ReplayGain + Mp3Gain = problem?

[tonitoni]

Hi!

I have scanned some files with mp3gain and I also use Winamp's ReplayGain. Both are seth to 89db.

Is there any bad things doing gain twice?

[J_Darnley]

It could be "over-gaining" the files if you scan with Winamp first and then use MP3gain to apply it again. Winamp won't know that file file has been altered.

If you apply the gain in MP3gain first then calculate with Winamp, it won't be done twice because Winamp will used the already gained output to calculate the final adjust which can't be made with MP3gain.

[Sawg]

Well, MP3Gain physically modifies the files. If you run your MP3s though MP3Gain and then scan them with Winamp's ReplayGain it should, in theory, see the MP3s are gained nicely and not make any modifications.

But no, it's not bad if you run them both.

[tonitoni]

Thanks for the reply Sawg & J_Darnley, that cleared it out!

It could be "over-gaining" the files if you scan with Winamp first and then use MP3gain to apply it again. Winamp won't know that file file has been altered.

I think I have altered some files that way, so what's the easiest way to remove winamp's replaygain values?

Edit
Figured out it myself! Replaygain re-scan in winamp.

Thanks for the help!

[J_Darnley]

Originally posted by tonitoni
Figured out it myself! Replaygain re-scan in winamp.

If there is any doubt about "over-gaining" files, you can always recalculate the gain in Winamp.

[Benski]

MP3Gain only has about 1.5dB of resolution in gain adjustment, so calculating Replay Gain should produce values in the range of -1.5dB to +1.5dB. Applying that Replay Gain on playback will still give a better experience than MP3Gain alone.

[J_Darnley]

To be very pedantic, Winamp should calculate gain in the range of -0.75 < x < 0.75 because MP3gain will see that being 0.5dB too quiet is closer than being 1.0dB too loud and vice versa. This is also the "average" gain because MP3gain does have a clip protect function, which I have turned on, so some files aren't amplified as much as they should be because of this.