0 Members and 1 Guest are viewing this topic. Read 13190 times.
According to who?The players are a lot more sophisticated than they used to be.
Well, I got waved off of that technique big time by others who told me that digital volume attenuation removes digital information, so, say, if you have 16 bits at full volume, you have 15 bits if you turn it down a bit, etc. So I stopped using it.
But you're still losing data,
Uh... no, you are not.
I'm assuming that you have sufficient bit length at the output i.e. 16 in, 24 out, which I think is probably typical. (no?) Take your example of having the volume "halved" - shift right one bit. Does the data change? No, not a bit (ha - OK, one bit exactly). It's exactly the same, extra zero-padding bits and shifts aside. Fair enough, take it down more than 8 bits and you lose data - but this should be well down below other factors like noise unless you have a gain issue in your system.In general, I think these "bit" and "data" arguments are forgetting that the digital signal is a representation of an analog signal, and that the real goal is to reproduce the analog signal as accurately as possible. So in general, I think that using the processing power on the computer for things like upsampling and dithered volume control is really moving in the right direction.
Thanks for the condescending reminder. "Realistic" does not mean "loud all day long" - it means "realistic."
Neekomax, I was thinking of sound quality.
Perhaps that is where our cases differ; there is no upsampling going on in my setup. My 16 bit music is played back in 16bit, streamed in 16bit, and my DAC uses 16bit conversion. Therefore, as far as I know, there are no extra bits of 'padding' to be lost without sacrificing musical data.
Fair enough, take it down more than 8 bits and you lose data - but this should be well down below other factors like noise unless you have a gain issue in your system.
Isn't 24 bit recording and 24 bit playback rather manditory for best fidelity?
In theory, yes. But the recording production itself is important (analog microphones, analog mic preamps, acoustics, mixing, etc.), and many of us have 16/44.1 recordings that sound better than a number of 24/192 recordings made from that resolution or higher masters. I think 16/44.1 recordings will comprise a respectable portion of most people's music library for some time to come, for the above reasons and because of availability (or lack of). Steve
neeko,in response to your output impedance question. It is advisable to have a ten to one ratio between source and preamp and preamp and power amp. By way of explanation if the preamp input is 10Kohms the source component should be 1Kohm or lower. For the sake of dynamics the source should actually be an order of magnitude lower,around 100 ohms or less. If the power amp input impedance is 27Kohms the preamp output should be less than 2.7Kohms for the same reasons mentioned above the actual output impedance should be lower much much lower. This ten to one ratio minimizes the possibility of high frequency roll-off due to interconnect cable capacitance in combining with the preamp and power amps impedances to create a filter pole inside the audio bandwidth. A filter pole is always created when two components are connected together by a cable with capacitance,the trick is to have the 3dB down point well outside the audio band,ten times higher isn't a bad figure to shoot for. Scotty
I guess I could start by finding out what the various impedances of my components are, right? Is that usually listed in the specs?
We would all think and hope so, and sometimes they are, but all too often you will see posts asking "Does anyone know the input (or output) impedance of ........" Steve
