A top audio engineer explains NPR’s signature sound

Print More

Will Fisher/Wikimedia Commons

What makes NPR’s broadcast sound so crisp and bright, while many local public radio stations have a bassier, boomier tone? Adam Ragusea, host of our podcast The Pub, wanted to find out, so he recently interviewed the guy who should know best — Shawn Fox, senior director of audio engineering at NPR. This is an edited transcript of the interview, which first appeared on The Pub #17.

Current: How did you come up with what’s been described as NPR’s “secret sauce”?

Fox: Oh, the secret sauce! When we were in our very early design phases of our new building, that was one of the things that came up very quickly — making sure that we continue on with the NPR sound. There was a whole detailed analysis of what was that sound? In terms of the phenomenon of being able to hear the difference between local and national audio quality, there are lots of reasons for that. All in all, what our local stations do and what we do at NPR are very close when you compare it to some of our commercial brethren. But yes, you can hear that difference.

Current: What is ingredient number one of the secret sauce? Is it the NPR studio microphone?

Fox: The NPR sound has so many tentacles. If we’re just focusing on the studio side, which was actually the easiest thing, it all starts with the microphone. We use a simple Neumann U87 microphone as the house-standard microphone at all of our facilities. They’re expensive, but that’s what we’ve used for years.

In the new building, we knew we had the old microphones — and microphones don’t die unless somebody really works hard at it — and we had more facilities, so we bought a few more. But it really comes down to the U87 with the bass rolled off.

Current: The bass rolled off? Could you explain what that means?

Fox: Sure. The U87 and most higher-end microphones have two switches on the back. One is a polar pattern, which is the direction of the microphone, and the other one is for the bass roll-off. When the bass is rolled off, you can’t hear the lower frequencies of my voice. The microphone itself takes them away.

And when I flip that switch, you get those lower frequencies. This is what they call flat mode; there is no attenuation.

Current: It takes out frequencies below how many hertz? We’re talking really super-low stuff, like below 150 or 200 hertz or so, right?

Fox: Real low stuff, somewhere below 250 or so.

Current: Yes, but if you’re at a station and kind of frustrated with the bassy, boomy sound you get on your studio mic, and you can’t get the engineering staff to do anything about it for you, one thing that you could do yourself is to look on the mic and see if there is a bass roll-off switch, and turn it on.

Fox: Yeah. I wouldn’t try to focus on trying to fix the studio, because there’s a whole different element of the acoustics with the studio. The reason NPR came to this standard — and this was decades ago — was because most of our listeners are consuming in an automobile or with something else in the background. Back in the day, and even to some degree now, you roll down those windows and hear those low rumbling frequencies. We wanted our voices to get above that so that they could be clear, open and understandable to improve our storytelling.

We came to that conclusion mostly because most of our consumers were listening to Morning Edition and All Things Considered in the automobile to and from work. And now, as more of our content is heard on headphones from iPhones and all the digital sides of that, we discovered that continuing with this is beneficial, because there is still that acoustic outside noise. But it’s not a fix for a studio. That’s the whole other element of why our sound is a little different. We want to make sure that we work with our on-air talent, our reporters and our hosts.

We are fans of being close-miked, and P-pops come into play there. But we make sure that we are within a foot of the microphone and usually a lot closer — close to six inches — in working with any of our on-air talent. That’s another element that goes into it.

Current: The thing about getting closer to the mic is proximity effect, right? The closer you are, the more bassy you are. So why doesn’t that happen to your people?

Fox: That’s where the bass roll-off also comes into play, and we also really work with them so that it’s not head-on directly into the microphone. It’s more to the side.

Current: If you imagine the air coming out of your mouth as a column, you want that column of air not going straight into the microphone but at a diagonal, sort of off to the side of the microphone.

Fox: That would be accurate, yes. That’s where we prefer it.

Current: So going back to special sauce and its ingredients, we talked about a couple of things. We talked about the microphone, and you’re using the U87. In my experience, the most common studio mic at stations is the Electro-Voice RE20.

Fox: The old RE20.

