http://www.geocities.com/mp3gain

Finally!
More of a milestone for me than anything really useful for you...

Version 1.0 (final) of MP3Gain is out. Nothing functionally different in the GUI.
The back end has a little tweak or two, but I don't have my notes on me right now and can't remember what it is...

EDIT: Ah, yes. I fixed a bug that showed the incorrect progress percentage for files over 21MB.

The other important thing (but only for Windows users, sorry), is that the back end now accepts wildcards but not recursion. So if you do "mp3gain Music\*.mp3" it will do all the mp3 files in the "Music" folder, but not in the "Music\Disco" folder.

It turns out that if you link to a different library, Windows will automatically expand wildcard command-line arguments for you. I didn't have to change my code at all. Big thanks to... er, dangit, I'm on the wrong computer again... the guy who pointed this out to me whose name I don't remember but it's in my archived e-mail on my other machine

For you Linux (and Mac) users, wildcard filenames and directory recursion are near the top of my to-do list.

I've applied VorbisGain to all my vorbis files, and I've played around with Foobar's ability to write replaygain tags to mp3s. Foobar is great and all, but I listen to my music in both Windows and Linux, and I like the portability of id3 instead of APE2.0 tags for my mp3s. Is MP3Gain the way for me to go? Does it have any drawbacks? Or do the Hydrogen Audio masses (that's you!) have some other preferred way of adjusting volume for their mp3s?

Snelg -

To have a solution we have to define the problem. You did yours well and found a simple, sound solution. I enjoy the product and use it on all my MP3's. Well concieved, well executed. Take a bow.

And thank you!

any chance of adding support in mp3gain for reading the ape2 tags that foobar uses, in order to skip the analyzation step that's already been done?

Now that would be neat.

Using Tag as an ape2 reading backend, using the mp3gain backend to modify files, and putting on new ape2 tags using Tag again..

ape2 tags just store relative gain values, so the original values will turn out useless if the original mp3's volume gets changed, hence the new ape2 tags.

edit: typos.

That's the plan, to make MP3Gain's use of APEv2 tags match foobar2000's use. In addition, MP3Gain will also have another piece of information keeping track of what changes have actually been applied to the file, so it can undo the changes if desired.
All this information will optionally be stored in ID3v2 tags also/instead.

-Glen

The Normal Installer warned me that the vb runtime files were out-dated, but I went in installing Mp3Gain anyway and it worked fine. Do I really need the vb runtime from the MP3Gain site?

Nope. If the GUI starts, then everything else should be okay.

Oh yeah, I forgot to mention the updates to the back end. I'll go back and edit the initial post now...

-Glen

Hello...

Do you think that you can manage to keep your promise? With promise I mean the (quote from a old mail to you) ...



You've replied that you will add it among the first things after the 1.0 version .

If you could add this in the next version it would be perfect... Anyway, thanks for your gem...

Cheers,

Mijo.

Just something I'd like to see in MP3Gain, and I know some others would too, is a display of the number of clipped samples, and possibly the resultant number after gain is applied...

The one-button "Track Gain but don't clip any files" function is still high on my "To Do" list.
But because it is currently possible to do this (even though it takes a little more than one click), I am first working on something that is not possible with the current program: storing "undo" information and analysis results in APE tags in the mp3 file itself.

Quick instructions for anyone who hasn't figured out the easiest way to do "Track Gain but don't clip" for a group of files:

• Load files into MP3Gain
  
• Do "Track Analysis"
  
• Sort by the "clip(Track)" column (by clicking on that column header)
  
• Shift-select all of the files that do not have a "Y" in the "clip(Track)" column
  
• Right-click in the list box and select "Apply Track Gain" (this will apply the suggested Track gain only to the selected files)
  
• When Track Gain is finished do "File > Invert Selection" (or just press Ctrl+V). Now all the the files that do have a "Y" in the "clip(Track)" column are selected
  
• Right-click in the list box and select "Apply Max No-clip Gain for Each file"

As for the "number of clipped samples" request, that's still a bit lower on my To Do list. Not much lower, but it's less frequently requested than storing undo info and the "Apply Gain but don't clip" functions.

And when I do add the # of clipped samples in, it will only be the current number of clipped samples. To show the number of samples that would clip for other gains, you'd have to do a noticable amount of extra computation for every single sample as it's decoded. Basically, you'd have to say,

"Is the sample > 32767?
Is (sample * 1.189) > 32767?
Is (sample * 0.841) > 32767?"

And that's just checking for gain +/- 1.5dB.

Just the number of CURRENT clipped samples would make me a very happy camper. One good reason for this is a situation I had a few days ago. I encoded an old Beatles album. The album volume was somewhere around 86 dB. But I couldn't raise it without clipping. So I raised it to 92 db anyway. Then I decoded with MAD. According to MAD, only one sample was clipped.

I just think that some people would take a tradeoff of a few clipped samples to be able to raise the percieved volume. After all, how many of us can hear 1, 2, or 50 samples (and that's provided they're consecutive).

Thanks for the great program Glen

Thanks for your clarification! BTW, I know there is a MP3Gain FAQ here, so if it isn't already in it, maybe someone could add your below hint to it.

All the best and thanks for your great work! I really appreciate it.

Cheers,

Mijo.