Podcasting: Ditching Skype For Mumble

In the last year, I’ve been a guest on a few different podcasts and I’ve been hosting my own, The Jukebox. During this time, I’ve podcasted with collaborators using an array of applications: Skype, Blab, Google Hangouts, and Team Viewer - to name a few. With each of these, however, technical difficulties came in spades. Even Skype, the consumer standard, is often bogged down by high-latency and technical difficulties. So, I set out to find a viable, quality alternative.

Before you continue reading, take into account whether you need video for your show. If you do, this solution won’t work. I’d recommend getting one of the aforementioned applications to work. My proposal, outlined below, is to use Mumble, which is an exceptional option for audio podcasting. I’ve done some thorough testing, though, so let me run you through it.

(Test audio can be found near the bottom of the article.)

First of all, what is Mumble? Mumble has become the popular alternative to Teamspeak and Ventrilo in the gaming community due to its low-latency performance, high quality codec, and general ease of use. It doesn’t tax your system like Skype does, either, which is nice. (Even if you’re just using Skype for audio.) Now, you do need a Mumble server. They are incredibly cheap, however, and I have one for about $5 a month that suits my needs.

(Note: there are plenty of public servers accessible on Mumble, too, which usually work just great. But… you don’t look all that professional to your guests when you’re hopping into an empty gaming room instead of your own dedicated channel for your show/network.)

Right, so let’s dig into logistics. Mumble is available to download on Windows, Mac, and Linux, and uses one uniform codec. Thus, your guests have to do a lot less. All they need to install Mumble, and set their input and output - just like they would on Skype or Hangouts. If you’re a podcaster, you probably want both to be your interface or USB mixer.

Now, Mumble defaults to ‘voice activity,’ meaning the microphone turns on when you speak. That’s no good for podcasting - the mic needs to always be on. Hence, all you have to do is go into the settings, pictured below, and set the ‘transmit’ option to continuous. Now, the mic is always on. Also, set the quality slider all the way to the right.

Now your microphone is acting as it would on Skype or Hangouts. If you need to mute it, just tap the mute key at the top - as you would in either of those two. There’s also a chat accessible to type messages or leave links. (I do that a lot on podcasts if I’m sending a guest something to pull up or read.)

Okay, so you’ve got your guests into Mumble. Your microphones are set; you’re all talking. How do you record the show? You hit the record button! Yes, plain and simple - the big red record button at the top. It’ll then open a dialogue that will display how long you’re recording for - a useful thing to have on screen, obviously. You have the option to record out to WAV, FLAC, or a few other formats. I use WAV.

Here’s the best part of Mumble recording - you can change the ‘mode’ to record in either downmix or multichannel. Downmix will, as you would expect, mix your audio down into a stereo WAV file. Multichannel will actually provide you with the stereo WAV files for each person in the conversation. This is immensely useful for a variety of reasons in post-production. Hell, I’d suggest having it on just in case something happens on one of the mics - noise, static, etc.

Once you’re all done, hit stop and your recording will output to your target directory. If you recorded WAV, you’re going to get it at 1058kbps, which is great. Now, just hop on over into your digital audio workstation of choice, drag your files in, do any post-production you need to do, and compress it down for podcast consumption. (My preference is Audition. You can use something as simple as Audacity for this with no issues, too.)

The bit rate of a WAV export.

I gauged the sound quality utilizing this Mumble method against Skype, Hangouts, Team Viewer, Teamspeak, and Ventrilo. (I’ve had non-stop technical difficulties with Blab.) The files I’m producing with Mumble are the best of the bunch, and I have the most options - especially because of the multichannel option. (Just throwing it out there: Ventrilo’s codecs are awful, and produced the worst of the tests.)

There’s a reason gamers use Mumble. It’s reliable. It doesn’t crash, it doesn’t tax your system, and audio problems are non-issue in comparison to Skype or other popular clients. Again, this is irrelevant if you’re recording a video podcast. If you’re recording audio, however, this is a very viable option. I think it’s the best, and I’ll be using it for every show I do.

(I’ve even created a simple tutorial to help my guests install Mumble if they don’t know how. Mumble guides you through the process very simply, though. It’s not much different than setting up Skype, minus the video capability.)

Over on Timo Hetzel’s blog, he came to this realization a whole lot earlier than me - like four years ago. “After ten episodes, combined runtime: 30 hours, we’re quite confidently using Mumble a vastly superior alternative to Skype for our purposes," he wrote. "We enjoy much more lively conversation without anyone cutting out, reduced latency, and increased audio quality. Skype is demoted to providing us with a moving image.”

(Audio test above. That was outputted as WAV and then converted to 320 MP3.)

So there you have it. Mumble: the audio podcaster’s best friend. Oh, and before this gets shared on Reddit and I get questioned - I’m not sponsored by Mumble or any of their affiliate hosts. I was on an endeavor to find a way to record the best audio podcasts, so I used all the aforementioned applications. Mission complete!

Oh - and pro tip. Click 'configure' at the top and uncheck 'text-to-speech' right away. That'll stop Mumble from announcing each person that comes and goes from the call.