INTERACT FORUM
More => Old Versions => Media Center 11 (Development Ended) => Topic started by: stefansmith on June 13, 2005, 06:48:05 am
-
Finally got round to analysing all my APE files -- because i wanted to use replay gain to get more consistent volume levels. I've also ben using replay gain to bring down the db level a bit -- because my external DAC, preamp and monoblocks have quite high gain levels, meaning that having the volume knob at 9 O'clock was already far too loud.
Using replay gain to bring down the db level, I get much better fine tuning on the volume control.
However: I *think* that using it takes something out of the music -- perhaps my ears are deceiving me, but it seems the music has less oomph, body, attack at my normal listening volume.
Anyone else noticed this, or am I just a victim of that all-too common syndrome "audiophilitis-psychosis-autosuggestionitis"?? :'(
-
I've use to think of that too. But after listening to all my songs in replay gain for a few days, I can proudly say that I "don't think" there's actually any noticeable difference in quality other than volume. Probably you will need to tune it up a bit more to see if there's any differenece in quality. If you think there is still difference try going into your replay gain setting and set a fixed adjustment. I've set a fixed adjustment of +10DB because I feel that replay gain is giving me a too soft volume altogether and thus I increase it such that the volume is of suitable level to me.
-
In my opinion it is possible, but unlikely. MC uses 32-bit signal processing and I doubt DSP volume adjustments has effect on audio quality in a perceivable way. If there are any differences they must come from the soundcard/DA converters/preamp chain. Some devices can behave differently with different volume levels. Only blind ABX testing (using volume level calibration just before the power amp) could tell if there really is any difference.
-
Even if the DSP calculations are perfect, if you reduce signal level you will degrade the sound quality. This is because you'll be using less of your D/A converter's output range.
For example, let's say your source material has peaks that come very close to 100% of the DAC's output range. This is an optimum case, since you're utilizing the full resolution of the DAC. Now if you reduce the signal level by 6 dB, the peaks come down to 50% of the DAC's full scale. In effect, you're now listening to music on a 15 bit DAC instead of the original 16 bits. You lose one bit of resolution for everey 6 dB of signal reduction.
The audibility of this degradation can certainly be disputed, but there is no question that information is being lost.
-
That's why I always use 24 bit output together with the Replay Gain. My DAC can operate at 24-bit. That gives 8 bits more headroom for 16-bit CDA tracks.
-
Even if the DSP calculations are perfect, if you reduce signal level you will degrade the sound quality. This is because you'll be using less of your D/A converter's output range.
For example, let's say your source material has peaks that come very close to 100% of the DAC's output range. This is an optimum case, since you're utilizing the full resolution of the DAC. Now if you reduce the signal level by 6 dB, the peaks come down to 50% of the DAC's full scale. In effect, you're now listening to music on a 15 bit DAC instead of the original 16 bits. You lose one bit of resolution for everey 6 dB of signal reduction.
The audibility of this degradation can certainly be disputed, but there is no question that information is being lost.
Keep in mind that MC is fully 32-bit, and can output a 32-bit or 24-bit bitstream to a DAC. In this scenario, there's no loss of precision from volume adjustments.
Also, it's amazingly hard even with high-end headphones in a silent room to distinguish any difference from extra bits after about the 13th bit.
-
That's why I always use 24 bit output together with the Replay Gain. My DAC can operate at 24-bit. That gives 8 bits more headroom for 16-bit CDA tracks.
On paper, that's correct. In real life, even the very best 24 bit audio DAC is limited by noise to about 20 bits. That's not counting the noise effects of any circuitry following the DAC, so by the time you get to your speakers it may only be 18-19 bits, perhaps less.
But again, the audibility of all this is certainly debateable. I know that there are sound differences between DACs of equivalent resolution, so this is a rather complex discussion...