How radio is made

How radio is made

About a month ago, I received a comment on my blog from Fred Sandsmark, a bandmate of mine at the Castro Valley Community Band, which is where I normally play clarinet during the school year once a week.

Fred wrote, in response to my June 1 blog entry on my latest piece for PRI’s The World:

I heard this report while driving (I don’t typically hear The World) and found it really interesting. Listening to it made me wonder about how you put such a report together — how you get your assignments (or make your pitches), how tight your deadlines are, how you find your interview subjects, if you favor in-person versus phone interviews (or if that’s just a function of where your source is), how much editing you have to do vs. how much (if any) is done at PRI, what technologies you use to collect and assemble a piece, etc. I know you’re probably too busy actually doing the work to write a blog post on this, but it might be interesting for those of us not in the radio biz to learn about.

So Fred, here you go!

First off, let me explain a little bit about what we mean when we talk about public radio, in the context of the United States. Most public radio listeners are familiar with the largest and most well-known public radio network in the country: National Public Radio. But there are others, including Public Radio International and American Public Media. There’s also Pacifica Radio and smaller, more local stations, like UC Berkeley’s KALX.

When you turn on your radio (in my case, KQED 88.5 FM in San Francisco) and are “listening to NPR,” some of the time you may actually be listening to NPR content. Basically anytime you hear anyone say: “From NPR News in Washington, I’m Carl Kassel,” or, “From NPR in Washington, I’m Neal Conan,” or mention NPR in any way, then that thing that you’re listening to — Morning Edition, Talk of the Nation, All Things Considered, On The Media — is produced by NPR and/or distributed on its member stations nationwide.

However, sometimes you might be listening to a show and they’ll say something like: “From Public Radio International, this is The World.” or perhaps “From American Public Media, this is Marketplace.”

That identifier is useful because it tells who which network produced that particular show. Most listeners brush over this and consider everything they hear on their local NPR station to be “on NPR.” KQED (and most other NPR member stations) produce relatively little local and original compared with the vast amounts of stuff that they license from NPR, PRI, and APM. Of the three, NPR produces the most amount of content — so it’s easy to understand why people confuse NPR with any public radio.

If your local public radio station is like mine (KQED), NPR breaks in at the top of every hour for a five-minute newscast, basically 24 hours a day — “From NPR News in Washington, I’m Korva Coleman.” So the listener is inundated with this notion of NPR 24 hours a day, even if what’s coming after that newscast is PRI’s The World, or APM’s Marketplace, or PRI’s This American Life, or APM’s A Prarie Home Companion. Yep, those last two — the most popular two shows on public radio — are not produced by NPR at all.

NPR has done a great job in the minds of American radio listeners of branding themselves in the minds of their listeners. Or perhaps more accurately, PRI/APM have done a bad job. NPR is significantly larger as an organization than other of those — so every time I’m on the radio (likely, on The World), my friends will email and say: “Did I just hear you on NPR?” and I’ll say: “If you mean PRI’s The World, then yes.” 🙂

Ok, now, the particular piece that Fred is referencing was an assignment that I pitched to The World directly.

Obviously, the best interviews are done in person — phone interviews are always a last resort, or what we use if we’re short on time and long on distance. I use Skype for all my radio-related calls and Audio Hijack Pro to record the sound, which conveniently splits it up into two tracks. Sometimes, many of my interview subjects are tech-savvy and have Skype installed themselves, which can sound quite good.

If an in-person is not possible, and if I/my interview subject have time and/or if my bosses have the budget for it, we’ll get a interview subject into a radio studio, where their voice can be recorded at very high quality and transmitted over an ISDN line. Ever listen to Fresh Air? Nearly all of those interviews are recorded with the subject sitting in another studio, while Terry Gross is in Philadelphia. And yet, they sound like they’re both sitting next to one another. That’s ISDN.

But booking studio time can require more advanced planning, and is costly — you have to pay the board operator on the other end, too. More often, in my case, what happens is that I’ll book a “tape synch.” That’s when I send someone out to my interview subject to stick a mic in their face, while I do the interview by phone. It’s like a phone interview, except now I have a good quality recording of that person’s voice. My syncher/stringer will then send me the audio via FTP and I can drop it into my audio editing software. (I often do tape synchs for other radio shows, too. I just did one this week at Yahoo for PRI’s Living on Earth. I usually find the assignments through AIR, the Association of Independents in Radio, of which I am a paying member.)

Back in the day, when radio reporters used reels of tape, or even cassette tapes, it was important for them to be “synched.” That is to say, important for them to start/stop at the same time so that later editing would be much easier. Now that the vast majority of radio producing/editing is done entirely digitally, that doesn’t matter so much — but we still keep the old jargon.

