astromo and glynor: Can we split off the volume discussion into it's own thread since I believe it warrants more discussion. I have always been under the impression that any reduction in volume via software was essentially reducing bit depth. Therefore, it was best to have it either disabled or maximized to 100% so that it can be controlled in the analog realm. I would be interested in hearing other points of view.
I personally prefer digital volume, and at the risk of leading the thread further astray, I'll go into some of the technical points. To be clear, the main reason I use digital volume is because it's much more convenient for me to control volume this way (block amplifiers, no remote control for the DAC, etc.). I don't think digital volume control is "superior" in terms of sound quality in most cases (although sometimes it can be), but there are some reasons why it's not necessarily a bad idea to do digital volume control.
If you (like me) listen primarily to redbook CD music in JRiver, and your DAC supports a notional 24 bit output (most modern DACs do), JRiver outputs at the highest bitdepth your DAC will support. That means that 16 bit audio is output as 24 bit, with zeroes padding the bottom eight bits. So in that scenario, you get some amount of digital attenuation "free" without getting close to the theoretical 16 bits of audio information: even quite significant digital volume attenuation (15-30dB, depending on the noise floor of your DAC) can still be more or less "bit perfect."
But even in situations where you might theoretically be compromising audio information there are a few reasons why digital attenuation may still be just as good as analog volume control:
1) 16 bit audio allows for a dynamic range of about 96 dB. Most music doesn't actually have anywhere close to that kind of dynamic range. You can test this by using the method mojave describes in this thread:
http://yabb.jriver.com/interact/index.php?topic=84970.0. He found that baking in -40dB of attenuation to a few 16 bit tracks didn't appear to affect the audio analysis of the tracks (i.e. DR and R128 stats remained the same). Generally speaking, that's been my experience when I repeated his experiment. Obviously YMMV, but I'd be willing to bet that it's a rare recording that gets close to using the whole 96dB. That means that in many cases even if your digital attenuation is "compromising" that 96dB envelope, you've got some distance to travel before you are likely to start throwing away actual audio information. Exactly how much distance will vary by your program material, but even, say, 10dB of "free" additional attenuation is nothing to spit at.
2) Analog electronics are not noise free, and analog volume controls can sometimes contribute significant distortion (Johnson noise from the potentiometers, etc.). I've had several amps over the years where the noise floor of the amp changed audibly depending on the setting of the volume knob, or changed when the knob was adjusted. All analog volume controls will contribute
some noise, but well-designed ones don't contribute significant or audible noise. Not all are well designed, however. JRiver's digital volume control contributes no distortion.
3) Even assuming one has a low-noise volume control, what's the noise floor like on the amp? Many amps claim a higher than 96dB SNR, but it's a misleading spec because it's usually calculated near (usually just below) the maximum power output of the amp (a volume at which most people are very unlikely to listen). To really meaningfully achieve 96dB of dynamic range, an amp would have to have a THD+N of about 0.0015% at normal listening levels. There are some very nice amps that have THD+N specs like that, but my current amps definitely don't, and I've never owned an amp that did. So it's worth considering whether the full bitdepth can even be resolved by the amplifier one is using to begin with.
As a side note, some modern receivers have a digital volume control (not to be confused with a digitally-controlled analog volume control, which is still an analog volume control), so for those receivers the digital attenuation is being performed somewhere, and it's just a question of convenience where that attenuation is happening. It's just a question of what's easier and whose volume control you "trust" more.
And I also support splitting, but have no power to do anything about it