So I have a little story to tell all of you about the creation of the first set of the Half-Life 2 episode sets. First you may need to know that I essentially record at night or very late in the evening from wednesday until saturday. These past couple of months I’ve been recording the Borderlands sets on saturday evening as that’s what we decided upon before starting the series. So I recorded on Friday night with the idea to edit and render overnight. I ended up playing for 2 hours.
After recording it’s 2am and I decide to stop and check the footage and import it into Vegas to do some quick editing and set it off. This process usually doesn’t take me longer than a couple of minutes, however this time the commentary file was only 3m 25s. This was not quite the 2 hours I was expecting. So I tried to repair it by importing the raw data and…. no luck. So I looked at the backup file I had on the other harddrive as maybe it got corrupted when being moved, but no luck. So I went to bed and continued in the morning as I was really tired.
The next morning I got back to it and saw the file was indeed 4.4gb, indicating that the audio had been recorded, assuming it hadn’t just padded the file with empty data. So I quickly found some info about the wave file header and how it’s really quite simple. It seems that the length part of it would probably just be wrong, if I could edit that I’d be set I thought. So I looked for an app that would do that…. Turns out I was wrong. So I then looked out for repairing wav-files and.. it mostly just came down to importing raw data into Audacity at that point. So.. I tried that again and that didn’t work with the same settings I recorded at.
So I ended up trying some different settings to see if I could get something out of it at all, instead of really loud noise or a loud pop and silence. Turns out I could get some sound out of it so I was convinced the audio was indeed there and.. the application had recorded with seemingly different settings. So after lots of trial and error, I found the correct settings and got just short of 2 hours of commentary. I then imported this into the Sony Vegas project so I could edit it together and get it rendering before the recording later that evening.
At this point it was probably roughly 1pm. But that’s actually when the bad part started for me. I had my commentary so what could be wrong? Well, roughly 3 minutes of it seemed to be missing. Not all in one part and seemingly no actual commentary on my part. Just 3 minutes of silence. This however did ensure that it was out of sync absolutely everywhere. So I had to make a cut essentially every minute or two and sometimes even more often and try and sync it up. This was such a painstakingly long process that it took me several hours to do and because I really needed the space for the recording later… I didn’t have time to render a bunch of footage to test it either. I synced it up to the preview though, so it just had to be good enough.
It was probably around 3:30pm-4pm and the rest of the time was spent Rendering the actual footage. I ended up finishing rendering the footage roughly 30 minutes before we started recording. I was quite lucky to even have made it. I also learned that I say some strange stuff that makes sense to me at the time but may be hard to get at times, especially when I had to sync it up afterwards as it could be really disjointed. I have noticed that at some points it’s not entirely synced up, but the difference is minor. I just hope it won’t piss you folks off too much as you all seem to adore the episodes.