HydrogenAudio

Hydrogenaudio Forum => General Audio => Topic started by: SpashChoke on 2012-12-04 16:04:37

Title: How does a player implement Replay Gain?
Post by: SpashChoke on 2012-12-04 16:04:37
With a music library of over 5000 songs I discovered what the Loudness Wars have done to music.
Replay Gain seems to solve this issue, by reducing the volume of overly compressed audio to a level that is in line with properly mastered music.
However, I can not find how this is implemented.

The iTunes website implies to use a limiting algorithm (as part of its Sound Check processing) in case the gain turns out to be larger than 1. This is an audiophile's nightmare. For the other end of the dynamic range, the bitstream needs to be re-dithered after multiplication with the gain factor.

I suppose Replay Gain is implemented by the DSPs in the players, but how is that done exactly? Ideally, the Replay Gain only influences the volume setting of the analog pre-amp following the DAC, but that would probably be impractical. Alternatively, I could imagine using e.g. a 24 bit DSP, using bits <21:6> to insert the 16 bits from the WAV file (leaving 12 dB headroom for Replay Gain >1) and applying the Replay Gain in the DSP, dither the result and only then DA Convert it.

Are there any players that work this way? What are good players in this respect and which ones should be avoided?

Thanks,
Jan
Title: How does a player implement Replay Gain?
Post by: saratoga on 2012-12-04 16:15:47
Multiplication by a constant.

Scaling the analog volume wouldn't help with clipping.