Current: That mic tends to have kind of a warm, bassy, almost gauzy sound to it that, in my opinion, does not work so well on the radio. So there’s the microphone, there’s the roll-off switch on the microphone to make sure that the lowest frequencies come off, there is the performance aspect of talent getting right up on the microphone but talking a little bit to the side of it so that the Ps don’t pop. I would assume that it’s also important to have a nice pop filter on there as well. What’s the next ingredient in the sauce?

Fox: It’s the actual studio itself; it’s the construction of the studio. We have a very low-reverberant studio, and we make sure that there aren’t a lot of solid walls. We want to take out the sound of the outside newsroom. Most of our facilities are all centrally located inside, right in the middle of our newsroom, so we make sure they’re isolated so you don’t hear anything else but the person speaking.

And a quick note on the RE20 — I’m a big fan of the RE20. I do see a lot of stations using it. I never veer away from anybody using the RE20. It’s a fantastic microphone. If we’re trying to be a little cost-conscious, we’ll bring in RE20s. They’re a good equivalent for the Neumann U87. But at stations I do see a lot of people using the RE20 and in that flat position, because that low-end bass attenuation on the RE20 can really be problematic at most stations.

Back when I worked in Detroit, we were an RE20 station. And a lot of on-air talent — and we get a little bit of this here at NPR — likes to sound a little bit more authoritative, and they hit the microphone into the flat position to get that bassy sound. In Detroit, we used epoxy to hold all of the switches into a position so that couldn’t happen again. And, honestly, we do see that occasionally here. We’ll hear something bassy and we’ll run up to the studio and, sure enough, somebody switched it.

Current: That’s hilarious.

Fox: So if you’re a chief engineer at a station that uses RE20s and you’re getting really angry at your on-air host for switching it, just get some Super Glue or epoxy; it’ll be fine.

Current: That’s a relatively easy fix. In terms of getting your studio to sound good so that it’s shielded from outside noise with soundproofing, you also need sound diffusion to make sure that the announcer’s voice is not echoing around the room. Are there cheap fixes that can be pursued? Or if your studio is already built out, are you just kind of stuck with it?

Fox: No, there are some little things. Over the past 10 years, you see more computer screens throughout every broadcast plant. And often that microphone is real close to the computer screen: Depending on how close, you can actually hear some of the electronic interference off the computer screen. The computer screen is the big issue. If it is just a little too close to that microphone, your voice is reflecting off of it.

Current: Right, because it’s a smooth hard surface that reflects sound.

Fox: Yes, this is one of the things I frequently see. Another thing for engineers or anybody at a station to do is to go into the studio, turn the microphone on, crank it to 11. Don’t talk — so you don’t blow your ears out — and listen to the sounds of all the fans that you have in the room. This is one challenge that we’ve had for decades here. You’re never going to get it to zero. You’re never going to get it completely silent. Nor do you really want to, because in order to get it silent you’ve got to move a lot of equipment out, and there’s a lot of cost with that.

But if you’ve got a CPU [central processing unit] right next to the microphone or somewhat close, move it two feet somewhere else and hide it under a cabinet. That helps out quite a bit.

And then there are a lot of studios built for two microphones that will be set up with six microphones for a talk show. There’s not much you can do; just try to space that out a little better. Those are the biggest things that I see on a regular basis.

Current: I haven’t actually visited the new NPR facility yet, but in the old one on Massachusetts Avenue, one of the things that always struck me was the sheer size of the on-air studios, which is counterintuitive because you’d think that the bigger the room the more echo you’d get. But with the right baffles and diffusion and all of that, the size allows all the different mic positions to be spread out really far apart, thus reducing any bleed from one mic into the next.

Fox: Our old headquarters on 635 Massachusetts Avenue was a great plant; it served our needs well for 20-odd years. You saw the large studios, but we consistently had problems with bleed in some of our smaller offline production studios. We resolved that in our new building on North Capitol Street. We made the offline production studios bigger. The actual on-air studios are larger than 635, but we were able to start with a clean slate and really put in a lot of time and effort working with our acousticians and our studio designers to make sure that we had defined what the secret sauce was and that they could design around that.

