Cleaning up Audio for a PodcastJuly 24, 2007 at 3:29 pm | Posted in Audio, Podcasting | Leave a comment
I’ve been involved with audio production as a hobby for quite some time now and enjoy playing with my own musical efforts. So when our company started to do podcasts and I thought the audio quality was poor, I donated some twiddling time to see how to polish them up a bit. I thought I’d post what I came up with here to help others who are doing spoken podcasts.
This process worked well for me doing a home-recorded, single person, spoken podcast. It would require further twidling to handle music, rumbling, or very poor quality files. I used a freemium (free for basic version, pay for premium) program called Audacity. These instructions apply to the free version.
The problems I was attempting to fix were:
- Audio was too soft. It was hard to crank it enough on my laptop. The audio required normalization, which cranks the volume to the maximum level (or whatever level just below maximum you set)
- Drastic variances in volume made the speaker difficult to understand. This is because the dynamic range was too high. Since most podcasts will be listened to in less than ideal environments (through earbuds while outside, through PC speakers), it’s important that the volume be consistent throughout rather than really loud that time you leaned near the mic and then waving in and out the rest of the time. A little compression was called for to take care of that
- Lots of hissing. Cutting some of the high frequencies out was the answer here. It wouldn’t help the cymbals in your favorite rock music sound very better, but for spoken words it’s perfectly fine
- Files were too big. Spoken words require a lot less audio fidelity than music. No one will care if your brilliant thoughts are in mono instead of stereo, for example. Switching them to mono, dropping the bit rate (a lot), and converting to mp3s made almost no difference in quality, but made a huge difference in file size
Here are the step-by-step instructions I wrote down for readying the podcast files for publishing:
- Load file into Audacity
- Ctrl-a (to select entire file)
- Effect, Equalization. Click on RIAA, then Load predefined curve, then ok
- Effect, compressor, ratio 8:1,.threashold -20, uncheck 0db normalization
- Normalize with default settings
- File export as mp3
There is also a onetime setup in Audacity for you to do (Note: This will affect everything else you do with Audacity too):
- Edit, Preferences
- Audio i/o tab: channels 1(mono)
- File format tab:select bit rate 32
This should make your spoken podcasts much easier to hear and understand. I hope this helps!