Help - Search - Members - Calendar
Full Version: Volume level for quiet MP3 files
Hydrogenaudio Forums > Lossy Audio Compression > MP3 > MP3 - General
snverhallen
Hi there,

Excuse me for my amateur question. I have a few live MP3 files that are on the quiet side, and I need to edit them so that the master volume level is a bit louder. I have looked into apps such as iVolume and MP3Gain, but I believe I'm looking for something else. All I need to do is change the master volumes for those files, I don't want to change my whole library and I don't need all of them at the same level etc.

Thanks in advance,

snverhallen
kornchild2002
MP3Gain can increase the volume of specific files, not your entire library. In fact, you can use MP3Gain to increase the volume of certain files at one level and other files at a different level. foobar2000 can also adjust RaplayGain values for each file as well.
2E7AH
you can try mp3DirectCut or mpTrim, although i don't know what issue you have with mp3Gain
snverhallen
QUOTE (kornchild2002 @ Mar 4 2009, 17:37) *
MP3Gain can increase the volume of specific files, not your entire library. In fact, you can use MP3Gain to increase the volume of certain files at one level and other files at a different level. foobar2000 can also adjust RaplayGain values for each file as well.


OK, and will it edit the music file and not the iTunes Library; because I need to post these live songs on the internet and they need to be at the same volume level as I edit them to. And the MP3Gain your talking about is this one right: http://homepage.mac.com/beryrinaldo/AudioTron/MacMP3Gain/

I'm on a Mac by the way.
tpijag
No they are surprisingly talking about the program that is actually called mp3gain wink.gif
Windows based.

terry

Your mac version may / maynot have all the same options.
snverhallen
Just used MP3Gain on an MP3 and it didn't change the volume at all. That's why I just need a plain and simple way of changing the volume with a simple slider thing. Like in iTunes with the EQ you can edit the Preamp setting, this is exactly what I need, except the settings need to be saved onto the file.
kornchild2002
Correction: you aren't using MP3Gain. MP3Gain is a Windows program but there is a Linux GUI. You can learn more about MP3Gain here. So please don't refer to the software you are using as MP3Gain since it isn't the same software I am talking about. The program you linked us to is supposed to be the Mac alternative for MP3Gain. Are you playing the files (or have them highlighted) in iTunes while you are gaining them?
snverhallen
QUOTE (kornchild2002 @ Mar 4 2009, 18:02) *
Correction: you aren't using MP3Gain. MP3Gain is a Windows program but there is a Linux GUI. You can learn more about MP3Gain here. So please don't refer to the software you are using as MP3Gain since it isn't the same software I am talking about. The program you linked us to is supposed to be the Mac alternative for MP3Gain. Are you playing the files (or have them highlighted) in iTunes while you are gaining them?


No, they are not even in iTunes. I think I have a solution: if I use Audacity 1.2.6a and use the Amplify feature, will that cause the MP3 to lose audio quality?
timcupery
You need to use mp3gain or a program like it that can edit the volume of mp3 files, in increments of 1.5 dB, by changing the volume value on each mp3 frame (1152 samples, 0.026 seconds) in the file.
You'll want to avoid clipping, or at least avoid much clipping. If the maximum volume level of a given file is 0.84, for example, you could raise it 1.5 dB to around 0.99. If you go over 1.0, the any part of the top of the waveform beyond 1.0 will be chopped off and truncated at the maximum of 1.0
The usefulness of mp3gain is that it can first calculate values for the file, and you can find the maximum amount you can raise the volume without clipping.
You can set up a Windows virtual environment on a mac, in order to run windows programs such as mp3gain or foobar2000.

edit: also, I'm not sure if audacity is the program you want to be using here. I think that Audacity can make changes to a waveform, but then re-encodes afterwards. (I'm not totally sure on this though.)
What you want is a program that can edit an mp3 file without decoding and then re-encoding again afterwards, which results in quality loss.
greynol
I think it's safe to say that any editing you do an mp3 file in a wave editor will require re-encoding when going back to mp3 format.

<rant>
There is an unreasonable amount of paranoia regarding clipping as an artifact. So far not a single person has reported being able to hear clipping as a result of the lossy encoding process using settings that would otherwise be transparent.
</rant>

When you're boosting the level of an mp3 beyond that of the original source, things are obviously different. However, I would only worry about clipping if it's audible. Compression/limiting of the original source prior to encoding would be the best option, if the original source is available.