Current: Are there any other elements of the secret sauce that you haven’t revealed yet, or are those things that you will take to the grave?

Fox: Well, speaking to the public radio community, one of the things that we don’t do is process our signal heading to the stations. We keep it as pure as possible. This has been up for some debate for the longest time, but as we currently stand we don’t compress; we don’t alter our signal leaving the studio to the satellite and hitting the stations.

We know that our over 600 stations and all their listeners have different needs. So we try to give the purest sound to the stations so that they can manipulate it for their market, with varying degrees of success. We have talked about, especially with our newscast unit, maybe starting to slide a little bit of compression into that. We haven’t done it yet, and I don’t know if we will, but giving that pure signal via the satellite to our stations is probably the last element to the sauce.

Fox in an NPR studio

Fox in an NPR studio. (Photo: Caitlin Sanders, NPR)

62 thoughts on “A top audio engineer explains NPR’s signature sound

  1. Since when is 249hz “real low stuff”? That’s in the low mids range. Real low stuff would start at least an entire octave and a half below that and continue down from there. Fox clearly knows his stuff, so I wonder if this is something that was perhaps misinterpreted by the author.

    • It’s a transcript of an interview from a podcast, so feel free to listen and decide for yourself if we misinterpreted it. Two things I would bear in mind: 1) It’s an extemporaneous conversation; 2) “Real low” is subjective.

    • Technically, this report is written at about a 3rd grade level. It is purposefully dumbed down like for a script. After all, this is NPR not engineering school. Propaganda is what they do.

        • That asshole you interviewed, the one in the suit and starched shirt like Ed Sullivan, he hasn’t done any work in many many years. He pays to get his hands manicured is his idea of work.

          He is not going to tell you, “There is a six-layer desk, and by one action it is possible to switch between all layers. “”FaderGlow””, which applies color to the fader slits, is equipped as a standard feature making it possible to set up to eight colors according to function.”

          • Studer is now part of Harman Pro, which is US based. I don’t know how much production and engineering is still carried out in Switzerland, but it’s pretty obvious that they are at least sharing engineering with Soundcraft and other parts of the technology group.

            In fact at this point, I’d say that Harman are using the Studer/Soundcraft brands to segment the same technology for studio/live sound.

          • I guess you are not keeping track of the trajectory of Soundcraft ever since they started importing their shared tech from Studer. Too bad that Sidney is passed on. I think he would have run a tighter ship than the “I’ll get back to you later on that” tech support, and “Go Look at YouTube” for product operational documentation. The best part is when the service dept knows something that they are not telling you, and then you have to find out the real answer from Soundcraft UK. But that guy doesn’t work there any more, the one who actually had the balls to tell the truth about something to a client who lays out the money. USA really has become the land of fronting and lies and it has filtered down to some companies right down to the street level. Reminds me of Comcast. Instruct them to port a telephone number out of Comcast. A month later it has not happened and no one (Obama?) is regulating them. It really sucks, it sucks badly. It is rot and corruption, since you mention it. Studer company culture has zero to do with what has become of Soundcraft. I would not mention the two companies in the same sentence, out of respect for the one that does their own work.

          • >Studer is now part of Harman Pro, which is US based.

            “Now” as in 20 years ago when they still had office in Nashville?

          • you have a hell of a chip on your shoulder there guy.
            if you dislike NPR so much, stop listening and stop poisoning their comment sections.

          • Interesting that you advise self censorship. What is this, North Korea? You do not realise the significance you what you advocate, particularly in the context of state run broadcast radio. You should go work for Kim Jun. You’d be right at home as an officer of the thought police. Never minds that the party members live in estates and everybody else has a little hut to live in and some cane to chew on, maybe beans if they are lucky. Sounds like many parts of the USA. Those “I’m your good neighbor” NPR talkers are paid executive compensation ($300k/ year for some) to prop it up. I admit Rush Limbaugh is paid more, but both of them are in the same business and have zero to do with working people ZERO. ZERRR-ROOO. Go self censor that. Let me know when NPR has some programming relevant to working people. I plan on living to be a hundred. I’ll be waiting. That is a long damn time to self censor myself like a good little cookie, a good little North Korean. USA home of war crimes and for-profit prisons and a trillion dollars in school debt, and the most people in jails and prisons in the history of the world and you won’t hear it on NPR.

