Ever since I started my podcast, I've been searching for ways to improve the quality of the voiceover, and today I finally made a breakthrough. Realizing that every show has been slightly better, I analyzed the changes I'd made along the way.
- Changing mics made a big difference; the Rode Podcaster I'm using now puts out better highs and lows than the cheaper mics I started with, and its built-in headphone jack lets me hear the effect of positioning. (As a dynamic mic, it boosts the lows as you move closer.)
- Experimenting with compression and EQ settings in Izotope Ozone is improving the punchiness.
- Recording in the closet instead of at my desktop helps prevent slapback echo and that "small room" sound.
- Recording during the day instead of late at night lets me add more volume and intensity to the delivery.
- Reading through the script out loud several times helps me find and fix awkward phrases.
- Recording multiple takes helps me warm up and get comfortable with the words. It also gives me an alternative phrase to paste in if a line isn't quite right. I usually find the second or third runthrough is the best.
- Being conscious of microphone-overloading sounds like P's and T's helps me remember to enunciate more carefully and move back from the mic when I say words with those sounds. (If thumps do get through, I've found the best way to remove them is to snip off a tiny bit of the beginning in my audio editor.)
- Deleting some of the inhalation sounds and pauses (but not all, as I was doing) helps move the narration along.
- Sipping water helps avoid tongue clacks that the microphone loves to pick up.
- Purposely clucking my tongue generates spikes in the waveform that make it easy to identify sections of the narration. When I make a mistake and want to redo a line, I'll cluck twice, creating a visual indicator for the improved version.
In short, I've been building up experience, and so I decided to encode those discoveries into algorithms that could run on a small DSP box. Here's the happy result, the Elvitron 401, named after the first before-and-after example that popped into my head.
The Elvitron 401 receives underripe vocals over USB and returns them to the computer with a more confident, mature sound. Internal DSP algorithms heighten intensity, punch, and even microphrasing.
Here's a before-and-after audio example:
- The Elvitron Experience (516KB MP3)
I'll add more background on the project soon. In the meantime, here's a high-res photo of the inside of the box.
Categories
Audio
Comments (4)
Read More Entries by David Battino.
This made my day. Great article! More people need to read this.
the list of 10 [ways to improve your vocals] looks plausible (though expensive gear)
I should emphasize that only two of the ten items have to do with gear, and that there are cheaper alternatives to the specific mic and software I listed.
I’ve been impressed with MP3 demos of the Giant Squid vocal mics, which cost $65, and you can emulate the functions of Ozone (if not the flow) with freeware multiband limiters, maximizers, and EQ. GigaVox Levelator, for example, is a one-button audio enhancer, though I’ve found it too heavy-handed.
@Orcmid: Yeah, I spent way too much time on the case design. But the items in the list are all techniques I’ve discovered and use, and the audio example is a true before-and-after excerpt. If only there were a big “experience” button you could push to make things better. It makes one realize how much gear is marketed that way.
This leaves me in a quandary. It was obvious that the Elvitron was an April Fool prank, and of course I looked to see what the punch was.
But the list of 10 looks plausible (though expensive gear), even to the double-clucking. Color me clueless.