|
|
發表於 2013-2-13 12:03:51
|
顯示全部樓層
From a pure digital signal processing point of view, it's impossible to create more information by upsampling, but it's entirely normal for upsampled music data to sound different - could be better or worse depending on the following:
1. The upsampling process will always introduce a small amount of distortion. The distortion from the best upsampling algorithm may be negligible but still exist, whereas a cheap algorithm can introduce a non-negligible amount of distortion.
2. With an audio signal bandwidth limited at 22kHz, those with a higher sampling rate (say, at 192K instead of 44K) will allow the same DSP filter in the DAC to have a lower resulting noise, hence improving the signal-to-noise ratio.
3. If the SNR improvement at (2) is above the SNR degration introduced by upsampling then from an overall system point of view, the signal quality has improved.
In real life things get a little more complicated:
- You need a really clean and revealing system to enjoy the noise reduction at (2), at a setup where the noise floor is high, the noise reduction at (2) won't do much good.
- Even in a clean and revealing system, the reduced noise floor will expose deficiencies at other parts of the audio setup, and some may perceived it as harsh or digital sound.
- The distortion introduced by upsampling may be considered by some as "taste". The high frequency noise introduced could give the perception of more high frequency details, more airy or even bigger sound stage. |
評分
-
1
查看全部評分
-
|
IP卡
狗仔卡
提升卡
置頂卡
沉默卡
喧囂卡
變色卡
顯身卡