This begs the question: why not use replaygain/mp3gain as it was intended? In most cases loud tracks (or albums) will be made quieter to match other albums which may or not need adjustment. Volume reductions will never result in additional clipping.
snverhallen
Thanks for all the info. I tried using Fission, but when I tried to increase the DB level it said it was already at the highest, which I find strange. I'm still amazed that there is no app that just raises the DB level like the EQ in iTunes. Oh well.

http://rogueamoeba.com/fission/ See the second screenshot, that is obviously the DB changer, and for me the slider was already at the top.
timcupery
@greynol - I meant to distinguish between standard wav-editing programs, and programs that allow you to edit the lossy audio stream but without requiring re-encoding. mp3directcut, and apprently fission as well. Such programs can even do graduated fades (limited by frames and 1.5 dB increments of course). I just wasn't sure if Audacity had the capability to edit without re-encoding - I think it doesn't.

thanks for the rant about clipping. it's good to know that this isn't very relevant from an audio-quality perspective.

I agree with you that it's better to use replaygain as it's intended, which means lowering the volume of most tracks; I use the same approach that you do.

however, even this way there are some albums, mastered at a low volume with lots of dynamic range, where the volume is significantly lower than replaygain would ideally place it. in these cases, I will raise the volume, but have worried about clipping. It's good to know that's less of a worry.

@snverhallen - there are plenty of programs for Windows that can raise the volume of mp3 files (and now AAC files too) - mp3gain, mptrim, mp3directcut come to mind.
I just don't know Mac programs.
greynol
For tracks where RG values are positive, clipping can be audible and has been reported as audible by various people. My rant was about situations where mp3gain reports the file as clipping before gain changes have been made, IOW, clipping that results from the lossy encoding process. Sorry for not being clear earlier. Hopefully this is better. If not, let me know.

Point taken about editing software. I have my doubts that Audacity is able to make adjustments to the global gain field. I'm curious to know what an mp3 file looks like when opened up in Fission. Does is look like a typical wave file, or does it look more like what is displayed in mp3directcut?
ameyer17
Actually, udial.wav. Strangely, transparent (to my ears, using my equipment, etc) in aoTuV ogg vorbis at -q3, but not -q5, -q6, -q7, -q8, -q9, or -q10. I've heard reports of similar behavior in other lossy formats.
EDIT: ABX logs in http://www.hydrogenaudio.org/forums/index....mp;#entry608477
EDIT2: If you mean on actual music or other samples that haven't been engineered to create clipping, I agree with you, though.
EDIT3 (Moderation): Removed your usless quote of my previous post.
timcupery
QUOTE (greynol @ Mar 4 2009, 20:07) *
For tracks where RG values are positive, clipping can be audible and has been reported as audible by various people. My rant was about situations where mp3gain reports the file as clipping before gain changes have been made, IOW, clipping that results from the lossy encoding process. Sorry for not being clear earlier. Hopefully this is better. If not, let me know.

Point taken about editing software. I have my doubts that Audacity is able to make adjustments to the global gain field. I'm curious to know what an mp3 file looks like when opened up in Fission. Does is look like a typical wave file, or does it look more like what is displayed in mp3directcut?

For tracks where RG is positive and really should be louder, but where the wav file is already close to 100%, I'll usually do some sort of compression prior to encoding. I usually do this manually, just using a wav editor to search for peak sections and lowering all of the peaks to a certain level. This difference is not audible (at least to my ear on ABX) and allows clip-free encodes.

edting software: mp3directcut is a nice piece of software, and I appreciate it, but it's really rudimentary.
theoretically, there's nothing to prevent a really nice of mp3-editing software from being programmed. That shows a display like Audacity or EAC's wav editor or whatever, and shows the boundaries of frames, etc. It would have to do a full decode of the sample-values, but that wouldn't be inordinate overhead.
I saw the Fission screenshots, and it looked kinda like Audacity, but I also wondered whether it would look like this when an mp3 was opened for non-transocde-ish editing.
lvqcl
QUOTE (ameyer17 @ Mar 5 2009, 05:21) *
Actually, udial.wav. Strangely, transparent (to my ears, using my equipment, etc) in aoTuV ogg vorbis at -q3, but not -q5, -q6, -q7, -q8, -q9, or -q10. I've heard reports of similar behavior in other lossy formats.

What is really strange for me is testing lossy codecs with udial sample. huh.gif
2Bdecided
QUOTE (greynol @ Mar 4 2009, 18:50) *
So far not a single person has reported being able to hear clipping as a result of the lossy encoding process using settings that would otherwise be transparent.
Now there's a challenge! Not one I'm about to rise to.

I assume you know that "good but probably not transparent" settings produce audible clipping? The most obvious link is here:
http://www.ff123.net/norm.html

FWIW the kind of mastering that produces this issue usually wrecks the sound so much that a little extra clipping is the least of the track's problems. But it might upset someone.

