Mushi
Member
From: USA
Registered: 2010-07-06
Posts: 252
I was wondering if anyone else has had experience speeding up their Japanese audio files in order to get through them more quickly and efficiently.
I originally stumbled upon this when I accidentally hit the 2x button on my video player, and found that I actually preferred watching some videos that way. The one major problem though was that the audio sounded like it was spoken by chipmunks, and that shift in pitch made the audio harder to understand.
But recently, I came across the atempo filter in ffmpeg, and now I've started using it to convert my audio files to 1.5x to 2x their original speed. It uses some kind of correlation algorithm to preserve the original pitch while increasing the speed. It can also slow down audio if you wish.
I find it quite amazing how this allows me to listen to twice as much dialogue in a given period, and yet the dialogue sound pretty much perfectly normal - just spoken more quickly. I feel like this technique has boosted my productivity. Has anyone else tried this?
Mushi
Member
From: USA
Registered: 2010-07-06
Posts: 252
Video players vary in how good they are at "scaling" audio, but ffmpeg seems to give better results than they do. Probably that's because it's a conversion program rather than a real-time player, so it can do whatever processing it needs to do at its own pace to create the output file.
I remember hearing somewhere that people read about 50 wpm or so faster than they speak. And input is easier than output, so it does make sense that the pace of audio materials would tax our patience, and maybe even make us tune out.
On the other hand, I find that the faster audio holds your attention better, and also tends to reveal patterns you'd otherwise miss, such as certain words and phrases that are repeated. You tend to catch those things better when they occur 30 seconds apart rather than 60 seconds apart! It ends up increasing comprehension in many cases.
