Hi guys, I'm new on this useful forum. :) I have a question for you and for expert. I always used MP3Gain to avoid clipping mp3. I usually rip from original cd, but I noticed that some volume tracks of album are too low, near 85db respect others tracks that are on 95db for example. I tried to go up and I noticed that I can raise up to 87db without clipping (analyzing with MP3Gain). I can safely go up or it's better to leave the tracks with default db?
ReplayGain/MP3Gain attempts to match the loudness of all tracks/albums, so you shouldn't have to monkey with the settings, just use the same setting for all of your MP3s. If some songs are still louder/quieter after applying MP3Gain, you can adjust them manually (by ear) as long as you don't run into clipping.
In album mode all songs are adjusted by the same amount so loud songs remain (relatively) loud and quiet songs remain (relatively) quiet as intended by the producer.
I tried to go up and I noticed that I can raise up to 87db without clipping (analyzing with MP3Gain). I can safely go up or it's better to leave the tracks with default db?
If you
raise the target volume you are leaving MP3Gain
less room to work (unless you allow clipping).
For example, if you set the target high-enough all of your tracks will be normalized/maximized as loud as they can go without clipping and there will be no "volume matching" at all. (Perceived loudness does not correlate well with peak levels.)
Many (most) songs are already normalized/maximized, including the quiet-sounding songs. That means the only way to match volumes (without clipping) is to reduce the level of the loud songs.
I understand. I usually use the option Maximizing + Apply Max No-clip Gain for Each file to avoid clipping with -1.5 or 3db and leave the quiet songs that don't clipping. So you think is it better to leave quiet songs as they are even if I can raise it without clipping, right?
And is it better the way of Apply Max No-clip Gain for Each file or Apply Max No-clip Gain for Album? I noticed that the second choice decrease more (not too much) volume output.
Nobody guys?
Well, the peaks of lossy files can be a lot higher than the original source. Even much higher than the ceiling. I'm not sure why, but that's how it is. So if you use no-clip the volume will decrease a lot. If you also use albumgain it will look at the highest peak of the entire album and then decrease volume of all tracks accordingly. Which will be even lower tham trackgain.
I got it. I noticed that there's no much margin of differences between Apply Max No-clip Gain for Each file and Apply Max No-clip Gain for Album. So I think it's better to decrease the right to avoid clipping with Apply Max No-clip Gain for Each file right?
Myself I'd go for the EBU R 128 standard (-23 LUFS with -2 dBFS True Peak).
This is the same as the ReplayGain in recent versions of foobar2000 with a ReplayGain adjustment of -5 dB, and Peak scan oversampling of 4x.
Calibrate your listening setup to this and you'll be able to go from watching a movie to listening to a rock track without having to touch your volume because your ears hurt. Do note you may have to turn up the volume knob on your system a lot more than currently.
Sorry but your answer is not what I was looking for...
No more answers guys?
You are being told what each thing means and what effect it has.
But you seem to ask for opinions, which is why you don't get the right answer.
Also, nowadays I think that most people don't use MP3Gain, and instead use other replaygain players that can use ReplayGain 2.0/EBU128.
Of course, MP3Gain still has a place for generic players (like in a car, smart TV, default smartphone players...), but even then, the same action (modifying file gain) could be done with other software.
So...
- Lossy codecs (MP3 being a lossy codec) generate peaks above full scale.
- Peaks in general do not correlate directly with loudness.
- Album gain is the mean value of the track gain of all tracks that represent the album. Album peak is, obviously, the highest peak.
- If you are used to listen to MP3s album by album, instead of randomizing tracks, album gain is preferred to track gain so that it maintains the original differences of the album, while making the tracks play near the desired loudness.
- If you are used to listen to MP3s randomly, or have many single tracks, you might prefer to use track gain so that the loudness of the tracks is the desired one (algorithm weak points aside).
Replaygain aware players can even be smart and know if they have to use album gain or track gain, but with MP3Gain, you're set with a setting and you're done.
I used to prefer track gain, except for gapless albums in which, of course, I used album gain.
As far as clipping prevention, I would not use it with MP3. If one track has a high peak, you might opt to lower the gain target for that specific track but in general, clipping prevention on MP3 causes MP3Gain to fail to set the desired gain on low to moderate volume tracks, and the effects of that clipping is rarely a problem.
You are being told what each thing means and what effect it has.
But you seem to ask for opinions, which is why you don't get the right answer.
Also, nowadays I think that most people don't use MP3Gain, and instead use other replaygain players that can use ReplayGain 2.0/EBU128.
Of course, MP3Gain still has a place for generic players (like in a car, smart TV, default smartphone players...), but even then, the same action (modifying file gain) could be done with other software.
So...
- Lossy codecs (MP3 being a lossy codec) generate peaks above full scale.
- Peaks in general do not correlate directly with loudness.
- Album gain is the mean value of the track gain of all tracks that represent the album. Album peak is, obviously, the highest peak.
- If you are used to listen to MP3s album by album, instead of randomizing tracks, album gain is preferred to track gain so that it maintains the original differences of the album, while making the tracks play near the desired loudness.
- If you are used to listen to MP3s randomly, or have many single tracks, you might prefer to use track gain so that the loudness of the tracks is the desired one (algorithm weak points aside).
Replaygain aware players can even be smart and know if they have to use album gain or track gain, but with MP3Gain, you're set with a setting and you're done.
I used to prefer track gain, except for gapless albums in which, of course, I used album gain.
As far as clipping prevention, I would not use it with MP3. If one track has a high peak, you might opt to lower the gain target for that specific track but in general, clipping prevention on MP3 causes MP3Gain to fail to set the desired gain on low to moderate volume tracks, and the effects of that clipping is rarely a problem.
Ook Jaz, thank you for your opinion. And what do you think about "Don't clip when doing Track Gain" under Options menu?
I indirectly said that in my reply:
Given that the album peak is the highest of the track peaks and that the album gain is the mean of the track gains, it means that clipping prevention based on album gain could reduce the target gain more than what it would for a single track. (because the gain will be reduced to all tracks to accommodate for the highest track peak).
And I already said that I don't use clipping prevention.
Oook! Now I understand! Thank you so much for being exhaustive!! ;) ;) ;)
Hello, does anybody in this forum know what default value is to be taken for 'size' when plotting spectrum in frequency analysis in Audacity 2.0 software , thanks ::)