1 (edited by kevinw 2008-04-25 17:20:17)

Topic: Getting better audio from Olympus DS-2 + ATR55 combo (or similar)

I use an Olympus DS-2 recorder (~$80 two years ago) with an Audio-Technica ATR55 mic (the $45 one that looks like a shotgun and has a 'tele' and 'normal' setting).

The mic is a monaural mic, so I had been recording in 'mono' mode. But I got curious about some warbly artifacts I was hearing at higher pitches. Interestingly, the recorder's specs show a much wider dynamic range for the 'Stereo HQ' setting than for the 'Mono HQ' setting.

I did various tests of speech and ambient sounds (e.g., running water into a metal sink, with lots of higher frequencies) and there is an enormous difference. By recording in stereo mode, you'll only get audio on one channel, but the frequency range is much wider. The spectrum plots show a very steep drop off at 7,000 Hz in mono, but in stereo mode the drop hits at around 14,000 Hz. In my recordings, at least, the amount of sound at that frequency was already quite small, so the cutoff didn't result in audible artifacts. The difference is noticeable even on laptop speakers, and obvious with a pair of earbuds.Be advised that the stereo files are about 4 times bigger than the mono files. This is another indicator that there is more information in each channel (otherwise it would just be twice as big), but it will cut into your recording time. I'd probably experiment with a slightly lower quality stereo mode ('Stereo SP") rather than the higher-quality mono mode if space were at a premium; the specs still show a wider frequency response than the mono HQ mode. I haven't tested this yet, though.

For folks using other recorders (W-200/300, or the newer DS models), be sure to experiment. The specs on the newer recorders show a better frequency response range in mono mode, though still not as wide as in stereo mode.

[UPDATE SINCE ORIGINAL: some web searching turns up that the Stereo HQ mode records at 128 kbps while the mono HQ mode is only 32 kbps. Was this in the owner's manual? Noooooo.....]

Bottom line: for best results from this combo, record in stereo mode. In the sound editor, just split the stereo tracks, discard the blank track, and switch the remaining track to 'mono' so it plays through both speakers.

If anyone is interested in samples or seeing the spectrum plots, drop me an email through my profile page.