The Podcast Pedant

As you’re listening to your own podcast, do your recordings sound flat to you? Do they fail to hold your interest? Another tip for recording: smile.

I’m not kidding. It changes the shape of your mouth, and therefore your tone, in a way that people naturally find more energetic and likeable. You’ll get more brightness and range. It also changes your own frame of mind, and impels you to have more fun, if you don’t fight it. You’ll enjoy yourself more; and therefore you’ll sound like you’re enjoying yourself more; and that predisposes your audience to enjoying themselves more too.

Will you feel stupid making yourself smile and talk? Yeah. You’ll feel like a complete idiot. But when you listen to your podcast a couple days later, you’ll hear the difference; and next time it’ll be a bit easier, and the time after that it’ll be easier yet.

It’s something to try.

So I finally got a hardware compressor.¹ I was beginning to get tired of applying the same software compression in Soundtrack Pro over and over again — and also tired of the occasional clipping which software can’t solve. I wanted to move just one step closer to a “live to tape” scenario, although I know for certain I’ll never really get there.

After a few days of obsessive review-reading and wallet-shaking, I settled on the RNC 1773 from FMR Audio:

The RNC literally stands for “Really Nice Compressor.” You’ve got to respect that level of forthrightness. It’s a no-frills stereo compressor: there’s no gate, no limiter, no expander, no dual mono mode with extra knobs. What it does have are two compression modes:

1. Normal Mode, which acts exactly like a compressor should. I played around with this a little, and it sounded okay. The attack is very fast and hard, which I found a little jarring. Doubtless I could adjust it, but I never got that far, because what I really wanted to try was:
2. Super Nice Mode, which chains three compressor circuits in series for a very gentle, very transparent compression effect that still retains all its power.

If you Google on the thing you’ll find a large number of sound engineers who swear that the RNC is the best compressor you can find for less than $2,000. I got mine for $175 (plus state tax, and with some cables thrown in) at Humbucker Music, whom I will attest are a great bunch of folks. I ordered it on Wednesday, and it was at my door Thursday afternoon.

I’ve used it twice now for podcasting. I wish I could say something like, “Wow! All I had to was turn the thing on and nightingales dropped dead from envy at my feet.” Unfortunately, it didn’t work that way. New sound gear never works that way. Even on Super Nice mode, I’m still working on tweaking the settings just right. My first attempt (for last week’s Escape Pod intro) used a 6:1 ratio, -8 dB threshold, +6 dB gain. I personally think it came out sounding overpowered, way too flat and pushy. For this week’s intro I used a 4:1 ratio, and made the threshold and gain even at 8 dB (which the manual recommends). It wasn’t flat this time, but I clipped frequently. This could mean I need to make my mic gain part of the equation too.

Don’t take this as criticism of the compressor. The RNC does what it’s supposed to, and it really is beautifully transparent. There’s no change at all in the sound’s tone or noise, just its volume, and that’s rare and lovely. I’m being honest with you about my trials to make the point that there’s no magic bullet. The more gear you have, the more skill you need to develop. Once I learn to use the thing properly, then I believe it will add a volume and clarity to my podcast that will make the investment more than worth it. I can already sense parts of that. It’s just a matter of getting all the pieces into place.



1. I should probably write a post at some point about the finer points of compression and what compressors do. For now, if you didn’t already know, take this for a definition: “A compressor evens out the volume of your signal by making loud sounds quieter.” ↩

Once, long ago, I was subscribed to a podcast. This podcast was a rather prominent member of The Sci-Fi Podcast Network.¹ So at the start of each file, it began with a TSFPN audio tag: “This is TSFPN.com. You’ve found the best podcasts in the universe.“

Which is all wonderful. Except that where the podcast itself sounded smooth and high-quality, the audio tag sounded like Alvin the Android Chipmunk. It was too fast and too high-pitched.

Half of you are nodding now: you’ve seen this happen before. There are a lot of ways to make this mistake, but one of the most common is to combine two sound files of different sample rates in the same Audacity track. It’s easily done, and easily fixed. Just put them in different tracks and then mix down.

But here’s the kicker. I listened to this podcast, and heard the exact same chipmunked tag, every week for four months. It was the very first thing you heard. It sounded terrible, and it was never caught. I eventually lost interest in the subject material and unsubscribed, but for all I know it’s still going on.

