As this is my first post on this board, it qualifies me as a NewBe! As such, I know I have little if any credibility here, and this is probably a little unusual for a first post, so please bear with me.
And now to the topic: I recently bought a Roland R-05 to replace the minidisc recorder that I had been using up until some years ago when it was stolen. I had first settled for a Zoom H1, but because it suffered from the battery drain problem (it used 47 mA when switched off), I ended up replacing it with the Roland R-05. I was maybe a little sceptical at first, not seeing too many tests of the R-05, so I decided to test the noise levels etc a bit while still being able to return it.
The conclusion: It is very good, and looks almost identical (in terms of noise performance and likely also dynamic range) to the equivalent Sony PCM-M10, at least when comparing the specs found over at
http://www.avisoft.com/recordertests.htm. It is better (noise-wise) than the Edirol R-09HR, and will offer either lower noise or more dynamic range at the same recording levels. I was being asked to consider also the L-09HR by the friendly salesman, but I am really happy I didn't spend the extra money (here it is 60% more expensive, as is also the Sony). For most of us in Europe, I guess the R-05 is the best value for money in its price class. In the US, maybe the Sony is?
One concern I've seen raised here is the input sensitivity. It is spec'ed at -7 dBu, but is in fact in reality more like -0.6 dBu at default gain for the LineIN. So maybe not that different from the L-09HR in practice (if that is according to spec then...) in terms of input levels?
If anyone cares for numbers, the EIN is around -122 dB(A) in the highest gain settings, and in the range between -30dBu to +20dBu (roughly) it is capable of recording with a dynamic range of more than 90 dB. It thus requires a pretty quiet microphone before the noise of the preamp starts to matter.
The built-in microphones are well matched to the preamps too, and offer a good, quiet recording as well.
I know some say EIN is not relevant for typical taper usage. I don't fully agree to that even if I understand the logic, but the relevant part of looking at EIN for taper use is that it helps to establish the dynamic range. A low EIN means a quiet recording -- which is certainly true for this unit (up to 98 dB(A) dynamic range).
There are probably more details on this than anyone cares about on my page here:
http://www.b4net.dk/?page_id=111 -- including specific noise data and other observations.
Comments are obviously more than welcome!
-- Per.