Podcast Audio 


With so much podcasting these days, a lot of new people are doing digital audio. I don't do podcasts, but once upon a time in Studio City, CA, I was trained as a recording engineer. Here are a few recording hints I wish more podcasters knew. 

Podcast listeners need sound that stays at a near-constant volume. Consider the two most common ways people listen to podcasts: with earbuds (while walking, jogging, shopping, etc.), and in the car. These can be noisy environments, so in order to hear the spoken word intelligibly, we have to crank up the sound loud enough to overcome the noise. Then, if there's a sudden volume spike, it can be deafening--especially with earbuds.

Recently I've encountered this in several different forms:
 
• A debate in which one speaker was considerably stronger and clearer than the other, entirely due to the recording technique--certainly an unfair advantage. I wanted to hear both sides, but I gave up and listened to neither. 
• Telephone interviews, where the person on the phone is much softer than the interviewer in studio. Usually it's the person on the phone we're most interested in, isn't it? And the phone itself reduces intelligibility. So that person's volume should be at least equal to the other's.  
• Podcasts with recorded boilerplate introductions, in which the main talk is much louder than the intro. 
• Live-audience events, in which the audience applause is much, much louder than the speakers. I've heard that twice recently.

That's one set of issues. The other is less common in my experience: distortion due to overdriving the electronics. It's an unpleasant, scratchy, messed-up sound. The solution for this is the same as for the other: use good microphone technique and keep an eye on the meters.

The microphones most podcasters use must be close to the source: within a few inches, ideally. That's my first recording hint. (There's a reason singers hold their mics close to their mouths.) You should be careful that the mic isn't being overwhelmed by the source volume, but that's not at all likely in the case of the spoken word. If there are P's "popping," get a wind screen (a foam covering for the microphone) or move the mic off center so it's not directly in the path of that little wind blast. Two related points are almost too obvious to mention: do your recording in a quiet room, and if you're recording a lecture followed by Q&A, see if you can get the questioners a microphone.

Then (second hint), watch your recording levels. If you can manage it, use a compressor for automatic volume control; or use software with automatic gain control or compression. A compressor keeps the sound confined to a near-constant level. Almost all popular music is heavily compressed, and so are news broadcasts and talk radio. That's why they work for you in the car.

You may not actually need to bother with studying that web link on compression. Chances are, if you go to your local music store, someone there would be very happy to walk you through the purchase and setup of a compressor. They really don't cost that much. The advantage of following the advice on that web page is that it shows you how to obtain and use a free software compressor. (Here's one for the Mac, too.)

If you don't use a compressor, then keep your volume as level as you can by using your meters. Each part of your podcast, including intros, phone calls, applause, and each person speaking--everything but actual silence--should stay in whatever sweet spot your software defines as ideal. That's usually about 2/3 or 3/4 of the way to the top of the meter's range. If your software uses "VU" metering, it should peak regularly between -3dB and 0dB (for the spoken word); if it uses "Peak" metering, or if you're recording music, it may be a little higher, between 0dB and +3dB. If you're not sure between the two, it probably doesn't matter a lot in actual practice. (If you hear distortion on playback, your levels are too hot.) The main thing is, whatever target you set, keep it constant throughout the podcast.

Final hint, and by far the most important: use your ears. I came into recording from a musician's background, and even though I find all the engineering information enormously useful, I never dare forget that the only real test is, "How does it sound?" (Oh, there's one other test: "Am I breaking anything?" If it sounds right and you're not harming your equipment, not much else matters.)

Don't just listen in a pristine, quiet studio environment, though. Music studios have incredible, great, huge studio monitors (speakers) hanging on the walls, and they also have a pair of little "reference monitors" sitting close by on the edge of the mixing console. These monitors' purpose: so the engineer and others can hear the music the way the rest of the world hears it. Sometimes I would cut a cassette tape, walk out to the parking lot, and listen to it in my car.

Podcasters should do something like that, too. Try out your recordings in a moving car, or with earbuds in the local shopping mall. Do this a few times, and you'll get a feel for what you need to do in studio to make it sound good in the real world.

Your listeners will be glad you did.

P.S. Since I don't do podcasting I haven't surveyed the software options and I can't recommend any. How about you--any recommendations? 

Posted: Fri - March 23, 2007 at 09:09 AM           |


© 2004-2007 by Tom Gilson. Permission is granted to quote up to two paragraphs of any blog entry, provided that a link back to the original is included or (in print) the website address is provided. Please email me regarding longer quotes. All other rights reserved.

Weblog Commenting and Trackback by HaloScan.com