That was definitive proof to me that the guy never listened to his own podcast. If he did, he’d have noticed and fixed this easy bug. He didn’t, and he started off on the wrong foot every week, and never knew it.² And that’s today’s lesson:

First, you should always listen to your MP3 file before you upload it. If it’s a long podcast, at least skip through it to make sure all the pieces are there and sound like they should. Never skip this, or you’ll be sure to embarrass yourself with some technical gaffe sooner or later. Even if it’s 6 AM by the time I upload, I always take at least a minute or two to jump through my podcast beginning, section transitions, and ending. Those are where mistakes are likeliest to happen.

Second, you should subscribe to your own podcast feed and listen to it with all your other podcasts. This means you’ll catch any RSS screw-ups without having to have your audience tell you about them; but more than that, it gives you the opportunity to evaluate yourself as a listener and decide if there’s anything from week to week that needs improvement. Are your levels uneven when you listen on your car stereo? Great, now you know. Did you drone on too long about something unexciting? It’s easier to notice that a couple days later, and you’ll be more conscious about it next time. Continuous improvement means continually evaluating your work, and the best way to do that is to listen to yourself the same way everybody else does.

This will seem like common sense to a lot of you. Some of you will find it inconceivable not to listen to your own stuff — after all, if you didn’t like to hear your own voice, why podcast? But in practice it’s very, very common to skip these steps.

You do so at your peril. If not the peril of losing audience and reputation, at least the peril that someday some smartass like me will make a blog post about you. And who wants that?

1. Which, if you click on the link, you will see has lately dissolved into a cheerful puddle of brightly colored goo. But that’s another story. ↩
2. Should I have dropped him a friendly e-mail? Probably, and in most cases I would have. But there were some personality factors, too, and… Well, I didn’t. So. ↩

(This began another recycled post from the Alley — but then I started elaborating, and now very few of the original words have survived.)

The home recording business is moving more and more these days towards digital interfaces on all their low- and mid-end stuff. This makes sense: almost everybody is recording into a computer now, so if putting a USB interface on your otherwise-average mixer will give it a competitive edge, why not? It simplifies life for the musician or podcaster, and sending a digital signal to your computer means you aren’t bound by the computer’s sound card. ¹

I’m of the opinion that it’s a very good idea to have a USB or Firewire interface for your audio — whether it’s built into your mixer, or a separate interface box, or even a standalone USB microphone. But if you can afford it, get Firewire.

Why?  What’s the difference between Firewire and USB? From the outside, they’re both just cables. You can chain stuff together or put them on hubs, and you can buy a lot of the same sorts of devices for each. USB 2.0 and Firewire 400 have similar speeds “on paper.” (480 megabits per seconds and 400 megabits, respectively.)

But even a casual look will show that USB audio devices tend to be cheap and Firewire devices tend to be expensive (barring the odd exception). And just about all A/V geeks will agree that Firewire is better, and even necessary for complex work. Why is that?

The difference is a fundamental one, but it’s internal to the way these things work, and pretty subtle. The following represents my best understanding based on a few hours of research and being a computer geek for several years. If I get any of it wrong, someone please correct me.

USB

USB is a hosted networking protocol — which means every device talks only to your computer and is utterly reliant on your computer to tell it what to do. You can have up to 128 devices on a USB network, and they might be chained and hubbed in all sorts of interesting ways, but they’re all just passing their bits back to the computer to decide what to do with them.

USB devices are asynchronous, which means that any device has the power to send any amount of data at any time. If two devices decide to talk at once, their data can collide with each other. If the traffic’s not highly time-sensitive, this isn’t a big deal. There are routines in place to manage it, and you’ll never notice if your mouse click happens a couple microseconds later. BUT. There are a few applications where it matters, and one of them is audio. An audio interface is sending a constant stream of sound data back to the computer. It rarely uses up the whole pipe, and so it’s still possible for other devices to talk — but the odds of collisions are higher, and if you get too much other traffic the errors can pile up beyond the computer’s ability to stay “caught up” and you lose some sound data.

