Attack of the Pod Penguins: Not Why, Why Not?
- 18 August, 2006 14:09
Producing a podcast on Linux is easier than you think, thanks to some versatile tools. Here's an intro to the fundamentals.
It's not that no one thinks of multimedia and Linux in the same frame. Many do. They're the Linux and FOSS evangelists, determined to remove all vestiges of proprietary software from their computing lives and yours, too, if you understand what's best. They're the engineers, fascinated with the challenge of bridging proprietary (and often very secret) hardware to the community of open source users. They're small commercial interests, focused on maximizing Return On Investment by minimizing the cost of entry into a burgeoning multimedia market. And, they're users, dabbling in open source for the first time, wondering how on Earth to make that sound card work like it did in Microsoft Windows.
Plenty of people think of Linux and multimedia in the same loose context. Few actually pair them for any useful purpose. Hardcore use of open source multimedia tools doesn't even approach what it could be in the computing world. With two other venerable computing platforms, it's difficult for some to imagine why anyone would need to use open source multimedia tools. It's not that they think it's impossible, or even particularly difficult. The problem is history. Both Microsoft and Apple have sizeable code bases including easy-to-access hooks into multimedia hardware. On the Windows side, hardware designers create new multimedia toys with that code base in mind. On the Mac side, hardware manufacturing is strictly controlled to conform to the Apple standards. While Macs generally take the upper hand in multimedia, both platforms deliver pretty consistently. It's not tough to see why open source-created multimedia hasn't made it onto the world computing stage. Most can't understand why it's necessary.
I've thought a lot about this in the past few years. Most of my weeknights and nearly all of my weekends have been filled with open source multimedia tools. I qualify as any one of the users noted above at any given time: engineer, evangelist, small business owner, and user. As a user, I rip CDs and listen to streaming blues broadcasts. As evangelist, I believe, heart and soul, that the world would be a better place without proprietary software, that the gains from computing should not be exclusive to those who can afford software. Or hardware. Or the seemingly endless cycles of proprietary upgrades required for both. As engineer, I've hacked my kernel and dreamed of schematics for a new sound card. I've also run a business on open source tools, albeit a small one, but one that measures real profit in triple-digit percentages. In all those roles, I tend to see widespread use of open source tools in multimedia delivery and creation as the next great plateau in wide open source adoption. It's still a destination somewhere up the mountain-the next base camp, as it were-but it is within sight.
So, when Don Marti sent me an e-mail a few weeks ago to tell me of his move to LinuxWorld.com (and to drop a line in the lake, fishing for articles), I jumped. My inner writer, a largely theoretical and speculative creature, has, for nearly eighteen months, been subsumed by practice. I've been obsessed with open source multimedia in the quest for the perfect sound for The Roadhouse Podcast and the IndieFeed Blues channel. That's meant some business, as well, to offset the long hours. I've put the tools to use in politics, as editor and producer of The ClarkCast, a weekly podcast by General Wesley Clark. My coder's heart has been filled with scripts to maximize process efficiency. "Tony," said my writer as I read Don's e-mail, "here's your chance to make me happy and write about multimedia tools in Linux." That writer speaks up so seldom and is such a good friend, it seems the least I can do.
So, we embark from this point on a thorough examination of multimedia tools in Linux. A series of articles lie ahead, ranging from the practical to the frivolous to the pragmatic. We'll dive into both hardware and software. We'll overlay the demands of multimedia on the broad rainbow of Linux distributions. For the engineers, I'll detail the processes in both hardware and software that make the music sing. For the users, we'll focus on application of the tools, from commands and frontends to shortcuts and customization. And, unlike those who wonder why, you'll be one who wonders, like me, why not.
Of course, we have to start with what's grabbed our attention these days. We'd hardly be a publication of any rank if we didn't delve first into the buzz, the flash that's lit our collective fuses; if we didn't discover how to stir some of that buzz ourselves. As the collective of open source publications so often does, and is right to do, we have an obligation to talk about what's hot. More importantly, our higher-level obligation is to convey how you can use open source tools to achieve the same ends as the big boys. At its core, that's more than some grudgingly-accepted publisher's mandate. We share your dream: to do for ourselves what others pay to have done for them.
Few trends in the technology world over the past few years offer more opportunity for DIY satisfaction than podcasting. Some estimate that in the two short years since the first widely available podcast, more than 30,000 have sprung from basements and kitchens, commercial studios and mainstream media outlets. Some proclaim the death of terrestrial radio. Others see their growth, especially in the commercial realm, as a complementary service, demanded by the listeners and fed by the advertising dollar. And some, looking beyond the hyperbole that always accompanies such new technologies, see it as both evidence and validation of the new criticality of citizen media. Take those assessments as you will. They're really not the point of the articles to follow.
The point is this. If you have something to say-anything, really-it should be said. Open source multimedia tools exist to help you create that message and to move it to the appropriate, self-selected audience. You don't need to investigate the latest podcasting trialware then pay the price for software that may only achieve half your ends. The hallmark of Free and Open Source Software has always been the ability to utilize a collection of small tools, well made, to achieve a known goal. In podcasting, you can, in fact, have it all. With minimal expense, you can create the message. Utilizing a full kit of open source tools, you can save, edit, tweak, refine, and perfect that message, in context, sound and delivery. If a DIY sound is your preference, it's yours. If you lean toward a more polished production, that's achievable with open source tools, as well. Whether your topic of choice is music, politics, technology, or any other you can imagine, open source tools can make it happen.
