This is a description of how we produce the audio files. There is undoubtedly other software and tools to do the job, however we find these more than adequate. They have the advantage of being free, Open Source software. We are using the Microsoft Windows® operating system.
We record direct to the hard disk drive of a PC, and then convert to MP3 format audio files and upload them to the Web host for this Web site.
The church is equiped with a public address amplifier system which has low-level audio outputs, as well as the expected high power speaker connections. The low level audio output is connected to the "line-in" socket of the sound card in the PC.
The recording software we use is Audacity. It is also available for Linux and Mac OS X systems. Audacity saves the audio files in its own project files, but can easily export to WAV, MP3 or Ogg Vorbis formats (Ogg is not particulary suited for voice). We record in mono. After the recording is complete, Audacity is used to edit the required portions: for example, the sermon is extracted to be uploaded to the Web site. The required portion is then exported from Audacity to a WAV file.
After the WAV file is produced, it is encoded to MP3 using RazorLame (a Windows front-end to LAME; Windows binary files here). Even though Audacity can export to MP3 itself, RazorLame gives more control over the settings you can choose. It is a simple process to drag and drop one or more WAV files into RazorLame. Even if you did use Audacity's MP3 export you would still need to install LAME, as that is what Audacity uses "behind the scenes".
For the MP3 files that we upload to the Web, a bitrate of 16kbps and a sample frequency of 22.05Ksps ("KHz") are used. This gives good compression but involves some loss of quality. Also note that such MP3 files may not play as MP3 files on some home DVD/MP3 players. You would need to use higher quality settings for such a purpose.
At times we need to produce many audio CDs, for example if we have a guest speaker giving a series on talks. In that case the audio is not converted to MP3, but the WAV file(s) is used by the CD recording software to produce the audio CD. For that we use the commercial version of Nero Burning ROM. This version (in contrast to the version often bundled with CD writer drives) allows simultaneous writing to several CD writers. We have found that using two CD writers gives good results.
To add a professional finish to the CD, we use blank CD media with an inkjet printable surface, and print using a printer that can print directly to the CD surface (Canon i865 in our case). This gives a much better finish than using either hand writen titles, or paper CD labels which are stuck on. We use either the bundled software that comes with the printer (and includes several templates and designs), or the Draw component which is part of the free Open Source OpenOffice.org software. A very simple CD template for the OpenOffice.org software configured for the Canon printer is here [7Kb].