The details of the protocol are typically implemented in software. That makes it very cheap and easy, as it pushes the work onto the CPU, and the devices themselves don’t have to be very smart. But it also means that USB traffic has a direct impact on system load, and vice versa. For most common applications this doesn’t matter — the traffic from your keyboard and mouse is so slight that it’s hardly going to bring your system to a crashing halt, and most of us wouldn’t notice if our external hard drives slow down for a second when our screen saver kicks in. For audio, however, it does matter. Same problem as collisions. Audio devices are pushing data out at a constant rate, and if the computer’s too busy running sound effects or switching programs or swapping out RAM to pay attention, some data can get lost. And then you get glitches and/or latency.

In case anyone’s wondering, audio glitches are bad. Latency (the delay between the creation of a signal and the final reception of it) isn’t quite as bad unless you’re trying to monitor yourself through headphones while you’re talking. If you are, a latency of a fraction of a second can be unsettling. Like you’re living in the future.

Firewire

Firewire is a peer-to-peer protocol,² meaning that every device on a Firewire network is equally capable of talking to every other device. Two video cameras on a Firewire network can share data with each other. A Firewire audio interface could save sound data directly to a Firewire hard drive. ³ Your computer is just another peer on this network, and has no inherent special status.

Firewire is always implemented in hardware, with a special controller chip on every device. So the load it puts on your CPU is much lighter than USB communications load, and you’re much less likely to lose any sound data just because you’re running fifteen things at once. Specialized hardware usually makes things faster and more reliable, and this is one of those times. (By the way, it’s also one of the things that makes Firewire more expensive. It’s also a reason Apple dropped it from the iPod Nano — there was just no room for the Firewire chip.)

But the real reason Firewire is more reliable than USB is more fundamental than that. It’s because Firewire allows two operating modes. One is asynchronous, as we described above with USB. The other is isochronous mode, and it lets a device carve out a certain dedicated amount of bandwidth that other devices can’t touch. It gets a certain number of time slices each second all its own. The advantages for audio should be obvious: that stream of data can just keep on flowing, and as long as there isn’t more bandwidth demand than the wire can handle (not very likely) nothing will interfere with it. No collisions, no glitches.

From a practical perspective, this also makes it safer to send a lot more audio via Firewire. That’s why most of the multichannel interfaces (18 channels, 24 channels, etc.) are Firewire devices, and USB devices usually just send a two-channel stereo signal.

So there you have it. For hooking up your mouse, keyboard or thumb drive, USB is plenty fast and plenty cheap. For hard drives, either one will do (although Firewire is somewhat more reliable). For audio devices, USB will do fine if no other devices are competing with it and if you have processor room to spare. But Firewire will always be able to handle more load with lower latency and no glitches, because it has resources it can set aside to make sure your audio gets where it needs to go.

…And that’s why Firewire’s more expensive and taken more seriously.

Just in case you were wondering.

1. Which is good, because most sound cards are crap for recording. And even the ones that aren’t have to worry about electrical noise from the rest of the stuff inside your computer case. ↩
2. Hello, RIAA search bots! No, I’m not talking about that kind of P2P. Go away. ↩
3. Note that it could. I’m not completely sure whether any actually do, but I’d appreciate information on this. ↩

Want a fast, cheap, probably obvious trick for noticeably improving your podcast voice? Stand up when you talk. It’s well known that standing straightens the air passages and gives you a lot more power. It’s why you always see singers and voiceover artists standing. I started standing for my story readings and intros when I moved my podcast down into the basement studio, and I could immediately tell the difference.¹

Obviously this trick won’t work for everyone. You need either a tall mic stand or a boom arm, and your content and style may preclude it. Couplecasters would likely find it too weird to chat with each other while standing; and if your podcast is mostly improvised monologue, you may find it difficult to (literally) think on your feet. Some people do it very well, or learn to, but it’s far more important to feel comfortable. Standing up probably works best for pre-scripted content, or improvisation with a semi-formal or formal structure to it.

But if you can make it work for you, and make it natural, you’ll be surprised at the difference it makes to your sound — and, very possibly, to your attitude and confidence in yourself and your podcast.

