Skip to main content

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: Does Normalization decrease the sound quality of an audio file? (Read 42618 times) previous topic - next topic
0 Members and 2 Guests are viewing this topic.

Does Normalization decrease the sound quality of an audio file?

Reply #25
With what settings would I get the loudest result without sacrificing quality (ie. clipping)?
What happens if I set the "with replay gain" setting higher than 89db?
What happens if I set the "without replay gain" setting at -1db or even slide it to +db levels?

You actually want to set the reference level lower than 89 to accommodate a library with a wider range of loudness. Anything that gets a positive Replay gain reading may potentially clip when played back with volume leveling enabled. The Replay gain specification suggests using 83. I personally use 85. 89 seems to work well for material only back to the 1990s.

MediaMonkey has a "clipping prevention" feature associated with its Replay gain implementation. It's not real clear how this works but I believe it lowers playback gain below the Replay gain suggestion if the Replay gain value would results in clipping.


In fact I don't listen to stuff back to back, one after the other. I listen to indivudal albums one at a time. In fact I don't much listen to musiuc through my computer, and I don't really care for having to adjust the volume continuously in between tracks.

What I want to be able to do is increase the overall volume level of indivudal albums for putting into my mp3 player. Without clipping and sacrificing quality, I want to be able to increase an albums overall sound. I mean I want to embed these replaygain values in the tracks forever. That's it.

So if you consider my questions against this background, what would you recommend me?

There's one other thing I don't understand. In Foobar2000, after scanning and updating the gain levels, I'm asked to set a value for "without gain" before the conversion process. I don't get it. These tracks have already been scanned and updated and became tracks "with gain" anyway, isn't it so? So the setting for "without gain" should be ineffective even if I set it to +1000db, since these tracks belong to the "with gain" category, ain't that true?

Does Normalization decrease the sound quality of an audio file?

Reply #26
Different editors may use terms like "normalize" differently.  In Audacity, the normalize function removes any DC bias, and re-balances the channels.  This could be useful in ripping vinyl as sometimes one channel is consistently a little louder than the other.
Amplify would be the function to scale both channels by the same amount.

Another thing to check with vinyl rips is whether the loudest part of the file is a click.  Remove that before scaling.

Does Normalization decrease the sound quality of an audio file?

Reply #27
Quote
In fact I don't listen to stuff back to back, one after the other. I listen to indivudal albums one at a time. In fact I don't much listen to musiuc through my computer, and I don't really care for having to adjust the volume continuously in between tracks.

What I want to be able to do is increase the overall volume level of indivudal albums for putting into my mp3 player. Without clipping and sacrificing quality, I want to be able to increase an albums overall sound. I mean I want to embed these replaygain values in the tracks forever. That's it.

So if you consider my questions against this background, what would you recommend me?
In that case, you don't want Replay Gain at all!  As I said before, RG tries to make all tracks (or albums) play with the same loudness...  It's NOT trying to maximize loudness!    And, since most tracks can't be increased without clipping, it works by reducing most tracks. 

Just normalize (maximize) all of your albums to 0dB  (some people recommend slightly below 0dB).    In other words, increase all of the tracks by the same amount.    This can be a little tricky if your album is separated into tracks.  The most straightforward way to do this is to re-combine all of the songs into a single WAV (or FLAC) file, normalize, and re-split the tracks.  (If you normalize the album as a whole, some tracks won't be normalized to 0dB...  Maybe only one track will be normalized to 0dB, but this is how it should be!)

Again, if these older recordings are more dynamic than your modern recordings, they can be normalized/maximized and they will never be as loud as the modern recordings.    In fact, your FLAC files may already be normalized and re-normalizing may do nothing.  (The only way to make them "sound louder" is to apply dynamic compression, which I don't recommend...  There's a reason you like that 70's sound, and the lack of over-compression is probably part of it!)

If the album is mastered properly, you shouldn't have to adjust the volume between songs.  Some songs are supposed to sound louder than others, some songs (even quiet-sounding tracks) will have higher peaks than others, etc.  If you normalize the tracks separately, you mess-up the relationship. 

Replay Gain has two modes/values.  If you choose "album gain", it will apply the same gain adjustment to all tracks.  This will retain the relative gain between tracks as originally mastered (assuming you haven't messed that up by normalizing the tracks individually). 

During playback in the "album" mode, Replay Gain will adjust the level once every time you start a different album.  The goal is to make all albums equally loud without adjusting the volume in the middle of an album and messing-up the way the album was produced.  When playing back in "track" mode, the level will be adjusted for each song.


Quote
There's one other thing I don't understand. In Foobar2000, after scanning and updating the gain levels, I'm asked to set a value for "without gain" before the conversion process. I don't get it.
I don't use foobar, but I if I understand correctly, I believe this is for playing tracks without replay gain.  If all of your tracks have replay gain, this setting won't do anything.  If you want to maximize your non-replay gained tracks and you want to avoid clipping, set this to 0dB (no change).

Does Normalization decrease the sound quality of an audio file?

Reply #28
So if you consider my questions against this background, what would you recommend me?

I went back to the beginning and see that you want to listen on public transport. The likely reason you cannot hear well is because the train/plane/bus is too loud. The best fix is to quiet the ambient noise it by using in-ear or noise cancelling headphones.

You can turn it up louder but you risk hearing damage with combined high ambient and headphone levels. Some devices just don't go up loud enough to overcome high level ambient noise and/or damage your ears.

If for some reason, you want to try to get around this, you'll want to apply some of the same dynamic range compression processing that the professionals use to make today's CDs sound so much louder than yesterday's. Technically speaking, this processing does not cause clipping but it definitely does decrease sound quality.

Does Normalization decrease the sound quality of an audio file?