          • As your Fearless Leader and Dictator, I demand that go outside and take some deep breaths, maybe go for a nice walk.

          • How about if you just turn off NPR and instead broadcast relevant news instead of Teletubbies brainwashing sister company to CNN. Hey let’s
            “Do the numbers” “Ack! Ack! Ack!” Nice job catering to the 1% to 10% max. The guy “Doin the numbers” is an ex-NSA/CIA intelligence man. Now THAT’s FUNNY Mr. state media thought police. When is NPR going to do right? Not ever. Not even if they sound like the fresh boy next door. When is NPR going to do real reporting on the #1 cause of bankruptcy in the USA, which is medical bills. Talk about totally irrelevant state run media propping up the Borg. You ought to be ashamed of yourself.

          • Those are jobs with high level security clearances. Mr. Ryssdal is a life member of the state party. It really is no different than North Korea.

          • Well this is the place to talk about how this works, once the person has high level security clearances. Goes like this, “any other officer or employee of the Federal Government may not provide
            any information about [insert subject area] to any person
            outside the Federal Government.”

            Kind of an interesting predicament for someone in broadcast journalism, don’t you think? Read this and get “the vibe.” https://www.congress.gov/bill/112th-congress/house-bill/2819/text

            It’s codified! Spelled out. Pentagon staff officer? as a prerequisite to one of those Voices of NPR. Pentagon staff officer and US State Dept talking major beaucoup clearances. This is the version you do not get, little duckling, who went to broadcast journalism school. It is the old two-sided two-faced coin, the script. Now read the script and do play-acting as “best friend” to the public. Ronald Reagan institutionalised the technique. There is an inside room, and if you do not have clearances you are not invited. I would not call that “Public.” -At least I’m dumb enough to tell you. Or is it tell-on you?

            Edit: USA, ha. Where you have congressional directive, “To prohibit (insert demographic group) from providing information”

            It’s official!

          • You don’t know anything about this little duckling, and since you apparently don’t know how to google such things, I don’t much trust your knowledge or judgment on anything else.

          • You are supposed to think for yourself, not “trust” me. I am not selling anything or running any game. I get no check from what I am telling you, it is not a con and there is nothing to trust. Google what? Look up how the USA is rated as a “free country.” Not very well. The context here is that maybe unwittingly you have put your foot on a frog (NPR) that is one of the primary social directors and propaganda outlets in the US. Btw, now that I have you on the phone, what is with the hypey title of your piece, “Top audio engineer.” I mean yes, the guy has a plum gig, but it is voice to broadcast. He is not making David Bowie recorded or mixing 200 inputs on a theatrical production. Does interviewing NPR (or anyone else) mean that you have to pander in the title? These fellow is not a top engineer to me and he is going light on engineering discussion. I have known and worked with some really top level audio engineers. I guess they are not ever talked about. Maybe there is more to your work for me to learn from. You know there are only so many hours in the day. This article is really about what microphone is specified for NPR studios.

          • That’s a good story. I’ve never heard it or heard it mentioned on NPR. You would think if it affects 25% of the people, that NPR would mention it 25% of the time. It would be interesting to see a distribution pie chart of what this broadcast organization broadcasts. I know that their programming varies widely between cities and markets. The highly educated “enlightened” cities are served “cool and meaningful” programming. Well let me tell you, it is a different serving outside of the educated bubbles, a different serving from the media. And I’m talking about you, NPR. See, out in the hinterlands is where the financial predators feed.

