Justin Lacey in audio, podcasting

Cured From Podcast Audio Separation Anxiety

I started an editorial website 7 years ago to motivate a friend and I to "write more." Thanks to the way that audiences have changed and what it takes to create content with groups of people across various locations, I had no idea that I would end up spending more of my time (and money) on video and audio production. I personally enjoy the creative problems that are involved with producing various types of shows and content, but it's not something I'm professionally invested in. I'm still just learning as I go. I thought by this point I had my bases covered, but I recently had a somewhat obvious revelation.

It has taken me close to 750 podcast episodes, ~300 videos, and thousands of hours of production work to realize - when I have different audio sources, I should record them into separate audio tracks.

Duh. I'm going to take another moment here to laugh at and collect myself.

Ok, I'm good now. You good?

But how? How did it take me this long to notice something so obvious? More importantly, what exactly has using multiple audio tracks made better? Well, if you didn’t leave with a smug look on your face by now, we can break it down a bit.

Wav files

How bad habits are formed

Prior to unwittingly embarking on the multimedia extravaganza that is an editorial website, I had zero audio production experience. I had, however, cut several amazing ringtones and countless custom sound packages for multiple versions of Windows. I would never come close to qualifying myself as an audiophile and I couldn't find a lossless file out of a listening party lineup. I am telling you though, that after I made these small changes to my production setup, I could hear a night and day difference in quality between my oldest content and newest content.

I stumbled into podcasting. It’s not like I was ever planning on producing something on the level of a rock opera. My original goal was to just produce a small, weekly audio podcast. At the beginning, I did some research and noticed most of my favorite shows (in 2009 mind you) were exported as 64 kbps MP3s which gave me an initial quality target to shoot for. The real problem I was dealing with was just getting the show off the ground with the most basic hardware I could find. To this day the website is a hobby so my priorities still remain focused on producing a show quickly and efficiently over spending a ton of time and money on qualitative details. After all the quality comes from the content itself, not in how it was made. That’s partially true.

Should have gone ham radio

Listen to a little taste of this nightmare from year one:

We all have to start somewhere I suppose. That was a bunch of amateurs recording on Skype using microphones from webcams and headsets. Woof. Thankfully it didn’t take me long to upgrade microphones (shout out to my Blue Snowball) so it no longer sounded like I was recording from a bathroom across an empty metal hallway. Better microphones did not fix the primary audio input issue here though:

The application I was using to record the Skype audio mixed all of the tracks together into a single track. ALL of them. /shudders

Since the soundchecks in our earliest shows were just a guessing game, the volume levels of the remote guests compared to my echoey host audio tended to vary wildly. I didn’t take the time to care. I was focused on making content and scheduling others to make content, but our recordings "technically worked." It got us off the ground and opened us up to new podcast possibilities as livestreaming was right around the corner.

We’ll do it live

Surprisingly, the first real improvement to our audio quality came about because we wanted to start producing live video content in the form of a charity marathon. This change forced some hardware upgrades and the addition of a producer to oversee the behind the scenes details while I hosted. We ended up recording our first live video podcast as a promotion for the planned marathon show. The live rapport made for a much better recording experience. From that point on, I made every effort to have local guests rather than remote guests over video conferencing.

The side benefit of this experience was that it changed the way we recorded our audio. With multiple local guests we started using a USB audio mixer that allowed us to adjust/monitor audio levels for each individual microphone. This result would all be mixed into one audio track while all remote guests were mixed into a separate audio track. We could then balance the volume levels of these two (as in more than one!) tracks in our live streaming/video capture software. (Thanks, Wirecast and OBS)

Our first step forward:

It was a relatively small upgrade, but it introduced me to the value of the options that come with more sophisticated hardware and software. At the time, I was primarily concerned with our live show’s audience and that meant we output a single track of audio content for them that we repurposed for the podcast feed later. I figured this was as good as I could do for the money. I was happy with the improvements and wouldn’t make any adjustments for hundreds of episodes.

