How we podcast
I thought it might be useful to let everybody know how we go about making the podcast. Over the last year and a half I have been trying to reduce the effort needed as much as possible, the more effort you need to put into the sooner people burn out. We prepare two shows per week, one news and one that alternates between what we're playing and a topic. This means that the shows that need a lot of preparation (topics) are only every other week and they have more time to be prepared.
The show preperation starts on Sunday night and continues on until Sunday afternoon with news gathering. This is easy as it's just starring news items in Google Reader that may or may not be interesting in the show. No effort is spent yet determining how accurate the stories are as there's still time before we record for other people to do that work for us, and I don't try and work out what's important and what isn't yet either as something new will probably crop up before we record anyway. Friday night is an important time for news as important things tend to break on a Friday it seems. This could be a coincidence, positive reinforcement on my side or people trying to hide news over the weekend.
On Sunday I take a look at the stories I’ve accumulated over the week and make some judgement calls about what is important and what isn’t. I’ll chase up anything I think is dodgy or hasn’t been reported much by anybody else and end up with a bunch of links open in Firefox as tabs. I add that tab group to my bookmarks and call the folder the date for that show. I then use the organise bookmarks feature to change the order of the news so that I’m happy with a running order. At this stage not all items listed may make it on the show, and I’ll usually have more items than I’ll need. During all of this notes are written in a notebook so that I know what I must cover with each item. Because a normal notebook would be too low tech I use a Livescribe Pulse pen. I also keep a notebook next to my computer and console to record notes as I play for the what we’re playing shows.
Come Sunday afternoon we’re ready to record. Tim drives over sometime in the afternoon and generally puts up with watching me play games until we really can’t put off recording something.
Since time and effort are so important the biggest thing we do is record as live. When we do a whole show that needs no edits then that saves a lot of extra work snipping pauses, erms and asides out of the audio. This saves me about two hours post-processing time over when we started, and it also means that the recording of the show itself takes up not much longer than we talk and more time can be spent with playing games.
Our actual recording equipment in the studio is a pair of microphones and a Zoom H4n recorder. We started off using USB microphones and recording directly onto a mac, but there were too many occasions where we lost audio due to application crashes or the USB mics being reduced to just recording static. A small dedicated recorder also has the advantage of being able to be used in the field. The files created are 2 channels with both of us being on each side of a stereo wav file. This is handy because it’s a perfectly standard format and because we’re in the same file there are no synchronisation issues. Recorders that have 3 or more channels will usually use a proprietary format that needs conversion or need to have the audio synchronised in the editing application. Because we’re just sitting at a desk there is bleed through of both of our voices onto each others channels through being picked up by the microphones and any synchronisation issues would be very apparent by adding an echo to the podcast.
We record two pieces for each show. The first is the pre-music introduction and this generally does actually take a few takes. This isn’t actually a bad thing as it means that we end up getting into the mindset to record while doing this and, if you’re interested, the record number of times it’s taken Tim to get the name of the show right, what the date or show number is and what virginworlds.com is called is six.
We next record the main show segment, which starts with me remembering that I’ve not thought of a nice way to start yet and just saying the same old thing. We then just natter through the next hour about what we’re talking and are done. If it’s a news show then I work my way through the tabs open in firefox (it’s important to have them opened ahead of time, which is why saving them as a folder in your bookmarks is handy as you can open them all at once, and I use my hand written notes to help remember what I’m saying. As we cover each story I use an add-in for Firefox called Zotero to add each page into a folder in that, which also has the current date in it. Zotero is a research tool that generates bibliographies and I use this to generate the show notes. When we’ve run out of things to say or news items to cover we stop recording and get back to playing games. If anything does come up during the show like a pause I make a note in the notebook of the time according to the recorder so I can address it in the edit.
Later on I will edit a show using Logic on the Mac. There is nothing special about Logic and any audio editing application will do. I have created a template for the podcast that contains the theme music with automation set up to fade the music down over time from the end of the main part. We have a two track setup with the music on track two and I drop the two wav files from the Zoom recorder (mounted as a disk drive over USB) into track one. I trim the intro audio and position it accordingly over the start of the music. The main audio is then trimmed at the start and end and a fade added to the very end from the start of the silence after we say goodbye for a second or so. This sounds better than just stopping.
My logic template has a compressor configured for both track. Track 1 is a standard setup to make the volumes a bit more useful and smooth out times when heads may not be as near to the mic as they should be, or voices are raised. A compressor has nothing to do with file size, but in fact works on the volume to effectively smooth over the differences and make it all as loud as it should be.
Track two has a second compressor, which is configured via something called a side chain to be a dipper. This means that when there is sound on track one (us speaking) it reduces the sound of the music. By doing this I should technically never have to worry about dipping audio down manually while we speak, which is one of the mistakes everybody will make from time to time.
If we recorded everything correctly then at this stage I know that everything is fine content wise. I will then check various parts of the recording to make sure the sound is OK, and use the waveform on the track view to guide me. If needs be I can use automation to push the compressor higher for individual sections, but I don’t need to do that very often. If other music is needed to be added I will drop that onto track two to take advantage of the compressor and check that it does indeed dip down a suitable amount. This part takes less than half an hour nowadays, and can take as short a time as 10 minutes. I no longer listen to the whole show to make sure I didn’t miss anything as I’m making notes as we go. Before that was the most time consuming part of the process as a one hour show would take two or more hours to edit.
When I’m happy with the show I hit bounce in Logic and it goes off to create an MP3 file. At this point I lose interest and turn the Xbox on. When I remember that the bounce should be finished I open the resulting MP3 into iTunes and check the start, end and make a note of the runtime. I then use iTunes to add album artwork and edit the ID3 tags. I then drag the file to the desktop, rename it and start my FTP client to upload the file to libsyn. Then I go back to playing games.
When I remember that the upload would have finished a while back I then head over to Libsyn and post the show. If it’s a what we’re playing show I just need to remember what we talked about, and I’ll have made notes about that. If it’s a news show then I select all the news items in Zotero and drag them into a text editor. This gives me the html for a list of items using a custom citation style that I created. This is then copy and pasted into the notes and published. I then ping feedburner to force it to update my feed. A similar thing is then done with the vanhemlock.com post about the show and everybody can then have a listen to what we’ve been saying.