Reply #29
Are there any decent multi-band DRC VST plugins available?

While it can decrease sound quality, when used carefully, DRC will not necessarily cause your stuff to sound bad.  I have a feeling some of these comments are the typical overly-polarized, knee-jerk reactions to the loudness war that we commonly see on this forum.

Does Normalization decrease the sound quality of an audio file?

Reply #30
I believe many audio editors have dynamic range processing facilities. I’ve used those of CoolEdit to remake a few CDs for in-automobile listening. Not knowing much about what I should do, I just played with some samples of wide range orchestral music until I had something that seemed halfway reasonable. It isn’t perfect, someone who knew where they were going could probably do much better, but it made a quite significant listening improvement under those noisy conditions.

Does Normalization decrease the sound quality of an audio file?

Reply #31
While it can decrease sound quality, when used carefully, DRC will not necessarily cause your stuff to sound bad.  I have a feeling some of these comments are the typical overly-polarized, knee-jerk reactions to the loudness war that we commonly see on this forum.

DRC is a type of distortion of the original recording. Information in the original recording is lost. I think it is OK to say that it "decreases sound quality". I agree that it doesn't necessarily "sound bad".

There are listening situations (e.g. transportation, exercise) where you need to reduce dynamic range to be able to enjoy the music. There needs to be a DRC dial on playback equipment. Why has this not happened?

Does Normalization decrease the sound quality of an audio file?

Reply #32
There needs to be a DRC dial on playback equipment. Why has this not happened?

I agree with the need. The most probable reason why a DRC dial has not been included on other than a minority of all playback gear is probably price/perforamnce.

DRC adjustments have been fairly common features of surround receivers, usually labelled something like "Midnight Listening" or some such.


Does Normalization decrease the sound quality of an audio file?

Reply #33
I don't think it is about price/performance. It's a simple DSP algorithm. Might be a usability issue. How do you explain to someone what DRC does. Can you get reasonable results with one or two preset modes?

The reason there is DRC on AVRs is because movie soundtracks actually have dynamics. The DRC rarely works in stereo mode. Popular music recording have all the dynamics squeezed out of a during mastering. DRC during playback is not going to do anything. Sadly I think I just answered my original question.

Does Normalization decrease the sound quality of an audio file?

Reply #34
I don't think it is about price/performance. It's a simple DSP algorithm. Might be a usability issue. How do you explain to someone what DRC does. Can you get reasonable results with one or two preset modes?


I don't know how complex of a box you have to build before the DSP resources to build a good DRC can have zero incremental cost. I think it would still be more than a basic MP3 player.


Quote
The reason there is DRC on AVRs is because movie soundtracks actually have dynamics. The DRC rarely works in stereo mode. Popular music recording have all the dynamics squeezed out of a during mastering. DRC during playback is not going to do anything. Sadly I think I just answered my original question.


Perhaps we should put ourselves in the position of people who actually have to sell product.  I think we end up with a chicken and egg problem. The manufacturer of the digital player questions whether adding a DRC facility will sell more players or allow him to get more for each player that is sold. The producers of  recordings question whether they can trust the DRC hardware that is out there to optimize the sales of his recordings should he produce them uncompressed.

Does Normalization decrease the sound quality of an audio file?

Reply #35
I think it is OK to say that it "decreases sound quality".

...and it is every bit as OK to say that it can be used to increase sound quality.  Personally, I would avoid making assertions about sound quality, especially considering the topic being discussed.

Does Normalization decrease the sound quality of an audio file?

Reply #36
I think it is OK to say that it "decreases sound quality".

...and it is every bit as OK to say that it can be used to increase sound quality.  Personally, I would avoid making assertions about sound quality, especially considering the topic being discussed.

Point taken. I was trying to adhere to the OP's language. Perhaps the original question needs to be rephrased.

Does Normalization decrease the sound quality of an audio file?

Reply #37
On Audacity if you use "Amplify" it'll raise the levels of the track to the maximum level WITHOUT clipping. Should be a quick and easy way to normalize, I suppose.

Does Normalization decrease the sound quality of an audio file?

Reply #38
Did the OP miss this...

Normalization:
- Usually useless (Cd's, for example, tend to be normalized already)

...because it's true. The example the OP posted had a peak at 0.92. Raising this to 1.00 isn't going to make much audible difference (if any).

DRC the old tracks, or get a louder mp3 player, or more sensitive earphones, or reduce the background noise.

DRC is the cheapest option, but if you're picky, it probably won't satisfy. If it was possible to take a 1970s style master, run it through a one-click free batch process, and make it as loud as a 2010 master, then people wouldn't pay tens or hundreds of thousands of dollars for equipment to make 2010-style masters. ;)

You might be able to get close enough. There are loads of DRC plug-ins out there. On the more complex ones, you could spend half a life-time getting the settings right. e.g. try http://www.stereotool.com/

Cheers,
David.

Does Normalization decrease the sound quality of an audio file?

Reply #39
There is a plugin called vlevel for foobar which works quite nicely as DRC. The settings I use are: strength 1.0, multiplier 2, buffer 5 seconds. (not sure if vlevel works with the latest version of foobar, I use 9.6.4) This DSP can be used for file conversion as well.

Does Normalization decrease the sound quality of an audio file?

Reply #40
Is vlevel a multi-band compressor?  Depending on how much compression you need, anything less might not be able to get the job done without creating a sonic mess.

Does Normalization decrease the sound quality of an audio file?

Reply #41
I tend to agree with greynol, and can suggest Stereo Tool as a free multiband compressor which I have at least toyed around with, and seems to work.

Does Normalization decrease the sound quality of an audio file?

Reply #42
vlevel

Basically, vlevel is just a volume normalizer, it doesn't color the sound in any way other then volume changes.