In the late 90s, when getting audio posted on the internet was a little more complicated. I remember using a headset mic to record audio with a Sound Blaster 16 PCI Card, and editing it using SB Studio on my 486SX PC.
Yes, humans evolved to the point of having opposable thumbs by then.
In the late Nineties, to stream audio on the internet, you used either Windows Media Encoder or my personal preference, RealAudio. Both methods need two files to work: The file itself, and then a file that acted as a kind of middleman. When you clicked the Play button, you were engaging the middleman file, which pointed to the audio file on your server. Once accessed, it loaded up in your Audio Player, and once it felt like it had enough lead time, it would play. The smaller your audio file was the better.
Having a smaller file was great for another reason: You didn’t have a lot of space for things back then. My first website was on GeoCities, and that let me have 10MB. Megabytes. By today’s standards, that’s microscopic, and yet we managed to get audio entertainment up on the internet long before podcasting, and long before anyone really cared much about audio quality. As a result, the bit rate on those early sound files was usually Mono 16 bit 22kHz files, and they’re pretty horrible. You won’t be listening to Dark Side of the Moon at that quality, and if you are I seriously question your life choices.
Today, we don’t have these problems. Definitely not one of length or server space if a podcaster like Joe Rogan can go three hours regularly, and not one of quality if you’ve heard anything from the gold standard of production quality, NPR. That said, if you’re just starting out, no one is expecting you to produce anything near the quality of Serial. Most independent podcasters are an army of one, while these big podcast outfits have the staff to dedicate to production. Most people who evangelize about podcasting will tell you that the message is more important than the production quality if you’re connecting with an audience. As time goes on and you gain experience, the desire to make your show sound better is perfectly natural. I have found that this is a stress point for some beginners because while they might have a decent audio editing program, they’re not quite sure what to do with it beyond recording. This is compounded somewhat by people who would like to make this sound complicated so they can sell you their solution. The good news is, this really isn’t complicated at all.
I think there are seven steps to making sure you have decent podcast audio.
STFU or GTFO
Simply put, if it’s in your recording space and it makes noise, quiet it or remove it. Clinking glasses, pencils or fingers tapping on tables, pets, kids, you name it. If it cannot be made to be quiet, it should be made to leave. The less noise you have to deal with during recording, the less cleanup you’re going to have to do later. Remember: “We’ll fix it in post” is a pipe dream, and sometimes you just can’t fix it.
Stop That Damn Racket
Some noise can’t be quieted and can’t leave. Air conditioning, for example. It’s also possible that the ambient noise in the recording space can’t be quieted much at all. It’s OK. Most of the Digital Audio Workstations (DAWs) available today can capture a ‘noise print’ from the file, and remove as much of that noise as possible. When you record, make sure you have at least five to 10 seconds of silence before anyone speaks so you can capture that noise print. I also like to leave five to ten seconds at the end of the recording. You can also use a noise gate that quiets the input if the signal falls below a certain level. If you have a budget, you can purchase DAW plugins or separate programs that will analyze the sound file and attempt to correct these automatically. Once you’ve removed most of the general noise, make sure you listen to the file and see if there is any other noise that needs to be removed manually.
Sometimes, you can’t remove a noise because you or some other talent is talking over it. In that case, I would assess whether you can remove the line and still retain the context and meaning of what the speaker is saying. That brings me to my next step….
They’re Smarter Than They Sound. Really.
This next step is not only good for getting a good sound, but it’s also a great idea if you don’t want you or your guests to sound like idiots. Get rid of the filler words like ‘Um’, ‘Ah’, ‘Er’, If you or your guests stammer through something and trip all over the words, remove it. If you or your guests use crutch words or phrases like ‘you know’, ‘like’, or ‘I’m just saying’, remove them if you can. Personally, I can’t stand the relatively new trend of people starting an answer with the word ‘So’. If I can remove it, I do. Shorten long pauses. Remove audible breaths. Finally, as you edit, you may become familiar with the talent’s pacing. Do your best to match that pacing throughout. The talent will sound much more professional, and if it’s a guest you have increased the odds of having that guest back.
The Audience Can’t Listen If They’re Deaf
I’m sure everyone has had this experience watching TV or listening to radio or podcast: The ads are much louder than the content, and you break speed records to wrench the volume down before you bleed from the ears.
Dialogue, music, remote interviews, and sound effects are sound files that come from a myriad of sources, and almost none of them are recorded or exported in the same way. Also, you may have spots within your own dialog where your volume goes up and down while speaking. It’s your job to make sure you don’t leave your audience riding the volume knob to hear you while running the risk of leaving them with permanent hearing loss at any moment. DAWs can do this through the Normalization function, while websites like Auphonic can do this automatically.
Aren’t You A Little Tweaker?
Next, we need to make the talent’s voice sound good, and to do this properly, I need to tell you a hard truth: Every voice is different, so the things that make one voice sound good, will not make every voice sound good. Another hard truth is that more the most part, no one sounds quite like they think they do. I think most people believe their voice is deeper than it really is because it sounds that way to us. When you play a person’s voice back to them, it’s entirely possible they don’t recognize it. In some cases, people will tell you upfront they don’t like hearing themselves, and that’s perfectly natural. Using equalization, this is very much a trial and error process where you’ll be tweaking the low end, midrange, and high end until you hit a sweet spot. Since every voice is different, it’s simply a learning process. For example, you don’t want to add more low end to someone who already has a deep voice, instead, you may play with the mids and highs to add some crispness and brightness to the voice. In some cases, you may decide to add some light compression, but I don’t think it’s always appropriate to do so. Your results may vary, but the only way to learn this is to do it.
Amalgamate The Fecal Matter
Now, we’re ready to assemble all your resources and add them to your multitrack space. Intros, outros, music, dialogue, sound effects, and anything else you need for your production. Lay them out on your timeline, and adjust the volume so that everything sounds natural and has the right flow. Nothing is drowned out, nothing is overpowering.
Are You F**king Listening?
Finally, listen to your whole timeline in real-time and correct any mistakes you haven’t caught. Once you’re satisfied, export your file at -16 LUFS. This is the standard for loudness with podcasts and ensures that your podcast will sound consistent with the loudness of the majority of podcasts out there today.
You won’t get it right the first time, but that’s ok! By following these steps to getting good podcast audio, you’ll increase the chances of having the kind of consistent high-quality podcast you want, and that listeners appreciate and share with others.