1. On top of the acoustic and psychological benefits of moving my podcast out of the living room. ↩

This post is adapted from a reply I made on Podcast Alley. The question, from a new member identified only as “flextone,” was about getting started with sound improvement using Cakewalk and/or Adobe Audition:

I would appreciate any tips for getting the best out of the sound file such as what compression, EQ settings, adding any FX such as a little reverb.

The answer, of course, is that there are no universal “right settings” for any of these effects. If there were, there wouldn’t be any need for parameters and controls — just a single “Make It Sound Good” button. That said, here’s the best advice I can offer for anyone looking to find their right settings, based on my experience producing an audiobookish podcast and editing a wide variety of sound files sent to me from the wild.

Before you touch anything:
Take that money you were thinking about saving for a better microphone, and go out and buy a $100 pair of studio headphones from a real music outlet. I mean it. Good headphones are more important than a good mic, because if you can’t hear your podcast you have no idea if you’re improving it. I talk more about that subject here.

Compression:
The best tutorial I’ve ever seen on compression is this one from dbx. The recommendations it makes are an excellent starting point. From there, I recommend you experiment and decide what sounds best for you. In general, moderate-to-fast attack/release settings are better for voice and moderate-to-slow settings are better for most music.

Also in general, it sounds better to undercompress than to overcompress. You’ll know you need more compression if you’re driving in your car, listening to your own podcast, and find yourself having to ride the volume knob.¹ You’ll know you have too much compression if every little sound you make is as loud as your most emphatic speech. This is boring and undramatic, and it increases listener fatigue.

Equalization:
First, you may not need it. Lots of people sound better without any EQ applied than they do with a great deal of fiddly sound sculpting. If you aren’t sure, try it both ways and do a blind test with a couple of honest friends. Don’t ask your spouse or family — they’re probably too biased towards your regular voice — but try to find friends who don’t care too much. (Have them listen with your good headphones, of course.) If you do need it, exactly what you need will depend entirely on your voice and the tone you want for your podcast.

I have an unusually ‘bright’ microphone, and I’ve always thought my voice sounded a bit too nasal, so I use the parametric EQ on my Mackie to give myself a little bit of 80 Hz boost and a moderate boost at 1000 Hz. I think it makes my voice a little deeper and rounder.²

Other general EQ thoughts:

• Experiment, then listen, then experiment some more. Record yourself with a ton of different settings, announcing each setting as you change it, and with frequent “EQ off” breaks for contrast.
• Small to moderate changes are much better than big ones. Just because you can add 15 dB to the high end of your voice range doesn’t mean you should. Ever. Unless you’re producing a cartoon.
• If you’re a man, don’t just blindly pump up the bass in the expectation that it will make you sound like Movie Preview Guy. It probably won’t, and even if it does, nobody wants to listen to Movie Preview Guy talk for half an hour.³
• If a total stranger can tell you’ve used EQ on your voice, you’ve failed.

Noise reduction:
You should learn about this and know how to use it for special circumstances: field recording, voicemail messages and bad Skype calls, that time when you never noticed the air conditioning was on, etc. It always distorts the signal, but sometimes that’s better than the alternative. Do not allow it to become a routine production step in your podcast. If you have noise issues every time you record, find out what’s going on in your environment and fix it. Don’t form the habit of relying on a magic button to fix it for you. It won’t.

Expansion/gating:
An expander is the lesser-known opposite of a compressor: it makes quiet sounds quieter, and leaves loud sounds intact. Depending on your microphone’s sensitivity and your own speech habits, you might find an expander useful for cutting down on distracting breath noises, lip and tongue smacks, etc. It can be useful for cutting down background noise in specific circumstances — it’s kinder and gentler than noise reduction — but again, it’s better to fix the problem instead. If you do use an expander, use it at moderate settings. Reducing breath so that it’s less audible sounds natural; an entire podcast with no breath noises at all sounds cold and alien.

That brings us to gates (aka “noise gates.”) A gate is simply an expander with a very high ratio, just as a limiter is a compressor with a very high ratio.⁴ It makes noises below the threshold disappear. I used to use gate effects fairly often, but I almost never do it any more. It’s too hard to get the threshold and attack/release just right so you don’t lose any signal and the transitions aren’t too harsh. And if you try to use a gate to cut out major noise problems, it still leaves the noise in when you’re talking, so the end result has the “on/off” clipping of a walkie-talkie.