Typically, I’m not handed an assignment by my editors — I’d say it happens maybe 25 percent of the time. (But it sure makes my life easier!) More often, I have to find a story on my own that I pitch to them. I spend many hours “working,” which really means scouring the Internet for possible stories. Pitches with places that I work with on a regular basis (like The World) typically aren’t long, perhaps a few sentences at most. It describes what the story is, why it’s news, and who I’d interview and (ideally) what other relevant audio can be recorded to make the piece more sound-rich.

Once they accept it, I make the arrangements for the interviews and do them. Since my days at Columbia, I’ve been recording on a Marantz PMD660. This is a popular digital recorder in the radio industry. Alternatives include the Zoom H4, the Marantz PMD620 and many others. The mic I was taught on was an omnidirectional Electrovoice RE50B. As the name implies, it records sound equally in all directions. However, for the last couple years, at the recommendation of my mentor and colleague Clark Boyd, I’ve been using an AudioTechnica 813a cardiod mic.

Cardiod mics record in an elliptical shape in the direction that the microphone is pointed in and are great as a handheld microphone for face-to-face interviews. They’re more directional than an omni mic, but not as focused as shotgun mics. I’ve met radio reporters who use shotgun mics for everything — I have the Sennheiser K6 with ME-66 short shot on long-term loan from The World, but honestly, I’ve only ever needed to use it once. For my purposes, the AT813a is exactly what I need.

For television/film production, or even for many radio documentaries that are being recorded in a more controlled environment, mic’ing is often done with a mic on a fishpole, but I’ve never done that.

I listen on these “cans” (that’s radio jargon for headphones), a pair of well-worn Sony MDR-7506s.

Now, when I get home with a bunch of “tape” recorded onto my compact flash card, I have to pull it all into my audio editing software. I use ProTools, an industry standard. Alternatives include Adobe Audition.

You can sorta see in this photo of what my workspace looks like, including my computer screen. Basically I have a bunch of waveforms that need to be cut and edited down into 20-30 sec chunks. Based on those chunks, I write a script that looks basically like this:

LISA MULLINS: Voters in Iran go to the poles this month to choose their next president. Incumbent Mahmoud Ahmadinejad will square off against three challengers. One is a fellow conservative, the other two are moderates. The four candidates have their differences, but their campaigns have at least one thing in common, they all have their own websites, Facebook pages, and FriendFeed and Twitter accounts. Reporter Cyrus Farivar [suh-ROOS FAR-ih-var] takes us online to follow the Iranian election.

CF TX 1: Iran has more Internet users than any other Middle Eastern country. So it’s not surprising that Iran has been something of a crucible for political speech online. Back in 2003, the Islamic Republic
became the first country in the world to arrest a blogger for his writings. [18 SEC]

SK AX 1: The battle for civil society is partly being waged in the cyberspace. [5 SEC]

CF TX 2: That’s Shahram Kholdi, an Iranian political science doctoral student at the University of Manchester in Britain. He points to the fact that in just the past week or so, Iranian authorities blocked, and then reinstated access to Facebook. [12 SEC]

and so forth.

In radio parlance, TX (“tracks”), refers to voice tracks, or the part of the piece where the reporter is speaking. AX (“actualities”) refers to anything an interviewee says or does. Sometimes there’s also AMB (“ambient sound”), which refers to any ambient sound that gets played under an TX/AX.

When I’m done writing that, I send it to my editor, along with the audio files of the AX. They edit the script, tell me what needs to be changed. Then, I record the TX in my closet. Yes, my closet. It’s the closest thing to a studio that I have, and it would be far too inconvenient/expensive to actually get me into a real studio each time I need to record something.

Then, I send the TX to my editor, and the show’s engineers mix it all together.

Yay radio!


  1. Fascinating! Do you have a general ratio of editing time vs. final product? In terms of initial editing (listening to the whole interview again?), re-edits, script changes and tweaks, how long does it take you to make a 10 minute piece?

    Also, when you’re interviewing people who aren’t used to speaking for radio, does it take some work to establish good rapport and get them comfortable with the mic?

    Thanks for this peek into your work.

  2. Thanks for this – very interesting – I wondered how it all fit together!

    BTW, I found a small case for a piece of your audio equipment under the passenger seat in my car. Black, zippered. Will hold onto it until I see you next.

  3. Cool synopsis! Thanks. I’m bookmarking this one.

    BTW: I got a Zoom H2 for recording my concerts and I love it. Used it for recording a live event and tried syncing it to video — PITA! But perhaps because it was DV NTSC video against 48K audio. Can’t say I wasn’t warned.