Page BreakWe should start at the beginning. In order to provide you with a comprehensive series of articles on podcasting with open source tools, we need to first create that full-function toolkit.
In the big scheme, this toolkit will include both software and hardware. It will include your choice of Linux distribution, applications for capturing your audio, tweaking the sound, encoding to MP3, editing your RSS feed, pushing the finished file to a server, and many other small tasks. On the hardware side, I'll make some recommendations for system hardware, including RAM and sound cards, critical elements of any multimedia system. We'll look at microphones and mixers, the core tools for your podcast. And we'll take a look at some hardware items that you may find a bit more esoteric: units for compression, expansion, limiting, noise gating, and pre-amping and processing your voice. Along this twisted trail, you'll also pick up more than a few useful tidbits of information on the physics of sound and how to use them to your advantage. Your podcast, after all, can be your public face. Whether you choose a slick polished production, or a truly DIY sound, the right toolkit and the knowledge of audio principles will allow you to focus more of your efforts on the creative side.
Let's start with the software-the tools with which you'll have the most interaction in creating your podcast. For the purposes of this article, we'll start with three tools. These are at the heart of my own podcasting experience. We'll save in-depth discussions of the software for full articles on those tools. While I use some other packages in small roles, these are the principals.
Available at audacity.sourceforge.net. We'll stick with the stable version of Audacity, which is version 1.2.4. The download is available in a variety of formats, including RPM, deb and source. If, like me, you're a Ubuntu user, you can apt-get Audacity from the Dapper Drake repositories. Audacity is licensed under the GNU General Public License (GPL).
Audacity will be your workhorse, providing most of the functionality you'll need to record, edit and process your podcast. Even after a year and a half of producing 3-4 podcasts a week with Audacity, it's sometimes hard to believe that it's actually free, open source software. The range of features available in the package is, simply, stunning. It easily rivals its proprietary counterparts on every level. It's also powerfully extensible, with a full range of free plugins available for the most obscure tasks. It's also scriptable. Using the Nyquist programming language, users can create custom plugins for nearly any task or effect. Depending on your version of GTK and your distribution, it may not be the prettiest audio tool around. But, it will handle anything you can throw at it with ease.
Despite the full range of features available in Audacity, we'll focus on only a few. The tools for editing, saving projects, compression, multitracking, analysis, and normalization will be the most useful to you, whether your podcast is based on music or talk. These features should be enough to set you down the path of practical discovery. We'll also talk quite a bit about the physics of sound when discussing Audacity. All in all, we'll take a couple of articles to walk through its most useful features.
Like Audacity, lame is available on SourceForge. From SourceForge, lame is available only as a source package. However, it is available as a deb package for both Debian and Ubuntu. With a bit of research, you may find it in the rpm format, as well. lame is released under the GNU Lesser General Public License (LGPL). The LGPL is a variant of the GPL which allows linking of some libraries into non-free software.
lame also takes us into a bit of a gray area. The MP3 format is proprietary, with licensing simultaneously held by Thomson Multimedia and Fraunhofer. Enforcement of the license, especially to individuals, has been almost non-existent. The MP3 format has, in fact, become the standard for audio delivery in the past several years. While a podcast can be encoded in the free-as-in-freedom Ogg Vorbis format as easily as in the MP3 format, your reach will be very, very limited. For our purposes, we'll assume (based on history) that the risk of royalty demands on your podcast by the license holders will be zero, and will be far outweighed by the reach MP3 encoding will create.
Page BreakOriginally, "lame" stood for "Lame Ain't an MP3 Encoder." The maintainers have faced facts and reality, now heralding it as just that. If you've used any Linux tools for ripping and encoding MP3s, it's possible you've already used lame. It's the core of many GUI-based ripping and encoding programs for Linux. In fact, by pointing Audacity to your lame libraries, you can directly export your podcast as an MP3 file. We won't talk much about that, though, as lame's real utility and power comes from command line customization. lame is easily the fastest MP3 encoder going. It also adds the flexibility to create your files with constant or variable bit rates. It's possible to create the id3 tags on the fly, to identify your file to most players. And, lame uses interesting algorithms for what it's creators call "psycho acoustics" - tweaking the sound of a file to more naturally parallel how your ears and brain interpret sound. That should tell you, as well, that we'll talk a bit more about audio physics as we work through the articles on lame.
normalize-audio is available at http://normalize.nongnu.org/. It's released under the GPL and is available for Linux as a source or deb package.
Normalization is the process of bringing multiple sound files to a standard volume level. normalize-audio performs this task quickly and easily. If you've listened to many podcasts (especially music podcasts) you already understand the value of normalize-audio. It's a fairly common mistake by podcasters to preview completed files using only headphones. Under those conditions, with the sound right on top of your ears, it's difficult to assess whether the narrative between songs, for example, is truly as loud as the music. It's a mistake I've made myself. Without normalization, those podcaster may never realize that their voice parts are much quieter than the music. Aside from the fact that most voice parts are mono, music is much more dynamic, with a greater range of both frequencies and volume than the spoken word. Even multiple spoken tracks may vary in volume. For the seamless enjoyment of your listeners, normalizing the audio in a consistent way is an important piece of the podcast puzzle.
Those are the principal software tools for creating your podcast. As I've said, I do use some others in the production of my own podcasts, but they're all secondary players. We'll touch on them as the need arises and move on.
Download and install these tools before the next "Attack of the Pod Penguins" article. In that article, we'll dive headlong into the first section on Audacity. You'll quickly see why open source tools and podcasting are a perfect match.