PLEASE NOTE:
You are using an outdated browser that is no longer suitable for modern web standards.
Please update (or change) your browser to view our site as it is intended to be seen. Thank you.
poppy field

Foxvideo Productions



Having spent the best part of 40 years looking through a viewfinder of a stills or video camera, sitting in front of a computer editing photos or videos, producing websites and managing content and social media accounts for organisations or individuals, it’s a very strange feeling to now say “I’m retired".

Despite no longer taking on commercial paid work, it’s virtually impossible to suddenly stop what you’ve been doing for the best part of your working life so, for the past few years I’ve been taking on work on a ‘pro-bono’ basis for church, charity or voluntary organisations. I hope that my experience, skills and equipment can now be put to good use by those who perhaps, don’t have the budget or funding to produce what they want or need for their media requirements.

Composite image

Audio to Text UPDATE

Filed in: Audio |Post production |Transcribe written by: Dave Farrants
cassette tape graphic

Browsing a pro video forum recently, I spotted an mention of ‘MacWhisper’, this is an AI based App to transcribe audio to text, similar to the Apps and online sites I tried a week ago to transcribe a badly recorded lecture on an audio cassette.

MacWhisper comes with 2 versions, a free version that included the ‘Tiny & Base’ models or a Pro version for $10 which describes itself as “industry leading transcription quality’. It also states it'll work in 100 supported languages and of importance to some, "no data leaves your machine".

I downloaded and paid for the pro version, it can import audio from mp3, wav, m4a, mp4 or a microphone (I also dropped a .mov in and it worked with that), I tried it with my lecture audio, this was a 50 minute mp3 clip and I stopped it after around 15 minutes. This App is not fast on my 2020 iMac, it’s recommended you use an M1 or M2 Mac, but speed is not important to me, I just wanted to judge the accuracy of the output. The result was about the same as the Apps and Online sites I’d tried previously, around 80-85% accurate, bearing in mind the low quality audio I was trying it with.

I then tried some high quality audio from a Peter Wells video on YouTube, this had some complex words and phrases, MacWhisper got it 99% right, this is from the MacWhisper transcript:

“Both hemispheres of our brain are equally vital for a balanced mind. Thus, in the debit-credit balance sense, the ancients responded to good by sacrificing something valuable. Mental equilibrium was achieved by offering their best lungs (should read LAMBS!) to Yahweh. A sacrificial offering fulfilled an old covenant between the people and Yahweh.”

MacWhisper offers output to plain text, .csv with timecode or to .vtt or .srt for import as subtitles into Davinci Resolve, which I tried with success.

This App is still in development with updates, so for $10 (plus Tax) this is worthwhile addition to anyone in video and a far cheaper option to some of the online offerings.

With AI this clever (have you tried ChatGPT?), how long before SkyNet becomes a reality…..
Back

Tags: Audio,Post production,Transcribe

< Previous Post   Next Post >

©2026 Foxvideo Productions, Dymock. Gloucestershire, GL18 2AU