  4. Ray,

    My pieces are generally in the three to four minute range, and I would say, story depending, it takes roughly 3-4 hrs (+/-) to produce that. That’s arranging interviews, interviews, writing, editing, and tracking.

    I’ve never had anyone who was uncomfortable around a microphone — except, and I kid you not, Chris Anderson (Editor, Wired magazine). He absolutely insisted that he hold the mic to his own face instead of me when I interviewed him once. Not really sure why.

  5. Thanks a bunch, Cyrus. It’s a great insight into the hours of stuff that’s pumped into the airwaves and onto the Internet every day. (Good, also, to read the PRI/APM/NPR discussion – I never thought of it in terms of good and bad branding.)

    I would imagine, too, that sometimes the interview subjects you want aren’t the ones you can get, or the ones you get don’t speak in a way that’s usable (lots of ums and ers and y’knows and rambling, or yes and no answers without context), which requires you to think on your feet in an interview to rephrase a question. Like print in that regard, where one is looking for that perfect quote.

    It all seems like a lot of work for something that’s so ephemeral – and that sounds so effortless on the air.

    Again, thanks for putting that together. Really interesting.

    Fred Sandsmark
  6. > Basically anytime you hear anyone say: “From NPR News in Washington, I’m Carl Kassel,” or, “From NPR in Washington, I’m Neal Conan,” or mention NPR in any way, then that thing that you’re listening to — Morning Edition, Talk of the Nation, All Things Considered, On The Media — is produced by NPR and distributed on its member stations nationwide.

    Correction: On the Media (like Car Talk and Living on Earth) is distributed but not produced by NPR. It’s produced by WNYC.

    I’m curious about your description of the editing process you undergo for The World. Do your editors there generally do a paper edit instead of a phone edit initially? I’ve always been taught that hearing a piece the first time around (like a listener would) is the best way to catch any potential problems. That’s how my editors at NPR have always operated, and I think that’s what I did the one time I filed for The World.

    (fellow AIR member)

    Scott Gurian
  7. Cyrus, I loved this blog entry. I’ve always been fascinated with public radio journalism. I’ve long wondered what the technology is that radio people use.

    I have a couple more questions.

    1. Do you use the microphone plugged into the small recorder for field interviews?

    2. What method do you use in your closet? (Love this idea..)

    3. Any advice for someone interested in trying a couple stories?

    Thanks again for the great info!

    (I’m friends with Alan Wiig. He told me all about your charmed life in France last year!)

    Troy Carle
  8. Hey Cyrus – very interesting! I see that you are pretty microphone savvy – would you have a suggestion for an all around digital voice recorder for anthropological fieldwork interviews? I don’t need them to be of production quality, but I found that in previous years my recordings have had way too much background noise, or the voices were not clear. This is key when I have trouble with the languages most of the people are using in the first place! I also plan on using the voice transcription software for my own voice. Thanks!

    Sara Thiam
  9. Troy: As I stated above, I use the AudioTechnica 813a plugged into the Marantz to do the interview. I’ve never used the Marantz’ built-in mic.

    What method do I use in the closet? I just close the doors/windows and make sure that everything is as quiet as possible. I turn on the light and hold the script up in front of my face with one hand and hold the mic with the other. It’s pretty easy.

    As far as trying public-radio style stories? Yeah, get some gear and find a story! I’d first play around with recording some people. Interview your friends so you get a feel for mic placement and the like. Then, once you’ve found a story you’d like to pitch, maybe try to think of what program it would be well-suited for. Do you have any in mind?

    Sara, you might try the Zoom H4 ( I know a lot of people are happy with that one. Maybe go into an audio shop and see what they suggest?

  10. I am working on my own radio project, and I have a few questions.

    1. I want to do some of the interviews over the phone (landline) which recorder do you recommend?

    2. After I record the phone interview, which software should I use to download it onto my computer? I need something very user friendly, I’m a content person not an editor!

    3. I’m so afraid of my PC crashing if I download software, do you recommend an external drive to download software, and use it for the editing?

    4. What do you think about the acoustica software?

    Thanks I really enjoyed reading your blog. Can you recommend some websites that have solid information on how to record etc?

  11. 1. I’m on a Mac, so I use Skype to call phones and then Audio Hijack Pro to record the calls.

    2. I’m not sure what you mean here. I’d just record the call with Skype.

    3. I’ve heard people of using external drives, but I don’t really understand why an external drive would be less prone to crashing than a regular drive.

    4. Never heard of Acoustica. Check out this site:

Comments are closed.