DanMan32 said:
Most radio stations use companders which compress the volume to a semi-constant level. You can often hear this during silent times of the broadcast where the gain control turns up, thus amplifying station/studio noise until some sound (talk, music, etc) comes through, and the auto gain control turns the gain down so that the volume is flat.
CD has a very wide dynamic range, and classical music is also very dynamic in sound level. So in order to accomodate loud portions, the soft passages can seem very soft. Also the 'average' sound level can vary from CD to CD. I believe though it is because of the CD dynamics that the CD input to the main pream/amp is lower than the radio input, so that a maxed out CD input level would not overdrive the preamp, causing distortion.
Movies on HiFi tape and DVD are notorious for seeming too low, then suddenly too loud during some scenes. Most DVD players are equipped to try and equalize (compress) the sound level a bit so soft sections aren't too soft, and loud sections aren't too loud.
I can add to this: The CD, recorded at 16-bits and 44,100 hertz, has incredible dynamic range as DanMan says.
But believe it or not, most commercial CD's are "mastered" to "squash" the dynamic range and make the quiets and louds very similar in volume.
Lots of music (recorded with microphones and direct plug-to-computer) is recorded at 16/44,100 (the same as a CD) or 24-bits and 96,000 hertz (higher quality; makes for better interim manipulation by human and computer). Music that's made "internally" (sounds created BY THE COMPUTER) are often initially recorded at 32-bits. Eventually it all gets "rendered down" to the CD standard of 16/44,100.
At any given stage of musicmaking, a lot of audio recording engineers will use compressors, limiters, and/or normalizers to do some real "magic" with the sounds.
For example, an engineer might use a compressor on a vocal track when the singer is having trouble controlling and/or fine-tuning their volume levels. A compressor can raise the level of the singer's quiet, intimate voicings, and put a "top limit" in on the loud parts. It makes a vocal more listenable, and less likely to get trampled by other instruments that might be in or near those vocal frequencies.
After all of the individual instrument and voice tracks are "mixed," the whole CD project stays in house or goes offsite for "mastering."
Mastering is where they take all the songs on a CD and they make EQ, volume, and other adjustments so that the whole album sounds like it "fits" right with itself. Some pop and top-40 type CDs are compressed and limited so that they'll be as loud as physically possible. Sometimes, this actually degrades the wonderful dynamic range (difference between loud and quiet parts) of a CD.
Then radio goes and ruins it even further, exactly as DanMan says! :x
Anyhow...just some tidbits for you to ponder.