InsectaPod Blog

We ran into some trouble recording episode 7 and had to re-record Anna’s narration. The problem stemmed from my attempts to even out the sound levels by using compression. Uneven sound levels is a pretty big concern with this podcast because an audio track can’t be any louder than it’s one point that reaches 0 decibels (the point before distortion). So even if the majority of the track is very quiet, if one point peaks as 0 dB, we can’t really raise the volume on the rest of it without causing distortion. And even if we do accept distortion at that one peaking moment, when we raise the volume on the whole track we raise all the background noise, too. So, ideally, the input signal should be very even and very loud so as to keep the levels at the same volume without distortion and to maintain a high signal-to-noise ratio.  A common tool to even out the levels is compression, so we tried using it on Anna’s narration in episode 7.

The thing about compression, though, is that while it squishes the loudest part of the signal closer to the average level, it also brings the lowest part up. When we set Anna up to read the narration the levels on the recorder weren’t as high as they should have been. When I sat down to edit the narration, I foudn that the noise-boost inherent in the compression compounded by the low signal-noise ratio resulting from us setting the input gain ineffectively made too much noise for the recording to be usable.  So we recorded it again more carefully and without compression and got a usable file.

Clearly, the input gain was a huge problem on that first recording, but I’m not sure that was the only thing. I’m suspicious that maybe the compression in the Flash recorder we use maybe isn’t tuned for spoken-word sure. There are no parameters to adjust, it’s either on or its off. I know that many podcasters use outboard compressors with great success. But I also know that making an expensive comp worhtwhile requires a room with some acoustic integrity, which we lack.

In any case, the more we do this, the more I’ve found that simplicity makes things easier. The first time we set up to record narration I hung blankets on the wall in a small room, rigged a condenser mic into a preamp and sonic maximizer before hitting the USB interface for recording on a laptop. The result was a mess, there were just too many things going on. Now Anna just reads into the same rig we do all the field interviews with. It’s easier, faster, and more reliable.

The flash card that came with the device we use to record interviews held just 126k of info. For that reason, we recorded those interviews for episodes one and two as mp3 files, we couldn’t fit very much info on a card at an uncompressed quality level.

Memory is cheap, though, and we jumped up to a 2 gigabyte card recently and started recording at higher quality levels. The problem is that it takes a lot longer to load up a file for editing. Anna transcribed the recording from a recent interview yesterday, but really had trouble being patient waiting for the file to load. It’s not really a big deal, just that the process becomes more cumbersome as opening a file or laying an effect like normalization or equalization on it takes 5 or 10 minutes.

I’m not really even sure if it makes sense to record .wav when everything gets shifted to .mp3 for downloading, anyway. I think it’s a good idea to record at the highest quality we can, though, because we might want to put the episodes on compact disc or something down the road. I know that when I listen to cd quality for a while and then switch to .mp3, i can discern the lower quality. If we record at high quality we can alway compress the files to make them more manageable, but if we record compressed we can never make them sound full.

I posted the second episode of InesctaPod Cast this morning. I’m happier with the sound quality of this one, and with the ease of editing the interviews and voiceovers together (hopefully that will get faster every time). The process of cutting up all the parts and then splicing them together into the narrative order took maybe 5 hours, but it was fairly smooth work. I used a different audio editor this time, Cubase instead of Audacity, and I think that made things go more quickly. Also, the included effects in Cubase are better than Audacity. That probably doesn’t make any difference as far as, say, the tube emulating distortion, but the compressor in Cubase offers a lot more control.

After spending that time yesterday creating the audio file, I updated the web pages and the rss file and posted everything. I also cut out a snippet of the epsiode audio for the InsectaPod MySpace profile (the entire episode is too big to upload to myspace). Probably another 2 hours spent on that stuff.

Worth it, though. It feels good to see a second episode up there.

On Sunday Anna and I took our new microphone to the Bug House at Michigan State University to see what kind of recordings of actual insects we could pick up. The Bug House is an outreach facility open to the public and targeting kids. They’ve got a lot of enclosed exotic bugs for kids to learn about. When Anna spends time there it’s mostly centered on gently asking youngster if they want to hold a giant centipede and reassuring moms the centipede poses no mortal risk.

I bought the microphone, an audio technica short shotgun, hoping that the high sensitvity would allow us to pick up some of the fairly soft noises made in the bug house, such as those of a scorpion scurrying across sand. This mic is similar, I think, to the one my friend Jon Slaght uses to record owls, and by my reasoning, if it’s sensitive enough to record an owl hoot at 100 meters, it should be able to get the pitter-patter of a scorpion at 2 cm. I knew it wasn’t the perfect tool for the job at hand, but I was thinking it would be a good tool to add to the collection. That may have been a naive assumption on my part, though. We managed to get some good stuff, like these Madagascar Hissing Cockroaches, and even the scurrying scorpion is audible, but most bugs didn’t make enough ruckus for the mic to pick up.

Which is too bad, the ability to record insects well was something I was banking on for InsectaPod Cast. I think the cockroaches are loud enough that we could have gotten something similar to the recording above with almost any $100 dynamic mic. I’ve spent some time looking for insect recording setups online, but haven’t found anything we can replicate. Most of the stuff google turns up is about recording loud groups of insects at a distance, such as crickets. If there’s stuff out there about capturing the tiny sounds of solitary bugs, I’d love to hear about it. At this point, it looks like the shotgun mic will be going back.

One of the things people on campus seem to be interested about regarding PodCasts is what kind of equipment is required to record one. The fact is, you can start out pretty inexpensively. Here’s a list of what I used to record episode 1.

• Microphone. Shure SM57. This is the bic pen of microphones- cheap, reliable and ubiquitous. It costs $100. Jeff Evans from the Landis Laboratory at MSU found mine on a sidewalk in Lansing. It was later stolen from me at a bar where my band was playing.



• Zoom H4 Handy Recorder. Flash recorders are kind of new on the scene and are replacing minidisc which replaced tape. This one cost $300, is compact, and has some decent features like built in stereo mics, xlr inputs for external mics (like the SM57) and built in audio compression. The 128mb included flash card holds 2 hours of mp3 and 9 minutes of high quality .wav file. I recorded most of episode 1 in mp3 because of the limited storage. I recorded all sounds, whether recorded in the field, interview, or narration, onto the H4 and then downloaded to the computer for editing.
• Audacity. This is a good open source (free) software for editing audio.

For recording a podcast that involves field or interview work, I think this is the bare-minimum. To just sit and recite something into a computer, you could use a $5 pc desktop mic, but unless you’re absolutely brilliant, why would anyone listen? I’ve already got some new mics I’m going to try out for episode 2, so there will be another gear list next month.