Podcasting 101
I just dropped my third podcast on aboutGroovy.com. It is an interview with Jeremy Rayner, one of the earliest committers on the Groovy project. There are enough interesting steps in the production process that other programmer-cum-podcasters might be inspired. Or at the very least, it might answer the inevitable question, "What was he thinking?"
Step #1: Buy a Mac
OK, maybe not, but it certainly eased the process considerably. 'Buying a Mac' is Step #1 for most everything I recommend these days, so you can safely ignore it. My wife doesn't even hear it any more. Some people say "um" or "uh" unconsciously. I say, "Step #1: Buy a Mac. Now, what were we talking about? Oh yeah, could I get onion rings with that instead of fries?"
Step #2: Download Skype
Each interview was done remotely. Scott Hickey, lead on the Groovy/Eclipse plug-in, is based out of Nebraska. Jason Rudolph, Grails committer and author of Getting Started with Grails, is in North Carolina. Jeremy Rayner, longtime Groovy committer and creator of the Groovy Ant tasks, is in the UK.
Skype allows us to talk for free. On occasion we'll pick up some digital static, but by and large the connection has been remarkably clear. This doesn't give you audiophile-quality output, but you can't beat the price. My recommendation is to find a quiet room, have a glass of water nearby, and have a hard ethernet cable connection to the internet. (WiFi is second-best, but you really feel the pain if bandwidth drops even for a minute. Sometimes Skype gets in a bad state and can't recover. The dreaded "Skype Echo" can really ruin a nice podcast.)
Step #3: Buy Rogue Amoeba Audio Hijack Pro
You might as well pick up Fission while you're at it. They sell them in a bundle for $50, and it is well worth the modest investment. (See the next step for details.) Audio Hijack Pro allows you to capture audio from any application running on the Mac. (Yeah, it's Mac-only.) Remember Skype? AHP grabs the audio and saves it to an MP3 file.
For Scott Hickey's interview, I dumped the raw MP3 up on the aboutGroovy server and called it a day. That is, I did that for the second interview. For the first interview, I clicked on the "Hijack" button in the upper left corner. The message said "Hijacker On", and we talked for twenty minutes. Only after we wrapped up did I notice the "Record" button in the upper right corner. [wince] Rookie mistake -- Scott was very gracious and we recorded the second interview right away.
Step #4: Don't fear the merge
When interviewing Jason a couple of weeks later, we inexplicably lost our connection midway through the interview. I didn't want Jason to have to repeat himself, so once we got connected again we just picked up where we left off. I knew that I'd find some piece of software that would allow me to join the two files after the fact. Little did I know that it would be the command prompt.
Unix-weenies are undoubtedly familiar with the "cat" command. You use it to display the contents of a text file ("cat myfile.txt"). While this doesn't sound like a likely candidate for MP3 files, stay tuned (no pun intended). You can also use cat to merge files -- "cat 1.txt 2.txt > merged.txt". Amazingly, you can use "cat" for your MP3 files as well. "cat jason1.mp3 jason2.mp3 > jason.mp3" Whodathunkit?
The one drawback of this is that cat doesn't update the metadata to reflect the new length of the track. Pull it up in iTunes and it will appear to be the same length as the first file in the merge (although the entire file will play without issues). Enter Fission.
Step #5: Don't fear the edit
In the quest for a simple audio editing tool, I found Rogue Amoeba's Fission. It does one thing very well -- opens an MP3 file, allows you to make basic edits, and saves it back to an MP3. (It does all of this in native MP3, BTW. No unencoding/reencoding nonsense...) I pulled up Jason's newly merged interview in Fission. It displays a nice waveform and allows you to play the file back. When you find a spot that you want to edit out, simply highlight the area and choose "cut". Voila. Even if you don't make any changes in the file at all, pulling it up in Fission and immediately saving it back again will fix the metadata so that the correct time appears in your MP3 player.
In my interview with Jeremy, he was fighting a cold. Fission made short work of snipping out all of the coughs and sniffles. I'd imagine that true audio experts might find Fission lacking, but for my bone-simple needs it is a perfect fit.
Step #6: Don't forget the tags
...like I did for the first two podcasts. Fission allows you to add tags for Artist, Album, Title, and the other usual suspects. Save the file again and you are done. Unless...
Step #7: Dude, don't you need a cool theme song?
I'd been looking for an excuse to learn GarageBand, and here it was. I whipped up about 30 seconds of lead-in music and saved it out to MP3. Use the "cat" trick for your bumper music and the Fission hack to correct the times again.
Unfortunately, my time in GarageBand foreshadowed what I'm going to need to do next. Being able to mix the volumes on individual channels is something I missed in my primitive Skype audio dump. Once I start interviewing folks face-to-face, having two microphones will become the next requirement. Knocking my overpowering "trainer voice" down to a reasonable level on par with the other normal person will make a huge difference. In my interview with Jeremy I would love to bring up the levels on his answers, but that would bring my questions to ear-shattering "Is it live or is it Memorex" levels. (In the meantime, maybe I could just ask the other person to shout their answers...)
Podcasting is meant to be a populist activity, and I certainly have the skills to embarrass anyone who has even the slightest bit of professional experience. ("Yeah, I swept the floors at a radio station while I was in high school. Even I know what an EQ is for.") But for right now I'm having fun. Hopefully this will give you to skills to do the same.
Posted on Tue, 13 Mar 2007 08:29 by default (431 day(s) old)
