This is a continuing series in which I interview great podcasters to learn about their podcasting setups. While the content is always the most important aspect of a podcast, the technical craft in bringing that content to the listeners also deserves attention. I hope this series will illuminate that critical piece of the puzzle.
Erik Hess is a man of many talents, from podcasting to web design to flying fighter jets. Yes, fighter jets.
What podcasts do you host?
We release an episode of Technical Difficulties once a week. One episode per month, we release an hour-long show where we chat with someone about their area of expertise. The other weeks we spend about thirty minutes covering one tech-oriented topic in detail. If fitting a lot of detail into thirty minutes seems like a difficult task, it usually is.
We make up for our short air-time by posting show notes that depart strongly from the traditional, context-free blizzard of links. Our notes provide time-stamped topic headers, asides with additional depth on difficult subjects, and extended commentary that would be too lengthy (and probably too boring) if we put it on-air.
We owe the new format to our silent co-host Potatowire, who wanted to create notes that could stand on their own, and in some ways overshadow the audio recording. They’re an experiment that’s still evolving week-by-week, but so far we feel they’ve been a great success.
What’s your physical rig? (Computer, Mic, headphones, other accessories.)*
My recording and editing machine is a 2012 Mac Mini with a Quad-Core i7 that I bumped up last year to 16 GB of RAM and a 480 GB SSD. There’s a 6 TB LaCie 2big Thunderbolt drive plugged in for extra storage, which comes in handy for uncompressed audio and video files.
As far as audio equipment is concerned, I talk into a Blue Yeti in cardioid mode on a Radius shock mount. That rig and a Nady MPF-6 pop filter are suspended from a Heil PL-2T boom mounted on the back of my desk. I wear a pair of Audio Technica ATH-M50 headphones which sound great and are extremely comfortable. They connect through the Yeti’s headphone jack, so it’s all USB from there to the computer.
In an effort to minimize clicking and clacking while recording I use a Logitech K811 as my primary keyboard. It’s comfortable, very quiet, and the backlighting has come in handy during early-morning editing sessions. My mouse is a Logitech G700, but I try to avoid it as much as possible while recording. Instead, I use an Apple Magic Trackpad in tap-to-click mode, which is as close to silent as you can get. When editing I revert to the mouse, but the trackpad is still handy for scrolling left-to-right through long audio tracks in GarageBand.
I used to use an iPad mini for mid-show research and communication, and while it was quiet, it ended up being more cumbersome than using a conventional keyboard and trackpad.
What type of room do you record in?
I record in my home office, which is a bit problematic from a sonic perspective. It’s uncarpeted and there’s not much on the walls, so it ends up being a pretty loud space. The Yeti’s directionality helps a bit, and I’ve put a rug on the floor. The shutters on the windows probably help as well. We’ve also got a big dog who likes to bark. You can hear him on a few recordings when I wasn’t able to edit him out.
Overall my office is a nice space to work, it’s just not ideal for recording.
What software do you use for recording and editing?
We start with Skype and eCamm Call Recorder. It’s dead-simple to use and has proven extremely reliable in practice. Gabe and I each record the show to make sure we have a backup, but it’s rarely needed.
Call Recorder dumps the episode into a stereo .mov file, and we use their bundled tools to split the sides of the conversation and turn them into separate, uncompressed mono tracks. From there I import them into an older version of Apple GarageBand for editing.
Thanks to a tip from Ben Alexander of Fiat Lux, we’re trying out a new web-based post-production service called Auphonic that automates a lot of the audio tweaks I used to do manually (and poorly) in GarageBand. Auphonic has got a lot of features and is worth a look for any podcaster who isn’t a highly-confident audio professional.
What do you use to host your podcasts online?
Hosting and RSS have been a big challenge for us. Our ideal service would support direct-URL file download, a rich iTunes-compatible RSS feed, excellent stats, some social features, and a top-notch cross-platform web audio player.
That platform doesn’t exist, so we’re currently using a half-broken hodgepodge of services to get as close as we can. As you’d imagine, that approach increases our workload and has often challenged our listeners’ patience. As a result, I’m not recommending our current “solution” to anybody until we can find something that’s simple and stable.
