Podcasting

If you like to talk and have something interesting to say, you may enjoy podcasting. A podcast is simply an audio file in MP3 or Ogg format that people can download and listen to on their own schedule.

A few years ago, I created a podcast called mintCast, which was designed to help new users get started with Linux. I had been listening to several podcasts, but I knew nothing about producing one myself. This article summarizes what I learned during the following two years. It should save you some time and pain, if you decide to create a podcast.

Entire books have been written about podcasting, and you should read some of them if you want to study the subject in depth. To keep this article to a manageable length, I focus on what I learned while producing mintCast.

The first step is to choose your topic and target audience. If your topic is too broad or too narrow, it will be hard to build an audience.

Going Linux has a great example. The home page describes the target audience (“New to Linux, upgrading from Windows to Linux, or just thinking about moving to Linux?”) and what it provides (“practical, day-to-day advice on how to use Linux and its applications”). Potential listeners can see at a glance if this is a podcast they want to listen to.

A podcast episode usually has multiple segments. Typical segments might include an introduction, personal news, industry news, opinions and commentary, interviews, tutorials, tips, recommendations of software or Web sites, and listener feedback. Whichever segments you decide to use, it's a good idea to stay consistent.

That doesn't mean you need to be absolutely rigid. A typical episode of mintCast would have these segments: introduction, personal news, Linux news, main topic, Web site recommendation and listener feedback. The main topic would be a tutorial, interview or discussion about a new distro release. This approach gave us a consistent format, with enough variety to keep it fresh and interesting.

You should write an outline before you start recording. Include the topics you want to talk about and the points you want to make. If your show has multiple hosts, Google Documents is an effective way to share your notes, because all the hosts can read and even edit the document.

Do not write a script of exactly what you want to say, and do not read directly from your notes. It's very obvious to listeners when you do this, and it sounds bad. It's acceptable to read a brief quotation, for example, when you're reading from a news item. Otherwise, just write out the points you want to make, and use the list only as a reminder as you're speaking.

Will you work alone, or with other hosts? I always found it easier to record a show with one or two other hosts. Having additional hosts means that you don't have to know everything, and you don't have to do all the talking.

In the first few episodes of mintCast, each host recorded separately, then we edited the recordings together. In later episodes, we recorded at the same time, allowing us to converse and play off each other. Doing it that way made the shows more interesting, and it was easier for the hosts to keep going. It did, however, add complexity to the process of recording the show.

There are many ways to record a podcast. Sometimes I used a small handheld recorder that saved the recordings in MP3 format and had a USB port that allowed me to copy the files over to my PC.

Most of the time, I used Audacity to record the show. Audacity is an open-source program that can record, edit and play audio files. You can install it from your distro's repositories.

A good quality microphone makes a big difference when you're recording a podcast. It doesn't need to be expensive, but do some research and read reviews before choosing a microphone.

A headset may seem to be a good choice, and that's what I worked with. It causes problems, though—unless you're very disciplined about always keeping your hands away from your head, it's very easy to bump the headset, making an ugly noise on the recording. A microphone boom can help avoid this.

Although Audacity has a lot of power and features, it's very easy to use. The default settings are fine for recording spoken voice. You can tweak the input settings by watching the meter while speaking a few words, then click the big red Record button when you're ready to start.

It becomes more complicated if you have two hosts and you're not sitting in the same room. On mintCast, we solved this problem by talking on the telephone, and each of us recorded our own audio in Audacity. We could then assemble the recordings as side-by-side tracks in Audacity. This gave us a very high recording quality.

When more hosts joined the team, or when we did user interviews, we needed to get three or four people on the call at the same time. We did not have the ability to do conference calls from our personal phones, and we didn't want to ask our interview subjects to try to use Audacity.

We tried Skype and Skype Call Recorder, but never managed to get that working. I know other podcasts use this software, and I certainly would try it again if I were doing a podcast now.

Some podcasts use the Talkshoe Web site. Once you've created a free Talkshoe account, you can create a call series and invite others to the calls. You can allow them to participate fully or just listen. Calls can be recorded and then are available to listen to or download at any time.

You can listen to Talkshoe calls on your computer or telephone. You can participate by calling in from a telephone or by using Skype or Ekiga. Talkshoe also sets up a chat room, allowing listeners on a live show to interact with the hosts.

Talkshoe can be your start-to-finish solution to podcasting. It allows you to promote, broadcast, record and publish your recordings. Why, then, should you bother with any of the other tools and processes in this article?

For mintCast, I could not bring myself to publish our shows without a considerable amount of editing. Each of the hosts had a verbal tic—for example, frequently saying “uh”, “you know” or “basically”—that was too embarrassing to leave in the recording. In Talkshoe, the audio is published exactly as it is recorded.

We also wanted to edit in theme music at the beginning and end of the show. We couldn't do that with Talkshoe.

Still, we often used Talkshoe to record the show. I then would download the recording and do the necessary editing before publishing it to our own Web site.

The mintCast team now is using Mumble to record shows. Mumble is a free and open-source voice chat program. It uses a nonstandard protocol, so you need to set up a Mumble server to which the Mumble clients can connect. With Mumble server and client version 1.2.3, you have the ability to record the Mumble call. You then can load the recording into Audacity for editing.

