Since 2006 I have been very fortunate to be able to study the Nez Perce language, Niimiipuutímt, on the Nez Perce reservation in north-central Idaho. (See the tribe's website here.) Here is an overview of my recording equipment for those who may be surveying the options for language documentation and fieldwork.
Recorder Marantz PMD-671 solid state digital recorder
Microphone AKG C-1000S
Mic stand AKG SA 43 (adjustable feet!)
plus... mic and USB cables, cheap earbuds for monitoring the recording, a (rather old) digital camera, spare 9V batteries for the mic, a cheap lapel mic in case of microphone meltdown, and let's not forget a laptop
my recording equipment

I ended up with a PMD-671 recorder after attempting to order a less fancy version (the Marantz PMD-670) but finding it on backorder across the web. I first recorded on a PMD-660 solid state recorder very kindly lent me by Cathy O'Connor; the 671 is not that very different from the 660 as far as I'm concerned (despite being about twice the size of the 660). Two nice features of the 671 in particular are its ability to verify a memory card before you attempt to record on it, and its ability to play back the recording from the card with a 2-second delay while you're still recording (if you can stand to listen to a present conversation while you verify one 2 seconds old). I use a 1 GB card which will hold close to three hours of high-quality WAV files. (For some critical discussion of the PMD-671, see this page.)

The "rock star" style directional mic is ideal for my recording situation as I record one or more speakers at a time (meaning a single lapel mic would not always be sufficient). The mic is powered by an 9-volt battery and must be turned on independently of the recorder. (Note that this has been shown to help sound quality on the earlier model PDM-660, as shown here.)

The few downsides to this setup are that the "on air" light on recorder is very small, making it hard to make it visible both to me and to my consultants (so they know whether they're on tape); that the microphone and recorder must be turned on separately; and that the data must be downloaded after 3 hrs.


Update: 12 March 2008