I use "Target Normal Volume" 96.5 db to run all my mp3 files
but i notice that almost 90% of MP3s have clipping (Y under clipping)
so, what exactly does clipping means?

does that means that MP3Gain did not adjust my mp3?
or it means that there are distorted sound in mp3?
what's the effect when clipping occur?

this is how I runs my MP3Gain
I add files, then simply click track gain ("Tags" "Ignore (do not read or write tags)" is checked)
I do this to all my MP3s, and all of them sounds fine although 90% of them have clipping
(well, some of the MP3s seem didn't get adjusted)

all i want is simple, all MP3s are adjusted into the same volume, all of them!
am I doing the right thing?
if I am wrong, tell me the right thing to do...
Thanks in Advanced anyway

You're asking for clipping when you adjust the volume so high. Leave it at the normal 89dB and you'll be fine.

No. If it's under the clipping column, it did adjust your mp3.

Tops of peaks get chopped off. Result is distortion which may or may not be audible, but with an aggressive reference of 96.5, there's a good possibility that it is audible. It really depends on how compressed the source is to begin with. Compression allows for louder music without clipping, though the more compressed the music, the higher the peaks deviate from the lossless original when converting to mp3.

Lower the reference so that less than 90% of your tracks are clipping or until you can't hear any clipping.

I don't understand the last part of this sentence. Are you saying that mp3 can't reproduce compressed music as accurately as uncompressed music?

No, it means that if you encode something with some error, 9.999 is more likely to get encoded as 10.000 than 1.263 is!

Hence audio that's slammed against the maximum is more likely to exceed the maximum after encoding than audio that rarely (or never) goes anywhere near the maximum.

Cheers,
David.

I always thought that clipping and compression were distinct, and often unrelated, phenomena. In fast, the more compressed, i.e. the less dynamics, the easier it is to avoid clipping. The fact that these are routinely used together is simply a characteristic of today's mastering.

Obviously clipped waveforms are going to be harder to reproduce with mp3. What I was asking was why would compressed music be harder to reproduce?

Maybe some of the clipped files were 96.5 dB to start with?

CDs with 96.5 dB volume aren't rare, and it's possible that the CDs have no audible clipping.

True, some files if adjusted to 96.5 db will sound just fine, but if you selectively adjust ones with audible clipping at this level to a lower level then you have defeated the purpose of MP3Gain - to allow you to listen to all of your music without needing to adjust your volume. As long as even some of the files in your collection need to be adjusted to a lower level, all of your files need to be adjusted to that same lower level.

I'm saying that more compressed music is going to have peaks that extend farther beyond those in the original lossless source than less compressed music. I regret that I'm not able to give you sufficient reason as to why, though guys like David, Benski and many others certainly can.

Definitely and it's possible that all of bds's tracks are 96.5 dB or greater prior to adjustment. Thanks for the correction.

Yes, but a lossy reproduction of such a CD may have audible clipping but wouldn't at a reduced level (hence the indication in Mp3Gain). I typically believe in auditioning such situations rather than simply assuming the clipping is audible because there's a good possibility that the clipping won't be audible.

I also agree completely with pdq that it defeats the purpose of the program to selectively adjust individual tracks. I never understood why people use the program to apply maximum gain (or minimum attenuation) to insure there is no clipping to each track (or album) independently; yet it does provide such an option.