Reverb:
This is one of the most overused effects in podcasting. Don’t apply reverb to your normal speaking voice just because it’s there and it sounds cool. It gets old very fast. Reverb is for music and used car commercials. Sounding like you’re podcasting from an empty cathedral won’t make your words any better, and will usually get in the way of people understanding them. If you do decide to use it, go with the subtlest possible settings that achieve your goal. Again, if a listener is able to say, “Aha! They’re using reverb!” — you’ve failed.

Anecdotal exception: I’ve used reverb for general speech exactly once. It was a story recording I’d received from another narrator, and it had high-frequency noise issues. Using SoundSoap took care of the noise but left the speech sounding noticeably distorted — that “tin box underwater” sound. A tiny bit of Soundtrack Pro’s reverb, using a small room size and a fairly short reverb time, helped to smooth out that distortion and bury it in the reverb effect. Even with low settings the reverb was noticeable, but worked okay in context — the voice for this particular story wasn’t supposed to be a normal human anyway.

Other effects:
By all means, play with everything until you’ve learned what it all does. Unless you have a specific reason for a specific effect, however, don’t actually use them in stuff people will hear. I’ve never heard a podcast where I thought, “Yeah, that guy needs more tremolo when he’s talking about cars,” or “If she’d only pitch-shifted her voice half an octave, I’d have learned a lot more about William Shatner’s butt!”

Most of those effects are gimmicks, and while the Beastie Boys may have uses for all of them, they sound pretty stupid on a single voice in isolation. It may be funny, in a vague sort of way, but “vaguely funny” doesn’t usually last the length of a podcast.

The mark of good sound engineering is that it’s transparent. If you do the job well, no one will even notice that you’ve done it. Your podcast will simply sound good enough that nobody pays attention to the sound; they’ll be listening to the words instead. Having people notice your effects usually means you’ve done them wrong.

Finally, listen to your own podcast.

Other comments and experiences encouraged as always, of course. This is one of those posts that I’m hoping will remain a document-in-progress.

1. That, or you need better mic placement and levels when recording. ↩
2. I’ve only had one criticism: my high school ex-girlfriend e-mailed me to say I was trying to talk too deep and sounded fake. But her experience with my voice isn’t that of the world at large, so I filed her opinion under H for “Huh.” and kept doing what I was doing. ↩
3. For women, I suppose the equivalent sin would be pumping the 6k to 8k range in the hopes of sounding more sibilant and sexy; but I honestly don’t think women go in as much for this particular sort of vanity. ↩
4. ”Very high” may mean “infinite” in both cases, but it might also mean something like 20:1. ↩

So the word is that Apple is now shipping the universal binary version of its Final Cut Studio applications. They said they’d have it out in March, and it looks like they made it just under the wire. For those who don’t speak Modern Mac: “universal binary” means it’ll run on the older PowerPC Macs and the new Intel ones.

This is of interest to me because I do my podcast production in Soundtrack Pro on my little Mac Mini in the basement. I have a much nicer dual core Intel iMac¹ but since I haven’t been able to run STP on it, I’ve been using it for everything except podcasting. I’ll still record on the Mini no matter what, but it’d be nice to do some of my editing upstairs where it’s warm and carpeted.

Crossgrade deals are available. The major problem here is, it doesn’t appear that I can get a universal version of just Soundtrack Pro. Instead they’re ditching the standalone applications, and making me an offer to pay two hundred bucks to get the full-blown Final Cut Studio. Now, $200 is a hell of a deal for all those applications. Would I use them? I don’t know. Final Cut would be nice if I ever do video podcasting. Right now I’m neither for nor against that notion. The other programs (3D animation and DVD mastering) just aren’t in my interest set.

So I’m trying to decide. “Crossgrade” or not? I realize this is exactly what Apple’s infamous for — making it very easy for you to spend more money with them to get what you’ve already got — but, well, they do make good stuff. And if I don’t do this, the only way to improve on my current workflow will be to ditch Soundtrack Pro for something else.  I doubt there’s anything for the Mac right now that’s better for fine-grained podcast crafting.

Opinions welcome.