Before making your first recording, it's a good idea to do a test recording for at least five to ten minutes. Make sure your recording levels are good, and that it sounds like you expect. Listen for background noise—things that you normally don't even notice, like ceiling fans and air conditioners, might interfere.

Even low-quality microphones seem to pick up everything you don't want on the recording. If you're eating, drinking or typing, listeners will know it.

Make sure you know where your mute button is. If your microphone doesn't have a mute button, leave the sound control panel open on your PC and find the mute control. A lot of things can happen that you don't want on the recording, and the mute button can save you a lot of tedious editing.

Don't worry about making a perfect recording. Unless you're using Talkshoe to publish the podcast exactly as recorded, you can edit out most mistakes. We often found ourselves rambling or stumbling over our words, and then we would take a moment to collect our thoughts and then go back and start over at the beginning of the sentence or paragraph. These “do-overs” helped us produce a better show.

Once you have the show recorded, the best tool for editing is Audacity. It's in the repo of every major distro, and it's very easy to use, even if you have no previous experience editing audio.

If you recorded in Audacity, you're ready to start editing. If you recorded with some other tool, Audacity can open files in any audio format. If you have each host's audio in a separate recording, you can add new audio tracks and paste each recording into a separate track.

I also like to add a separate track for music. We had a theme song at the beginning of the show, a longer version of the theme for the end of the show, and a five-second clip from the theme that we used as a bumper between segments. I put all of those clips into the audio track.

Audacity allows you to select part of the audio, then cut or copy and paste it. A scrolling display shows a waveform of your audio for each track, making it easy to select exactly the part of the clip you want to work with. You can zoom in or out on the waveform display. After doing a few shows, I could recognize my “uhs” from the waveform before I even heard them.

If your audio is in multiple tracks, be careful about deleting material. If you delete a clip from one track, you've now changed the timing for the rest of that track, and it will be out of sync with the other tracks. If you delete a clip, be sure to select and delete that time frame from all tracks so that they will stay in sync. Audacity allows you to click and drag across all of the tracks if you need to.

But, if you need to remove unneeded noise from one track while a host is talking on another track, you don't want to delete it. Just select the noise in the waveform, then click the Generate menu and choose Silence. This will silence that part of that track, without affecting that track's timing compared to the other tracks.

Audacity has several useful tools on the Generate and Effect menus. If you want to remove a sound without silencing it, you also can select the clip and Generate a tone, a chirp or a noise that sounds like static.

It was not unusual for our initial recording levels to come out way too low. The Effect menu includes an Amplify command, allowing us to select an entire track (or any portion of it) and amplify the volume.

If the volume varies too much in different parts of the recording, the Compressor command on the Effect menu can bring the levels closer together. The volume will be increased for the quiet parts and reduced for the loud parts.

If you do have a consistent noise in the background, such as an air conditioner, the Noise Removal command on the Effect menu often can remove it. You need a few seconds of audio where you're not talking, where the only sound is the background noise you want to remove. When you click the Noise Removal command, a dialog box pops up explaining how to use it.

Save often as you work. Although crashes were not frequent, Audacity did crash more than any other program I've used in Linux.

Finally, Audacity can export the finished file to the appropriate audio format. Podcasts are typically MP3, though many also are offered in Ogg format.

Remember that all those CDs and MP3s you've purchased over the years are copyrighted—you can't use those songs in your show. However, there are sources where you can get music without spending thousands of dollars.

The Podcast Themes Web site has music specifically created for podcasts. A few songs are available for free, and many others are available at a low cost. Various licenses allow you to buy exclusive or non-exclusive rights to the music.

Unique Tracks is another Web site that offers royalty-free music. They offer individual songs and albums in many different styles: classical, uplifting, rock, heavy metal, hip-hop, New Age and many others. Many of the songs include the full-length song as well as shorter versions, and some even include a loop of five to ten seconds, which is ideal for a bumper.

Once you have a completed recording, you need to publish it so that others can download it. If you recorded in Talkshoe, it's already published. Even if you recorded in Talkshoe, you still can publish it to your own site.

Any Web-hosting company will offer a template for publishing a podcast. If your site uses WordPress, you'll need the PodPress plugin. In any case, you'll have a Web location to upload the finished files. Then, create a blog entry on your site and link the audio files.

Although your site probably already offers a feed, you can use Google FeedBurner to get information allowing you to analyze your traffic. Once you're set up in FeedBurner, it's a good idea to submit your podcast to the iTunes store, the Zune Marketplace and Podcast Alley. All of those will make it easier for people to find your podcast.

Once you have listeners, some will give you feedback about the show. It's best if your Web site is set up with a forum where listeners can discuss the show. On simpler sites, they may be able to leave comments only on the blog post. If nothing else, at least get a Gmail account and publish the address.

Read the comments and consider them seriously, especially the negative comments. I know it's painful to read criticism after you've put so much work into the podcast, but those criticisms often can help you make a better podcast. Encourage listeners to write in with comments and suggestions, and thank them for their feedback.

When I first started using Linux, the podcasts I listened to really helped me a lot. I was reading books and magazines, of course, but listening to a friendly voice really helps make a new topic accessible. It's also nice to be able to learn new things while driving to and from work.

If you like to talk and know a thing or two about Linux, why not help others increase their mastery of Linux?