A friend and I have been putting up a podcast every week or so and she got a question from someone about how to do it.
First, of course, you need to record. I use a Zoom H4N Recorder to record the source audio. My general rule is to record just about as high quality as practical. I record 48k 24bit stereo audio uncompressed — this translates to 24 bits of detail per channel 48,000 times per second. All this gets saved to a WAV file which is the audio equivalent of a raw file from your digital camera.
Recording, regardless of what you’re recording with is to try to get the levels right. In my case I record at 85 on the Zoom. Recording too high will cause clipping while recording too low will just throw away dynamic range.
If you have a choice it’s better to be too low than to be too high. If you have enough dynamic range you can almost get away with murder.
Next up is removing noise. No matter how good a recorder you have you’ll invariably have some background noise unless you’re in an anechoic chamber. Highlight a piece of dead air and capture the noise print:
The beginning and end of your recording are typically a good time for some dead time.
Then I normalize to -0.1dB to get me to a good starting point.
If you’re recording a long segment you’ll have parts that are quieter than others. You can apply a bit of compression to the top-end to get the levels a bit more even.
At this point you’re pretty much done. Figure out what format you want to save it in and you’re good to go!
I also trim off the high end a bit to make it sound less tinny.
You can have my original files that I used to play with this to try you’re hand at processing.
I have samples of various record levels from the Zoom along with a test from my iPhone.
For this example I used Adobe Audition, but you can do the same things with Audacity which is a free download.