I’ll start where most of our new listeners do, and that’s with iTunes. We struggled for nearly two months to get our standards-compliant RSS feed accepted to the iTunes directory and finally gave up, as Apple support, and even the iTunes engineering team had no idea why our feed was constantly being rejected by their system. In the end we pointed iTunes to our Soundcloud beta feed so we would show up in podcast apps without our listeners having to manually add the URL.
Soundcloud is an essential and occasionally frustrating part of our workflow. On the plus side, it has a great embedded web player, a large community of listeners, and excellent sharing options. It also lets us link directly to time stamps in our show notes, a key element of our section headers and fancy pull-quotes.
Once we got into Soundcloud’s podcasting beta program we had access to an iTunes-compatible RSS feed. As a result, we have them to thank for finally getting our show out in front of most audiences.
On the minus side, they don’t offer a copy-paste compatible direct download URL for their tracks. This is pretty much the only thing holding us back from using them exclusively. We have lots of listeners who prefer Huffduffer or just like downloading and listening to podcasts manually, and we’d prefer a non-hacky way of enabling that.
To provide that missing direct-download capability, we use Buzzsprout. They were our file host for Generational, and we’ve been very happy with their price and ease-of-use. We just kept our account going when we transferred to the new show and everything has been working fine.
Until very recently Soundcloud’s podcasting beta didn’t support full HTML (links, lists, etc.) in the show notes field, which meant no links. For a notes-focused podcast that wasn’t going to hack it, so we offered a separate RSS feed just for the notes. That gap in Soundcloud’s capability has now been fixed, but those RSS feeds will have stay up pretty much indefinitely.
What’s your basic workflow for recording a podcast and taking it to the published stage?
We start by collaborating on a list of potential topics and guests in Google Drive. We narrow them down and prioritize the list as a team, then try to select the next show with enough lead time that we have an opportunity for research. Recording day usually goes pretty quickly, and we do our best to knock out a couple of episodes at once so we stay ahead of our release calendar. Some weeks that works well, and some weeks it doesn’t.
Gabe usually creates the show outline as an iThoughts mind-map or a markdown document, and we use that to shape our conversation during the recording. As we work through the outline, Gabe and I keep a back-channel open through iMessage. We’ll probably be moving that over to Slack for our next recording session, since we already conduct the vast majority of our other communications there.
Once the show is saved to disk, I start editing in GarageBand. The show uses a Statamic-based CMS we extended ourselves, and all episodes are saved as markdown files with extensive YAML front matter and custom template snippets for the fancier elements. We don’t take a lot of notes during the show (I’m easily distracted) so I create the initial core of our show notes while I edit, roughing in the general structure, key links, pull-quotes, and time stamps. Our header images are taken from the incredibly rich open-access vaults of Flickr Commons.
Once the editing and rough draft of the notes are complete, I upload the audio files to Soundcloud as a private recording (so it doesn’t go out over iTunes) and push the show notes shell via github to our staging server. From there, Potatowire digs in and adds the detail, backstory and helpful asides that make our show notes really shine. Gabe follows up with an editing pass, adding his own flavor as well as any critical elements we may have missed.
On release day I do a final once-over, upload to Buzzsprout, commit the changes to the show notes, and deploy the site to our production server, which hosted on Webfaction. Immediately thereafter I make the Soundcloud file public, which updates iTunes and sends the episode out to our listeners’ podcatchers. Finally, we tell the world it’s live by sending the episode to Huffduffer and posting it to our Twitter account.
Would you like to change anything about your current podcasting setup?
In general I’m quite happy about how our setup has evolved, but there are a few places we could still improve.
The CMS is growing and changing every episode as we extend it to handle new or troublesome situations. That sort of regular tweaking will probably continue as the show develops, and if it stops then that’s likely because we’re not doing the show anymore.
Our single-machine all-digital workflow means we currently can’t break our guest’s recordings out onto a separate track for editing without the perils of double-ending. While our new show has cut back on guest appearances significantly, that’s still the next major hardware challenge I’d like to overcome.
Unfortunately that will probably be preempted by our next major software challenge. Since our GarageBand ’11-based workflow is already obsolescent and there’s been no indication that the latest iteration of GarageBand will restore the lost podcasting functionality, I’ve been looking at some other, more capable alternatives like Logic Pro X or Adobe Audition CC. This will likely happen sooner rather than later, but first I’ve got to find the time to learn the new software.