Friday, August 27, 2010

A Little TWiST Tech!


TWiST #9 is in the bag, and we are just waiting for it to be posted (looks like we are going to be going with a Friday morning posting; at least that's been the standard the last few weeks). Having done a few of these, I thought it might be kinda' fun to introduce you to the side of podcasting that's not often seen, the technical side.

Podcasts can vary radically in form and flavor. Some podcasts are absolutely bone-simple. Rex Black's podcasts are really just him and a microphone; no production, no post editing (or very little); what you hear is how it happened and what you hear is what you get. Note, I mean absolutely no disrespect whatsoever by that; Rex provides solid information in his podcasts and they are very focused. They also happen to be some of my favorite "repeat listen" technical podcasts. On the opposite end of the spectrum is the production value put into shows like Scott Hanselman's "Hanselminutes" and Dan Carlin's "Common Sense" and "Hardcore History". Truth be told, I look to these two as somewhat of the "gold standard" when it comes to podcasting, and they are the ones that I want to try to model the TWiST podcast after (at least as far as production values; I think if TWiST started talking about Ancient Rome or Neo-Prudentism, we'd lose a few of you (LOL!) ).

So what does it take to produce a podcast? A month ago, I would not have been able to answer that question. Now, however, I have some ideas and some enhancements that are being worked in to make the job a little easier and move faster, so with that...


  1. Subjects: Matt pretty much takes care of the real content of the shows. He determines who he is going to interview and then sets up the calls/interviews.
  2. Clean Recordings: Again, most of this right now is in Matt's domain. I don't do too much with this. Matt sets up the initial recording, makes the call, and captures the conversation. Sometimes he works with an in-person interview (as was the case with Selena Delasie), but most of the time it's a scheduled call via Skype with the participant. Both are compiled into an audio file and the audio file is posted to a site for massaging and review.
  3. A trusty PC with good audio editing software: yeah, I know, there are a lot of awesome and amazing programs that are available for the Mac, but I've been a contrarian just about my entire adult life… I've used a PC for music production ever since I made my first software sequencer nearly 20 years ago, and deliberately did it with a PC, so I've stubbornly kept at it. Today, the choices for doing quality audio for PC are staggering, and the really cool thing is how many options are totally free. For me, the editing tool of choice is Audacity.
  4. An "Audio Bed", which basically means a structure for creating podcast formats and the ability to fly-in audio pieces. I do this so that the show has a predictable structure, and also so that I can time elements in the podcast.
  5. "Bumps" or audio pieces that allow the show to have transitions, to cue when the show is about to end, or to allow for a message to be inserted when it might be relevant.
A Matt and I have experimented with the show, we have discovered some interesting things, some of which are items that I deal with on the technical "after the fact" end. The biggest challenge is that the volume of the participants can vary. As an example, when I did a level check on the participants on one call, the volume levels were all over the map, even though during the call we all sounded like we were at the same level. Short of having a control booth with everyone's call being routed to a different channel in a mixing board, this is something that I handle post production, where I go in and separate the audio into different tracks, scrub the tracks for audio artifacts where possible, and then level the volumes so that they are close to even.

Another issue that varies with those being interviewed is the challenge of preserving speech and dialogue "ticks", or smoothing them out to have the flow sound more professional. I tend to delete the obvious "um's" and "ah's", especially when there's dead air surrounding them, but some shows go beyond that and cut when people stutter or stammer words. I draw a line between doing editing for format's sake and doing a total clean-up to where the person in question has all of their personality ripped out. As for me, I rarely um or ah, but I do stammer and repeat at times. If it's distracting, I'll edit it out, but if it's just part of the delivery and feels natural, I'll often leave little things like that in (again, it helps show the personality of the subject).

One of the recent experiments that we have been looking at is the 'forum" or "multi-call-in" podcast, where many people are on the same call. This has it's own challenges, because everyone is sharing the same pipe for bandwidth, and the more people are on the call, the more sonic artifacts get introduced. Some of these can be cleaned, but the more people on the line, the fewer "dead" gaps there are that this can be applied. We recently discovered that there was a "slap back" effect when people on calls were not using headsets, but using their audio speakers to listen in on the calls. That can be a real challenge to "clean" because, again, the best place to make those mods is in the dead space, and those echoes and pops take up not just the dead space, but the live conversation as well.

From a testing perspective, one of the things I love about doing this is that I constantly have to learn new techniques to gain an efficiency or speed up the process. There's an absolute deadline; the podcast has to be posted at a particular time (well, OK it doesn't have to be; I'm sure if STP wanted to move it, they certainly could :) ), but it helps me to know that I have a limited time to do any tweaks or make any show development changes or create process notes. That means each week I get a chance to try something to see if it will help or hinder the process. Those things that improve the process, I take notes on and continue to do. Those that don't, I also take notes on and tend to avoid. What's really helpful (and sometimes stressful) is that there's a hard sense of the time requirement for each episode, so I can't dilly too long on getting things "perfect", but I can learn how to get closer to that point for the next episode.

So there you go, a glimpse into my current world of podcast production. I'm still a bit of a greenhorn with this, but the good news is, I'm getting better each week (well, I think I am; in any event… "if you have a different opinion, I'd love to hear it ;)!"). Oh and don't worry, we're not going to be asking for "a buck a show", and thankfully, I don't get sick all that often* (LOL!).




\* This comes from the podcast "Common Sense With Dan Carlin"… his producer is always referred to as the "alleged producer Ben" who may exist (or may not ;) ), and is a running gag in each of the shows with their middle bump being the "Ben is Sick" routine, where they make their pitch for "a buck a show". Actually, some of the time, the things Dan comes up with are pretty funny, and many times the "Ben is Sick" bit can eclipse the rest of the show (LOL!).
Post a Comment