Well dennisrtyler did what I should have done and found a previous thread while I was typing, but instead of delete al this I post anyway
.
In general, you can make either a 'data' or 'audio' CDR.
You can burn the 24bit files unconverted as a 'data' cd which would be playable on a computer that can playback 24bit files but would not be playble in a CD player. This retains full audio quality.
To make a 'red-book' audio CD that should be playable in a regular CD player you'll need to convert the 24 bit files to 16 bit files and burn with an audio CD recording application. That might be taken care of by your burning software, or with an audio editing software (such as Audacity). In doing so you inevitably throw some information away, but depending on the source material that may not be detectable and there are things you can do to minimize the quality loss. Using dither is one of those things, see below.
There are two ways to convert the 24 bit files to 16 bit. The most basic way is to just truncate (throw away) the bottom (think quietest) 8bits. A better way is to use some sort of dither, which in non-technical speak essentially 'smoothes out' the ragged bottom edge where those bits are removed. There are different forms of dither and they can sound different, mostly evident when you turn up the volume on the quiet parts and listen to the decaying reverb and the 'space' around the notes. Audacity has several options for dithering, check the help files. You can set the options to automatically dither to your output files as you choose. It's relatively simple. It's been a long time since I used Audacity but the default dither options are probably fine.
Other things to keep in mind-
The dither stage should always be done as the very last step, after you do any level changing, editing, fading, eq, or anything like that (if you do any of that at all). Most people will at least adjust the gain if nothing else to make the loudest parts of the recording just a few db's below the maximum level which makes them easier to play without cranking up the volume and also shifts the music 'up the bits' so that when the bottom 8 bits are thrown away you don't loose as much down there.
Hope that's not to technical. There is loads of more detailed info on this site if you do a search and also in the Audacity help files as well as other places on the web.
Oh yeah, If the recording is something important, you'll probably want to keep the origional 24 bit files.