Work at SourceForge, help us to make it a better place! We have an immediate need for a Support Technician in our San Francisco or Denver office.
I have been using MP3Gain for a long time. I happened to notice something with a particular song. After adjusting to 92db with MP3Gain, the bass doesn't have the "punch" that is present in the original file. If MP3Gain only tweaks the volume, then I should be able to make the two files sound identical by adjusting my volume control. But if I turn up the volume on the MP3Gain'ed song, it still doesn't sound as good. When comparing with Audacity, the bass peaks seem to be reduced a lot more than the average music content. Here are the two files:
It *does* only tweak the volume. If I take your original, mp3gain it, then use foobar to dump both to wav using replaygain info, this gives me two files that are the same volume. If I take one and invert it, and mix it with the other, this should give me a silent file. Which, it does, except for what I presume is foobar's random dither noise around the 0 crosspoint. This is all exactly as expected. However, if I do the same test using your original and normalized mp3s instead, I get tons of leftover data. I don't know what happened to your normalized one, but it does not appear to be simply mp3gaining.
Here's the same statistics view for your original vs. your normalized mp3.
As you can see, there is still a LOT of data there. In fact, I can still hear and make out the song.
Thanks for once again jumping in to explain stuff, shorty-dammit.
I took a look at the files, too, and "normalized" is definitely _only_ mp3gain-ed. My guess about your data difference is that the "target" volume for the original normalized.mp3 is around 92dB, not the default 89. With the target at 92, when you do Track Gain on "original.mp3", the end result is precisely the same as "normalized.mp3".
So back to the actual question:
exar50, since you're already familiar with Audacity, it's relatively easy to show that the changes made by mp3gain only affect the amplitude.
Just open "normalized.mp3" and do "Effect - Amplify…" to increase the amplitude by 6 (or 6.02 if you want to be even more precise). You'll have to check the "Allow clipping" box. The end result will sound identical to "original.mp3"
Thank you for all the analysis, it helps a lot. I thought I was going to have to dump MP3Gain, which I didn't want to do because I really like it. I feel better now!
I didn't mention that my comparison was NOT done by listening directly to the MP3 files. Instead I burned the songs to CD and played it in my car. So now I suspect my CD burner program. I am using Acoustica MP3 CD Burner. I'll have to check all the settings and do some experimenting.
Strange. I wonder where I goofed up then. Was kinda late, hehe. I *thought* I loaded both up in foobar and did a replaygain scan of both in their current state, so that any volume differences would be accounted for, and then dumped to wav, inverted, and mixed. Should have been basically zero left in both cases then. Even with one already having been through mp3gain, rescanning both for the same target should have nullified itself. At any rate, yeah, must be something else causing any sound changes. :)
snelg, you are correct. I boosted the mp3gained file by 6 and it perfectly matches to the original. Thank you