Just submitted 2 characters, Vineto and Anjaneya control.
For all you lurkers here with microphones: Gather your strength and head to http://www.goldwave.com/release.php.
The trial version should be more than enough. Here are some tips for at least adequate sound quality:
- Go to Windows Control Panel --> Sound --> Microphone --> Set 44100 Hz and 24 bit.
- Turn Microphone Boost ON and "Listen this device" OFF
- Set the recording volume to a sufficiently low level, lower than 30 %.
- Install Goldwave trial version
- Create a new file, with a preset of 30 seconds and MONO
- Adjust Volume control options --> enable 24 bit PCM for both input and output --> Set the mono source channel to AVERAGE (the default setting "left" is just stupid)
- Do your magic
- Remember to select and cut off the remaining silent part with DELETE. Now your file should be just the right length.
- Select the whole file ---> Click tab EFFECT --> Filter --> Noise Reduction --> Default preset
- Save the file with the correct name file, choose OGG, 44100 Hz, 1.0 q, 240kbps, mono
Good luck to everyone, "we can do this!"
For those using Audacity or any other recording program, I'd like to condense the former into three main points:
1. Make sure your recording does not have clipping on it. Clipping is when the waveform signal from the microphone is so strong that it "peaks" beyond the valid area, creating "plateau"-like waveform that sounds like a demented kazooist during a Saturn V launch.
2. Make sure you're getting good enough signal-to-noise ratio on the recording. This is quite easy to do by monitoring the waveform; as long as the general thickness of the waveform while you're talking is clearly distinguishable from the "silent" level, you're good. For obvious reasons try to minimize the background noise level, too...
3. Familiarize yourself with the noise reduction utility on your recording program of choice. Audacity especially has a very powerful noise reduction tool that actually allows you to analyze your particular recordings for a noise pattern, making it more accurate at detecting which parts of your recording are noise and which ones are not. To do that, I recommend leaving some trailing or leading silence in your samples while you're recording them. After amplifying your track, you can then select a bit of the noise, open the Noise Removal tool, and click on "Get Noise Profile". The tool will close without any dialogue, but the noise profile is now stored. Now you select the track fully, click on "Noise Removal" again (or "Repeat Last Effect") and click OK. By my experience, the default settings are quite fine.
4. Submit samples preferably with lossless codec (PCM WAV or FLAC), or if you must use compression (for bandwidth or email limits), use highest-quality VBR OGG Vorbis.
...FOUR main points.
MP3 codec in particular is to be avoided at all costs. There are some technical reasons for this.
MPEG-3 codec is inherently lossy so using it places some limitations on what we can do with the submitted voice samples. Same applies to OGG but to lesser extent because:
MP3 is designed specifically for music, not for speech. In addition, voice acting samples often have at least some level of noise in them unless they're already properly treated or recorded with professional studio grade equipment... and MP3 codec considers audible noise as "signal" but doesn't really know what to do with it, and often the result is that the codec produces all kinds of strange artefacts (tingly noises, weird harmonics etc.) that can be almost impossible to remove.
OGG Vorbis is a more modern codec and better than MP3 in every possible way.
Possible suggestions for reducing noise / improving the SNR:
-make sure microphone cable doesn't run right next to some transformer - those are often notoriously bad sources of RFI that an unshielded cable will pick up like an antenna. This is a typical, common source of 50Hz "mains hum" type noise. I'm getting that too, but there's not much I can do about it...
-connect microphone directly to the sound card / motherboard's mic input on the back of the PC, instead of using the front panel connector. Sometimes, the front panel connectors can be a source of noise, be it caused by ground loops (typically causes a grinding or whining noise that varies based on things like CPU load), or RFI from the devices inside the computer (such as graphics card). Even if the sound card itself were shielded against such interference, the front audio panel wirings are typically a weak point in computer case design.