really good solution for autocaptions is Whisper, a free and open source solution from Open AI.
in reply to NV Access

I recommend downloading Buzz, which allows you to transcribe audio files and export to caption files. So long as your video's audio matches the time stamps of the video file, you'll get far more accurate results.
in reply to NV Access

I use NVDA at work for testing accessible web sites, keep up the great work!
in reply to Paul Houle

@UP8 Thanks Paul! We're glad you find it helpful! Please do share with others.
Unknown parent

NV Access
@rats @KaraLG84 Thanks! I actually usually make them myself so I know they are accurate - I needed to do it for a video the previous week which was in Spanish (and I don't speak Spanish) - so having figured out how to do it for that, I thought I'd try it for an English one. Lots of feedback on the experience, so I think I'll go back to hand crafted captions this week :) Thank you for taking the time to articulate those issues!
Unknown parent

NV Access
this is the auto generated captions on the video you linked, for reference
@rats @KaraLG84 Listening to that, it actually doesn't sound too far off, but definitely not 100% (Envy access instead of NV Access, etc)