Cheers,
David.
ameyer17
QUOTE (lvqcl @ Mar 5 2009, 09:03) *
QUOTE (ameyer17 @ Mar 5 2009, 05:21) *
Actually, udial.wav. Strangely, transparent (to my ears, using my equipment, etc) in aoTuV ogg vorbis at -q3, but not -q5, -q6, -q7, -q8, -q9, or -q10. I've heard reports of similar behavior in other lossy formats.

What is really strange for me is testing lossy codecs with udial sample. huh.gif

True.
To be fair, it was because someone else claimed there was a problem without providing ABX logs and I stupidly didn't believe him/her.
greynol
Here's an example of how I think samples should be produced for such testing:
http://www.hydrogenaudio.org/forums/index....st&p=558514
uart
Ok could someone please listen to these two mp3 samples. The first one is the original and the second one was mp3gained to 105dB.

Original mp3
Click to view attachment


After being mp3gained to 105dB
Click to view attachment

Uploaded HERE. Make sure you listen without applying any replaygain!


That second one doesn't sound quite right, what went wrong? Hey all I wanted to do was to make all my songs the same level as my loudest song. Greynol said it would be ok tongue.gif (joking)
greynol
I hope you really are joking, otherwise I think you need to go back and read what I wrote.

QUOTE (greynol @ Mar 4 2009, 17:07) *
My rant was about situations where mp3gain reports the file as clipping before gain changes have been made

[JAZ]
QUOTE (lvqcl @ Mar 5 2009, 17:03) *
QUOTE (ameyer17 @ Mar 5 2009, 05:21) *
Actually, udial.wav. Strangely, transparent (to my ears, using my equipment, etc) in aoTuV ogg vorbis at -q3, but not -q5, -q6, -q7, -q8, -q9, or -q10. I've heard reports of similar behavior in other lossy formats.

What is really strange for me is testing lossy codecs with udial sample. huh.gif



That is because of the filter. the higher tone is filtered in -q 3, and that's the one that can clip, because it's the louder one.

QUOTE
I have my doubts that Audacity is able to make adjustments to the global gain field.

No, it does not. Audacity always work with decoded data, and exports to whatever format you specify.

QUOTE
theoretically, there's nothing to prevent a really nice of mp3-editing software from being programmed.


What do you have in mind? What's the point of seeing the whole audio wave if you can only change its volume? Explain that "nice mp3-editing software"?
uart
Yes I understood what you wrote and I know your intensions, especially after you clarified it. However last time someone was enquiring about mp3gain-ing their collection to match their loudest songs (because of the usual newbie complaint that replaygain makes all their songs too quiet) I warned against exactly this type of clipping and you came along with the same rant. You have to understand that this is exactly the thing that many of these newbies actually have in mind when they start talking about wanting to make their songs louder with mp3gain. So yes it's kind of a joke, but not completely a joke, it has a moral.
timcupery
QUOTE ([JAZ] @ Mar 5 2009, 13:49) *

QUOTE ([timcupery])
theoretically, there's nothing to prevent a really nice of mp3-editing software from being programmed.

What do you have in mind? What's the point of seeing the whole audio wave if you can only change its volume? Explain that "nice mp3-editing software"?

The usefulness is that you can see what you are editing more easily. Yes, obviously you can only apply volume-based edits, or cutting and pasting, on a frame-by-frame basis, and this is even limited depending on the degree of frame-dependence of the mp3 in question. However, greater visibility of the audio data that I'm working with would be really nice. Mp3directcut doesn't give an ideally clear picture here, partly b/c you can't zoom in very much.
Maggi
QUOTE (kornchild2002 @ Mar 4 2009, 18:02) *
Correction: you aren't using MP3Gain. MP3Gain is a Windows program but there is a Linux GUI. You can learn more about MP3Gain here. So please don't refer to the software you are using as MP3Gain since it isn't the same software I am talking about. The program you linked us to is supposed to be the Mac alternative for MP3Gain. Are you playing the files (or have them highlighted) in iTunes while you are gaining them?


I don't see a reason why it should not work using MacMP3Gain, at least the very first sentence on that site clearly states:

QUOTE
MacMP3Gain is an AppleScript Studio application which brings an Aqua GUI to the command line version of mp3gain, ...


so it's a Mac-GUI for MP3Gain and since the source code for MP3Gain is available, I'd presume the author of MacMP3Gain compiled the commandline version on a Mac and built a GUI around it.

Vice versa, on the MP3Gain download page you'll find

QUOTE
The MacMP3Gain page has information about a Macintosh version of MP3Gain.


Maybe it's just the player that cannot benefit from those RG tags or isn't configured to use it ?

Or am I just missing something here ?

Cheers,
Maggi
This is a "lo-fi" version of our main content. To view the full version with more information, formatting and images, please click here.
Invision Power Board © 2001-2009 Invision Power Services, Inc.