Hi fellow tapers, I'm new here but have lurked behind the scenes for a while & read through the *whole of this thread* before making the decision to buy a DR100mk3! I just wanted to say that this forum has been by far the most useful source of info I've found on this device so thanks to everyone! I don't want to repeat what's already been said, but FWIW:
First impressions
The DR100mk3 is certainly solid and the design appears well thought through (or evolved, since it's mk3 after all). The ergonomics are great. It's a tad bulky, but couldn't really be any smaller without compromising connectivity and all of the tactile controls. Having a real switch for phantom power is so useful & the fact that this requires an "ok" confirmation prevents accidentally blasting the inputs with 48V while recording!
Originally I was looking for a replacement for a dead Sony M10:( but that machine was really one of a kind in size, battery life, quality, etc. I do have an ageing LS10, but the mics are too noisy for the quiet stuff that I record. My main rig is a Fostex FR2LE, which I love. I've had this it first came out (9 yrs?) and, although I constantly gnash my teeth dealing with the dreadful user interface, I've never had a glitch & I know I can rely on it. Actually, it's great over a shoulder, it's the top panel controls which bug me.
Now, I didn't expect to say this but... the DR100-3 beats the FR2LE in pretty much every respect. Size and connectivity win hands down. In terms of sound quality using the XLR inputs, I did a few comparisons using the AT BP4025 stereo mic which I often use for ambiences etc, and I really couldn't hear much difference between the two recorders. If anything, the Tascam has cleaner pres, but the recorded noise level sounds pretty much identical on both machines.
I have to say that I'd hoped for lower noise internal mics, but I think I have to accept that built-in mics on handhelds aren't going to offer the kind of quality and low noise that you'd otherwise pay as much again for. They are useable for louder stuff though. When close-up they don't spit & the shockmounting is very soft: when prodded the mics wiggle around easily within the protective cage. The result is that handling noise is minimal, unless you actually rub the case. With a furry windjammer and the 80Hz Low cut filter on, it's possible to walk around outdoors and not have obtrusive wind noise. The stereo spread is contained, but ok as the mics are actually angled out at 90 degrees inside the mesh, not pointing straight ahead as they appear. There's good rejection of rearward sound, though the off axis frequency response is rather uneven. I guess this is to be expected. That's why decent mics cost $$$!
I like to measure what I hear (it's the tech background!) so I've just done a few rough'n'ready (unweighted) noise measurements and the internal mic noise works out to ~30dB spl. This is compared to the BP4025 at ~16dB. (The 4025 mic noise spec is 14dBA, so my measurements are pretty close). But the really interesting figure is with mic inputs shorted (150R), which shows just the system noise. In comparison with the FR2LE, the Tascam, is actually *better*, with the equivalent noise = 1dBSPL compared to 2.6dBSPL for the FR2LE! Of course, with a high sensitivity mic like the 4025, these figures are all a bit academic, and taking into account measurement errors, I'd say these were too close to worry about. But it's good to see that Tascam take this seriously.
Incidentally, Phantom power is completely silent, though I'll have to check this again next halloween...
The figures confirm what I hear: the internal UNI mics aren't quiet enough for serious low-level recordings. I think they'd be ok for quick grabs of stuff: close-up recordings to be played back at lower levels, samples, city ambience, even mountain streams, but not for anything with any kind of "silence" as the backdrop. It's not so much the hiss, but the veiling effect that this has on the sound.
Finally, another measurement which might be useful, as I don't think this is in the specs. Plug in power = 2.5V through 2.5K (ie it gives 1mA when shorted). Pity they couldn't have upped this to 5V. So don't chuck your battery boxes!