1. Bought it with this year’s tax refund. Hooray for having a baby! ↩

This is something that’s been bugging me for a while. As I keep moving along my own hardware and software path, adding bits and pieces as the money meanders in, I’ve come to the conclusion that the books and forums and that smartass who tells you what to do because he’s been podcasting two weeks longer than you all have it wrong.¹

Let’s say you have $100 to spend on your podcast. You want those dollars to go as far as possible to improving your sound. All of the above sources will tell you “buy a decent mic.” Either a USB microphone like the Snowball or Samson, or an MXL mic and a Behringer mixer to plug it into.

Am I right? Have you heard this advice before? It’s wrong.

Unless your current rig is a tin can with a string, the first money you put into your podcast should not go into improving your input. That’s step two. Your first hundred dollars should go into buying a good pair of headphones.

I’m speaking specifically to podcasters who do postproduction here. If you’re a “talk-and-send” podcaster, you don’t need this advice.² If you spend any time after you record on editing, compression, trying to improve the sound quality, you need to be able to hear your podcast. Most likely you’re listening through your computer speakers, or your iPod earbuds, or whatever Best Buy was selling with the best-looking numbers and graphs on the back. You probably think you can hear your podcast. You’re probably wrong.

I learned this lesson back in December. Before then I’d been doing the Best Buy thing, and I was editing my podcast with a $30 pair of Jensen headphones. I thought I was doing an okay job.

Then a little before Christmas, I went looking to upgrade a few things. I went to Guitar Center and tried a few studio headphones in the $100 range: Sony, AKG and Sennheiser. They all sounded great. The AKG set was by far the most comfortable, but I went with the Sennheiser HD-280 because it was a closed design so you could hear more at lower volume.

I took them home and listened to the podcast I’d uploaded the day before. The difference was night and day. I had no idea how many sound artifacts I’d been missing.

Since then I’ve been able to tune my noise reduction better, get more accurate EQ and compression, and catch more clicks and pops that I’d have missed otherwise. And that’s why I think you should put headphones before a better microphone: with good output, you can do better sound adjustments and compensate for defects. If you don’t have good output, you can’t tell what you’re not hearing.

The same advice would apply for speakers, of course. Even expensive computer speakers made for gaming won’t give you the sound reproduction you need: they’ll pump up the bass and try to make things sound artificially good. You don’t want pleasant sound, you want accurate sound when you’re editing. You need studio reference monitors. But those are very expensive, and lacking any knowledge of good monitor speakers in the $100 range, I’d say go for headphones first.

Then buy the better mic as soon as you’re able, because you’ll want it even more after you’ve got the studio headphones. Bad sound will start to get on your nerves, because it’s obvious and glaring. But I contend that if you can’t hear the mic properly first, there’s no point in improving it until you can.

1. Unless, of course, that smartass is me. But I’ve been wrong on this too. ↩
2. Nor my broader advice, which would be “Stop doing that.” ↩

A simple thread about cheap microphones on the Alley disintegrated into a multi-front debate about elitism, personal finance, and exactly what price range for recommendations constitutes an unacceptable barrier for starting podcasters. It’s all too tedious for me to trouble you with.

But from the flames emerged this old bugbear, with a throaty growl and singed fur:

Michel: I know you mention content a few lines later, so don’t think I’m ignoring that, but content is so much more important that sound quality. Good job if you have fantastic sound quality and a boring show. It’s the best sounding crap out there.

Sigh. You know what? I’m getting tired of that whole argument. People keep bringing it up as an excuse to ignore the technical aspects of their craft. It’s like arguing that roots are more important to a tree than leaves. Yes, you need to start with content. No content, no show. But if you don’t eventually open up and sprout some production quality on the outside, that tree is gonna die.

The ideal podcaster is constructively dissatisfied with everything. We should all want better content next week than we’ve got this week. And we should want better sound quality, too. That doesn’t mean spending a lot of money. I’ve given lectures on how to get better sound out of free tools. Eventually you do hit a threshold where you need to scale up to better equipment — but not for a while. That’s fine. In fact it’s probably better than starting expensive, because getting good sound out of cheap gear will give you the skills to use the good gear properly.

But you need to care. You need to listen to your show, pay attention to what’s good and bad about it, and strive to make the whole thing better. If you’re apathetic about any part of the experience you present to your audience, your audience will be apathetic too.