Live show

I have returned with knowledge

After years of live shows from one location, eventually cast member schedules dictated that we needed to go back to prerecorded shows from multiple locations. It felt like a downgrade, but as I settled into a new routine I was able to take what I had learned from our live shows and apply some of those techniques to recording remotely as a group. I didn’t want to lower the bar for our audio quality so I convinced guests to purchase their own (semi-)professional microphones instead of going back to using gaming headsets. The video software from our live shows also allowed me to continue to use a two track recording (local and remote) which was better than going all the way back to a single track skype recording. I was very content with this solution, but every now and then its limitations would bug me:

  1. Remote guest audio was limited to whatever was received by the host computer. This means that audio glitches (robot-voice) and connection issues were regular problems. Sometimes we lost or had to rerecord entire shows/segments. Other times the audio compression quality would vary between guests, especially when compared to the clarity of the locally recorded host audio.

  2. Video conferencing software tends to prioritize the person speaking over guests trying to jump in to add comments. Therefore, when multiple people were talking at once some cast members were muffled.

We were this close?

When you lose enough episodes to connection issues, you start to look for backup plans. I started asking a few of my guests to record the episode on their end so I could use their single track audio as a backup if something went wrong with my setup. Ironically, after we started doing this I never had to use any of the backups so I mostly forgot about it. Things were going well, but the stage was set for the real game changer.

Live show 2

In the past year, Ryan Billingsley from our collective here, joined my podcast and was able to come by do some recordings with me in the same room. The shows went well and eventually he upgraded his home setup so that he could join in on remote episodes. We got to talking about audio quality and during one of his first remote shows with Coop (also from the collective) he threw out the idea of all of us recording our own audio. I just assumed he meant a single track mix to be used as a backup like we had been doing.

As we continued to talk I realized he meant that he’d be recording just his local microphone audio, and Coop would be doing the same. They would then deliver the two files to me so that I could mix it together with my own microphone audio after we were done recording. It sounded great, but I was worried that syncing up my audio track with theirs would be a pain.

Before shooting down the idea because “change is hard,” I took a moment to actually dig into both my audio and video software for the first time in years. It quickly became obvious that I had what I needed to export multiple stand-alone audio tracks alongside my standard single track mix. Using that mix as a reference, it would be easy to sync the other stand-alone tracks in post-production. It felt like I had learned to use podcasting software all those years ago, but I failed to read the last page of instructions. The results when compared to old episodes speak for themselves:

2016 Audio Example:

vs. 2017 Episode:

Everyone sounded clear. You could even here the smaller, colorful comments people would make as others were talking. It added a lot of flavor and quality to the episodes. I was thrilled.

A podcast process

Here’s my current solution for recording a podcast with multiple remote guests.

  • Locally, I record my host audio in Audacity as a .wav file.
  • Locally, I record a 3 track video in OBS Studio that I later strip to .wav files using Adobe Premiere.
    • Track 1 is my host audio (serves as a backup to my Audacity file)
    • Track 2 is a single track of all of the guest audio (serves as a backup to their recording)
    • Track 3 is a single track of my host audio combined with the guest audio (used to sync guest audio tracks in post production)
  • Remotely, each guests records their audio track in Audacity as a .wav file.
  • After the show, each guest uploads their audio track to a shared Google Drive folder.
  • Using Adobe Premiere, I export the audio tracks from my video file.
  • Using Levelator, I normalize the volume levels in all tracks.
  • Using Audacity, I import all of the audio tracks and sync them up referencing the combined track.
  • Finally, I export the final file as a 128 kbps MP3, tag it appropriately, upload and publish via Soundcloud.

Final Live Show

My final solution does involve a few more steps for all involved when compared to how we started. However, after hearing the results I have no issues requesting that guests record and provide their own audio. There are some solutions that handle all of this for you like zencastr.com, but this is working for us now and it saves me from another monthly subscription fee.

7 years in and I’m still learning things. I like that. Hopefully, my process above will help a few podcasters out there. I look forward to reading new ideas that I can use to continue to improve my own shows over the next 7.


comments powered by Disqus