How we produce our podcasts
Shortly after the first 50 episodes of the Macworld Podcast, I was tasked with producing (though hosting just every other week) these weekly spoken-audio efforts. While I have made adjustments to some of the tools I use, my workflow has changed very little.
Gear and location
We typically record from one of two locations. One is the in-office studio that we’ve dubbed the Podcave. That studio is equipped with four Shure SM58 microphones with pop-filters suspended from Heil HB-1 boom stands, Sony MDR-V6 headphones for each participant, a PreSonus FireStudio Project multichannel audio interface, and an iMac running GarageBand to record everything. The other location is my home studio where I use a vintage AKG 414 microphone along with an Apogee One USB interface plugged into my Mac Pro.
The locations have a measure of isolation and prophylaxis in common. The office studio is well insulated from outside noise, and we have made many efforts to keep people inside the room from adding unwanted noise. For instance, the table-mounted boom stands are padded at the base so that any table-top taps and whaps don’t travel up the stand to be picked up by the mic. Likewise, cables are strapped well out of the way. And the pop filters are placed far enough from the mic’s business end that it’s nearly impossible to be so close to the mic as to overdrive it.
Because my home is in the country, I have little outside noise other than the occasional rooster crow and horse whinny. A Do Not Disturb sign keeps family members out of the area. I use as many fan-less devices as possible. And I have a floor-standing tripod mic stand with the same kind of pop-filter arrangement so I don’t overpower the mic.
Because of my home’s location, in-studio guests are extremely rare. Instead, I record our remote guests over the free Internet voice messaging application Skype. The application I use to do this is Ecamm Network’s $20 Call Recorder. One advantage of Call Recorder is that when installed, it’s found directly within Skype. Just open Skype’s preferences, click the Recording option, and create your recording settings. A small Call Recorder window will now appear whenever you launch Skype. To begin recording both sides of the conversation, just click the Call Recorder’s red Record button.
Once you’ve finished recording, Call Recorder creates a stereo QuickTime movie file: One channel contains your audio track and the other the contents of the Skype track. Call Recorder includes a set of tools for easily splitting and converting the resulting file into separate audio tracks, which you can then drag into GarageBand for editing.
Under ideal conditions we don’t use that Skype recording. Skype can be unreliable—cutting out, dropping calls, or producing unattractive audio artifacts. And even when it behaves itself, in a multiparticipant podcast, you face the issue of having all those voices on a Skype track, where their volumes may be wildly different.
Instead, we ask the participants to record their own part and then upload it to a server where I can retrieve it and incorporate it into my GarageBand project. We do this to get a clearer recording. If a participant doesn’t have the ability to record their side of the conversation, if it’s too much to ask of a particularly busy guest, or if something goes wrong on their end, we cross our fingers and fall back on the Skype track.
One challenge of using guest-recorded tracks is that you have no control over their gear, their microphone’s gain, or the room they’re recording in. The recording levels may look great and sound perfectly acceptable on the Skype track, but when you receive each participant’s recording, the volume may be too hot or too low, there could be background noise, or someone's annoying habit of drumming on their desk, could be picked up by their mic.
At one time I attempted to fix the worst of these problems using GarageBand’s controls and effects plug-ins. But the built-in tools can be a little broad in their abilities, and there’s no command that essentially tells GarageBand, “Make sure that every track’s level is about the same, even when someone whispers or shouts.”
When faced with a challenging mix like this, I leave GarageBand behind and turn to Adobe’s $349 Adobe Audition CS6. Audition is made with audio editing and cleanup in mind rather than music, and as such, it contains a lot of solid tools for making the most of our podcasts.
With one command, I can balance volume among tracks. Removing constant noise is as simple as sampling the offending sound and applying and tweaking a filter. If I have to deal with a Skype track with poor EQ, I have several controls at my fingertips for making the track sound more natural. If someone lacked a pop filter and there are loads of plosives, Audition has a couple of different ways to tackle them. And, unlike GarageBand, Audition includes a ripple-delete feature. Just select a portion of the recording you’d like to delete, impose the ripple delete command, and the problem audio is gone and the content that follows is automatically placed where you marked the cut. GarageBand requires several steps to do the same thing.
Once I’ve cleaned up the audio, I export the mix as a single mono track. This I import into a new GarageBand Podcast project.
Why not just export from Audition and be done with it? The reason is simple: Enhanced podcasts. There are generally two kinds of audio podcasts—audio-only podcasts and enhanced podcasts. What makes the latter “enhanced” is the inclusion of graphics and chapters. (I provide the details on creating an enhanced podcast in How To Create Podcast Chapters.) When you include chapters in your podcast, listeners can easily skip to just the portions of the podcast they wish to listen to. If, for example, you include three topics in your podcast, listeners can easily skip to the one that interests them. For particularly long podcasts, I feel it’s the courteous thing to do.
Regrettably, most audio editing applications are clueless when it comes to enhanced podcasts. GarageBand, however, isn’t. It takes a bit more time to create such a podcast, but I feel it’s time well spent.
With the chapters, graphics, music, and podcast information added, I choose Share > Send Podcast to iTunes and then export the podcast as a 64 kbps mono file, which provides sound good enough for spoken word while maintaining a reasonable file size. Our content management system has a built-in script that takes the podcast, prepares it for streaming and downloading from our site, as well as sends the iTunes Store the information necessary for our podcast subscribers to obtain it there.
With all that done, I’m generally ready to start planning the next week’s podcast.