          • This is going to sound like I am telling you a joke, but it is true. So I am in the car and turn on NPR News to get some nice fresh U87. I think of you, Mark, and I make certain that it is in fact NPR programming. Check/ confirmed. The lead story is about the type of housing and comforts for finance industry titans and which cities are suitable for them to live in, cities with lots of super-fine restaurants to choose from, and lots of executive housing fitting to their wants. The next story is about refugees being brought into a US city and put into a small town. Then, a pause with the local news, then back to NPR material. Final story is about craft breweries and “hoppy malty.” Now Mark, is there anything at all here for working class Americans? Anything? The finance class luxury housing story? No. Having your town invaded by assigned refugees? Well it is not for the US people, that is for sure. And finally, who consumes craft (boutique) beer? 2% of the populace? Where is the story for the people who work the hours and try and pay their bills? The next time you are in the PetSmart or Pets-R-Us store, ask the cash register worker which craft beer they prefer. Then ask them how it is when they go to the doctor and if they get a physical every year. Yes, NPR does not fail to deliver pre-pack psyop as news. News?! This Is News?!!! NEWS? Two of the stories are about people pleasuring themselves, and the third is not about Americans. This is National “Public” Radio, that has zero relevance to 98% of the American public.

  2. Very cool to read this! When you listen to NPR shows in a podcast format, it becomes painfully obvious that no compression is used. You don’t have the benefit of the Orban multi-band compressor doing its thing before the FM transmitter. The levels are all over the place, and hard to hear sometimes in the car or in a noisy environment. The U87 rolloff starts around 150Hz, which is pretty high as low-cut filters go, but a good way to counteract proximity effect

    • I completely agree about the need for compression on NPR podcasts. I’d much rather use the NPR app/podcast vs RF (better quality/signal) but don’t because of the lack of compression!

      • >because of the lack of -> production

        PS A mix engineer could easily put a multiband compression on a L,R output of that console. The tools are there in the console. One mix for broadcast, one mix for podcast. This is the extent of difficulty in life? Just reading an article on a 52 discrete channel surround mix set up for a museum installation. I bet you a dollar the NPR engineer can handle it. Maybe they have to look up the .pdf file manual so they can get $10k use out of their $200k console.

      • Adam, 6 dB increase in loudness is about a 60% increase in volume to the listener. Sounds like a crisis of professionalism to me, a quality control issue. Good of you to call it out. This is why we have standards, and why professionals attend to technical standards in their work. Reference: AES, Audio Engineering Society.

  3. A technical correction: when bass is “rolled off” it doesn’t mean that bass can’t be heard (which is what Fox’s quote says). What it means is that the amount of bass is reduced (vs when the switch on the mic is off). If you look at the EQ response of a U87, it’s actually reducing bass up to around 400 Hz (not 250). As a pro audio engineer, I get the gist of what they meant, but it’s not technically accurate.

    Interesting read!

    • Huh? Stop confronting the dumbing-down in progress. It’s what they do! Someone might be offended that you went off-script, Tonto.

    • That’s a $200k Studer digital console in the picture. This article is just so much fluff and bs. It is lying by omission is what it really is. Listen to NPR if you want a controlled script giving you part of the story. Kind of like CNN and NPR are two sides of the same coin – state level propaganda, top down message.

  4. I think this would be a lot better interview if the host would share less of his opinion and let the expert guest share more of his instead. You are not a studio engineer who maintains studios that generate a signature sound, he is.

    • I don’t think I crowded Fox out of saying anything he wanted to say. This is a transcription of an interview from a podcast, maybe it doesn’t work as well in print, but I think if you listen you’ll see our back-and-fourth was quite natural. Sorry you don’t like my style.

      • This isn’t about your style, it is about the basics of interviewing. The interviewer should not phrase questions or make statements that the interviewee may have to contradict. Instead, ask open questions that are not leading. It isn’t polite to contradict, especially when you are a guest. Notice how he doesn’t immediately tell you that he likes RE20’s? You suggest flipping switches despite engineer’s decisions in the studio, and only later does he tell you that he epoxies things down to stop people like you!

        If you want to provide your audience with useful information, talk less and listen more.

  5. I realize I’m joining this discussion a bit late. Can anyone tell me if these mics are all the older style U87 or if some or all are more recent iterations such as the U87ai?

