How We Podcast – The Tape and Tech that Makes it Happen

When we launched First Ring Daily back in 2016, it was always our intent to properly build out all the tech needed to produce the podcast ourselves. The early days were rough as I had no idea what I was doing and launching a daily podcast is a lot more challenging than it may seem which pushed us down the road to call in some outside help from Andrew to make sure we got the content-ball rolling correctly before diving into the tech side of the equation.

On February 1st, we started producing the podcast 100% in-house which has a couple of advantages; flexibility on time when the show is recorded and the ability and to let me have 100% control over the process because I am secretly a control-freak. The downside is more work for myself as I now edit and upload every episode but with the proper workflow, that process now takes all of five minutes to complete.

So how do I make it happen? Well, if you have been watching the show since the 1st, you’ll know that it doesn’t always go smoothly and I have learned a lot…which I will detail in this post.

The technology for the co-host (Paul) is quite simple, he needs a webcam that for some reason is always slightly blurry and we don’t know why, Skype, and a quality mic. On my side, it is significantly more complex.

The software that I use for every episode is xSplit, OBS, Voicemeter Banana, Adobe Premiere Elements, and Skype. All of these pieces of software (minus Premier) work together to create, at the same time, the podcast.

In our original setup, I was not using OBS or Voicemeter, and I’ll explain why those pieces of software are now critical to this production of the show.

xSplit is the software that brings it all together, the video feeds are pumped into this software and it is what I use to record, scene switch, and on Friday’s, push the content to YouTube.

The first problem, audio and video were not in sync. If you watch a couple of the first episodes that I produced, there was an issue with audio and video being out of sync. After each show, I was painfully tweaking each audio feed in Premier to align lips with audio as I had trouble finding the root cause which means that this went on for nearly a week.

The first problem is that bringing Skype video and audio into Xsplit is a dicey experience, the video drops frames and the audio is marginal at best. Even though Xsplit supports natively the ability to rip the Skype video feed, I don’t recommend using it.

Solution: I configured Xsplit to ‘screen capture’ a monitor; what I do is make Paul’s video feed full screen and record the screen capture into Xsplit which works quite well.

Second Problem: Frames per second. On my end, I am using a higher-end Canon camera that captures 1080P at 60FPS while the screen recording is at 30FPS and you can’t force it any higher. When my feed was placed next to Paul’s in Xsplit, it caused audio-sync issues that once you see it, you can’t un-see the issue.

The solution seemed simple, change my camera to 1080P at 30FPS, but because the Blackmagic capture card and Xsplit do not play well together, the feed refused to show up randomly at this setting and required endless closing and re-opening the app to see if it would find the feed that day.

The solution to the FPS problem is to use OBS as it works much better with the Blackmagic card and then once again, screen-record OBS for my feed and then use that as the ‘brad’ cam. The added benefit here is that additional latency of screen recording myself and Paul is exactly the same and the feed into xSplit is also the same; voila – a podcast with no video/audio sync issues, right? Wrong.

Everything was going great until what felt like half the planet kept complaining about audio levels. Either Paul was absurdly loud, I was too quiet, my mic was peaking or the solar winds causing feedback across the channels. This problem was much harder to solve until I realized what the issue was: Skype.

For reasons only known to Microsoft, audio coming across from Skype is not the same each time you make a call. Often, Paul was too hot (loud) and it drowned out my voice. Solution attempt #1 was to adjust Paul’s voice using the volume mixer in Windows 10 but every day when I rebooted my computer, that setting would go back to the default max volume and ruin the audio balance.

Worse, I had no insight into how each episode would sound while recording as I didn’t have local playback of my own voice as I feared it may cause an echo. Double worse, Xsplit’s audio meters are about the size of two hydrogen atoms and there was no way to see them from across the room at the podcast desk. Effectively, each episode was rolling the dice if the audio would be ok for that episode and if not, time to spend an hour in Premier adjusting it.

Solution: Voicemeter Banana – this software has quickly become a favorite of mine; this software is a digital mixer with some awesome configuration settings. What you do is in all of your apps that have a microphone and speaker setting, you route them through the app and you can isolate channels and adjust the volume of each feed; they also have very large volume meters that I can see from the podcast desk.

Setting up Voicemeter is complicated but once you understand the basics of it, you can control all your microphones and audio feeds from one piece of software. If you look at the image above, the Sams Report mic is Input 1, FRD mic is hardware input 2, and Skype audio is hardware input 3.

I have all mic/podcast audio routed to the Aux output (this is channel B2) but I have Paul’s audio also routed to A1. The ‘A’s are where I can hear sound: A1 are PC speakers, A2 is not used, and A3 are my headphones (yes, this makes it possible to have the audio come out of your speakers and headphones at the same time). This is a very rough introduction to Voicemeter and if you do download the software, check on YouTube as there are excellent tutorials on how to configure this software.

To bring the audio into Xsplit, I set the microphone to Aux Input (B2) and the speakers to VAIO – the VAIO is where you hear the intro and outro music play. Xsplit records all audio from both the mic and PC but what I do is after the intro, I mute the VAIO channel; this is important because it records all sounds including email/Skype notifications. By limiting this channel, it makes sure you only hear us talking and not all the sounds caused by various aspects of Windows 10 / software running on my machine.

One thing you need to consider is that running this type of setup is taxing on your hardware. My machine, which you can read about here, is overkill for just two people but roughly speaking, it should support up to 4 video/audio streams without dropping frames. When just Paul and I record, I’m using about 40% of my machines capacity to record the video at 1080p (each file is about 6 gigabytes uncompressed).

This setup is overly-complex but seeing as no one has complained for about two weeks, I think I finally have just about everything figured out and tweakable if something doesn’t go as expected. And if I do screw something up, I can try to fix it in Premier but recording correctly is far easier than fixing in post-production.