Podcasts: They’re still a thing. A How-to.
It’s interesting to be writing this on Medium, which was started by Ev Williams, one of the founders of Twitter, and formerly of Odeo.
Odeo’s business model was all around podcasting. It didn’t do so well, but out of the failure of their podcasting business, came Twitter, invented by Noah Glass, one of Odeo’s engineers. I worked there for seven years trying to keep the site alive and secure. Podcasts were the furthest thing from my mind then, but now, I publish podcasts with the people over at Wicked Grounds cafe and stream our recording sessions live (Warning: NSFW).
After ten episodes, I think I know how to do this sort of thing.
There are few services today that do what Odeo did (managing online podcast publishing), but there are still many people creating podcasts and they are still an important and useful source of news and debate in our current Youtube obsessed Internet culture. Here’s a quick guide to how I run one of them.
I’ve probably had about ten people ask me “What equipment and software should I buy?” Start with a solid microphone. The warm, rich sounds that you hear on talk shows and NPR come from microphones like the Electrovoice RE-20, Shure-SM7B, and the Neumann BCM705. Most of the time we want to hear a large diaphragm dynamic mic with good rejection of outside noise. If you have a perfect recording studio environment, then you can use large diaphragm condenser mics like NPR does, but without an amazing room the condenser mics will pick up every last detail of your noisy bedroom. (oh, and keep the cats out of the room. They’re meow-y.)
You’re looking for a microphone that sounds good with your voice, so this may take some experimentation. Personally, I use the SM-7Bs. They sound amazing on most people and have variable equalization built in.
You will also need a microphone stand or desk boom. The R0de PSA-1 is a great choice here if you want that radio look and don’t want a stand taking up precious floor space in your room.
Don’t be tempted to use the microphone on your earbuds or talk into your laptop at a distance as this will never capture the richness of your voice. Your audio will be all noise and no signal.
You will also need an audio interface with enough microphone pre-amplifiers to handle your guests. If it’s just you, you could get away with a simple two channel interface such as one of the Focusrite Scarlett interfaces, but I record groups of 4–8 people so I tend to use the Focusrite Saffire Pro 40. It has plenty of inputs and outputs to get the job done. My only complaint with the device is that the Shure SM7B microphones that I have require an immense amount of audio gain (they are dynamic large-diaphram microphones) to work. The Saffire 40 barely provides this and at times I wish I had a stronger mic pre (such as the Focusrite ISA428) to work with.
Additionally, if you are dealing with multiple people, you will want to purchase a headphone amplifier and headphones so you can all hear each other at the same time. I’m currently using a ART HeadAmp 6 for this purpose and it’s excellent. I build a simple headphone mix with the included Saffire Control software for my interface and send it back to the show guests.
There are many discussions on what constitutes good software for editing. I’m on a Mac Pro and I tend to lean towards Pro Tools. The problem with Pro Tools is that while it’s the industry standard, it’s obscenely expensive. I only own it because of a student discount I got while studying music production at Pyramind. Otherwise it would have been close to $1000.
On the cheap, most Macs come with Garageband (although many of the podcasting features have been removed, sadly.) Logic Pro is another excellent, low-cost choice, and many of the aforementioned audio interfaces come with “light” versions of professional software like Ableton Live or Pro Tools.
You could also use the open source audio software, Audacity, to do this, but I seriously dislike it for editing and never use it. A nice how-to lives on their wiki and offers an incredible amount of detail on the publishing process.
So long as the software can record from your interface, and you can use some sort of plugin to apply processing to your sound later on, you should be good here. Use what you are comfortable using and don’t concern yourself too much with the software.
The Recording and Editing Process
I record every person at 24 bit/96 Khz in a separate channel. I record at high resolution so I have headroom to work in. This way, I can isolate specific people and adjust volumes with plenty of room to spare. Let’s not get into the audio resolution debate here — you want headroom, and 24 bit is where it’s at.
On each channel, I apply the following, in this order:
Equalization — To filter out low-end rumble and warm the vocals.
Expander/Gate — To push the noise floor down when the person is not speaking
DeEsser — To eliminate sibilance
Compression — To rein-in the audio levels on the channel
On the master bus, I like to use a maximizer or leveller like Waves L1 to ensure that my mix doesn’t digitally clip and to make sure that we are printing a nice hot signal for publishing. That all gets mixed down to an AIFF file for the next part, when we publish our work.
Once you have your final stereo AIFF or WAV, I like to use LAME to encode an MP3. I am typically producing a 96Kbit or 128Kbit stereo MP3 stream for publishing to iTunes (more on that in a moment) with all of the appropriate tags and EXIF data in place so that when the MP3 plays, the right data shows up in iTunes.
This as it turned out was one of the hardest problems I had to solve. Tagging doesn’t make much sense. I had to write a number of scripts, which I am making available here as open source. These scripts produce the RSS file for one or more podcasting channels, based on a simple JSON index file that you edit once after you make a new show.
If you don’t want to run scripts like I do, there are tools such as Podcast Generator and Wordpress plug-ins that will do the heavy lifting that my scripts do.
In any event, what you’ll end up with is something that generates an RSS feed and a collection of MP3s on a web server, someplace. When a user subscribes to your RSS feed, the client (in most cases, iTunes) will periodically download the RSS file. When the user doesn’t have one of the files in the feed, this will trigger one or more downloads of the actual MP3 files stored on your server (not Apple’s) which are referenced by the feed.
Once you make a podcast and have both your MP3 and RSS file ready on a publicly accessible web server, you’ll need to make sure people know it exists.
Follow Apple’s guidelines, to the letter. Submit the podcast to iTunes, Advertise it on your Twitter, Facebook, and other social media, and Submit the podcast to all of the available directories like Feedburner, Feed Shark and Podcast Alley.
There are many great links back on Audacity’s Wiki for places to promote your podcast as well.
Good luck with your podcasts! May you enlighten the world with your voice.