  6. Hey Adam I do not know how well it mixes, private sector and public broadcasting, or the limits of your mission. Maybe you should go over to Pinewood and see what they are doing and maybe see who is the top audio engineer over there. About a 50-60 mile drive for you, and then again on the return. Or check out the on-site Home Depot specifically for supplying production support. You have the production facility Tyler Perry is building where he just purchased a dormant army base not too far from Pinewood. Maybe you can find a top level audio engineer there, someone who takes on challenges and does pre and post production. Could get kind of intense. You might need some help. Maybe you have to get editorial permission.

    • I thought this conversation was about the acoustic sound of NPR. Some people get a little carried away. This is human nature I guess.

      • This article is highly incomplete. That’s the con and you are singing the song, “the acoustic sound.” Broadcast goes through voltage, and lots of it on the transmitter side. Last time I mixed a broadcast, the room acoustics has a lot less to with it than the console and processing. The “acoustic sound.” What, you work in a hi-fi shop selling turntable cartridges? Reminds of Yamaha and their “Natural Sound” con job that has zero to do with anything. I like how it is now a conversation for digital storytelling. Its the corporate quaalude to put you to sleep. I can deal with NPR being state run propaganda, but I an not going to have people do half-way mumbling and elevate and romanticise them. Obviously I do not need an idol for idol worship.

        • I would prescribe klonopin in your case or maybe some sour diesel cannabis. Chill my friend. In the light of eternity, it does matter very much at all.

          • As we have seen way too many times, when abused ANY drug can be deadly. I have been on it for 26 yrs. but use it very sparingly. The cannabis route is much safer plus you get high as a lagniappe.

  7. Nice acoustic diffusers on the back wall in the lead photo. Simple and effective. Anyone could build them.

  8. The crime of this interview is that real life is hard and takes hard work. Real knowledge is complex and takes hard work. Audio systems are decidedly not “flip the low-cut switch and plug in the microphone.” This is an insult to any serious person that has to work hard and learn real skills. Per the usual pattern, NPR role-plays and softens the message to the point of fraud. Meanwhile, they are praised as enlightened, as “Special.” There is a weird subtle intimidation, “You want access to this interview, you better flatter and keep it polished.” Real life is not polished. Upper class is polished like at a country club where everyone has manners. The only thing “public” about NPR is the transmitter aimed at the public. This article is a crime. Don’t do it. Do not think. Do not commit thought crime. Accept.

  9. I quit listening to NPR thanks to the Neumann u-87. I am not one who can tolerate every his through the nose every Swisher spittle in the mouth every smack of the lips. Ever since this signature sound you claim came to be in effect I have found myself trying to listen to NPR once or twice only to have to cut it off within 3 minutes because of somebody sounding like they’re trying to eat a bowl of cereal in my face while they talk. Leave high definition for visual high definition is not meant for the mouth. This is why there’s a trending cliche known as NPR mouth used to describe anything obnoxious about annoying mouth sounds in advertising and radio. So good for your sound Engineers I’ve got the sound so perfect they’re driving away listeners. Good job.

    • Couldn’t agree with you more.
      Some speakers are worse than others…Every lip smack, every tooth suck, every swallow…I can’t stand it.

  10. Reviving an old topic- the U87 bass roll off switch affects frequencies below 1KHz, virtually eliminating everything below 50 Hz, according to Neumann’s U87 operating manual.

  11. Pingback: Hongkong Podcast Editing Tips | Best Podcast Editor

  12. Pingback: Go straight to hell boys | Sonic Journeys

  13. Pingback: WHAT’S THAT SOUND? | Goldsmith Media Group

  14. I have such a hard time hearing NPR in the car, I’ve almost given up. The top end and high voices come through clearly, but so many of the hosts and reporters have low tones to their voices, or “throw away” the parts of their sentences and it gets lost. I’m constantly turning the volume up and down based on who is speaking and I miss a percentage of what is said. I feel like all levels need better balance in terms of volume. I don’t know anything about sound engineering, what I know is – my husband and I can’t hear NPR well.

Leave a Reply

Your email address will not be published. Required fields are marked *