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: --scale vs. mp3gain (Read 12742 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

--scale vs. mp3gain

Well, as I said it in the title, I am interested which of these two (the --scale switch and the mp3gain) is better for normalization of mp3 files? And what exactly is the difference which makes either one better?
Only the best is good enough.

--scale vs. mp3gain

Reply #1 –
Well, the thing is that you actually don't normalize your MP3's through either --scale switch or MP3Gain.

I think that those 2 methods of modifing the MP3 is used to prevent them from clipping.

The differences:

--scale switch is used when encoding and MP3Gain you apply after encoding.
By using --scale your resulting MP3's should be a bit smaller because you can only lower the --scale switch AFAIK.
By using MP3Gain your MP3's will actually sound a way quiter than the original if you apply gain of -1,5db or more so I can say that --scale is more precise.
The problem is that you can't decide if the --scale 0,9X is enough to prevent clipping...

I hope it helps...
Sorry for my poor English, I'm trying to get better... ;)
"The greatest trick the Devil ever pulled, was convincing the world he didn't exist."

--scale vs. mp3gain

Reply #2 –
You can use Wavegain to scan your files, not applying the results, but just taking the scale value Wavegain will give you.  After pass the value given by wavegain into your --scale x.xxxx infos

Soren

--scale vs. mp3gain

Reply #3 –
Quote
You can use Wavegain to scan your files, not applying the results, but just taking the scale value Wavegain will give you.  After pass the value given by wavegain into your --scale x.xxxx infos

Soren
[{POST_SNAPBACK}][/a]


If your response was addressed to me you should know that I want to have the MP3's as close to the original without clipping that was introduced during the encoding. So no scaning can help me with this

These issues were already discussed here in [a href="http://www.hydrogenaudio.org/forums/index.php?showtopic=10637&view=findpost&p=248330]Wavegain vs. MP3Gain[/url], Clipping! and mp3gain threads and I'm quite sure that nothing new will be introduced to me

Edit: I think that this link could be quite useful for many people...
Sorry for my poor English, I'm trying to get better... ;)
"The greatest trick the Devil ever pulled, was convincing the world he didn't exist."

--scale vs. mp3gain

Reply #4 –
Neither --scale nor MP3Gain will have such adverse effects as k.eight.a would have you believe.

MP3Gain is better because you have more freedom in how to adjust the volume. You can also see what you're doing.

--scale doesn't actually "normalize" MP3's per se, but rather it attenuates them.
What I mean is, say you use "--scale 0.97." This will not normalize the MP3's you're creating to 97%, but rather it will make them all 3% quieter than their input files.

--scale vs. mp3gain

Reply #5 –
mp3Gain is better since you'll never know what --scale you have to use to prevent clipping. In addition, mp3Gain allows you to undo your changes. If you want to change the volume later you have to use mp3Gain anyway (unless you want to encode the song again  ). Btw. mp3Gain is lossless...
--alt-presets are there for a reason! These other switches DO NOT work better than it, trust me on this.
LAME + Joint Stereo doesn't destroy 'Stereo'

--scale vs. mp3gain

Reply #6 –
Quote
--scale doesn't actually "normalize" MP3's per se, but rather it attenuates them.
What I mean is, say you use "--scale 0.97." This will not normalize the MP3's you're creating to 97%, but rather it will make them all 3% quieter than their input files.
[a href="index.php?act=findpost&pid=261165"][{POST_SNAPBACK}][/a]

...I know it!
Quote
mp3Gain is better since you'll never know what --scale you have to use to prevent clipping. In addition, mp3Gain allows you to undo your changes. If you want to change the volume later you have to use mp3Gain anyway (unless you want to encode the song again  ). Btw. mp3Gain is lossless...
[a href="index.php?act=findpost&pid=261181"][{POST_SNAPBACK}][/a]

This is true, but one step down to prevent clipping is a way quieter...
I think it was Precisionist who said that when you apply -1,5 db to a file which is near full scale the result is about 70% of the volume AFAIK, so MP3Gain is not so precise...
PS: I know that I'm exchanging normalization and ReplayGain values (because I'm not educated in this territory) but I hope you understand what I mean...
Sorry for my poor English, I'm trying to get better... ;)
"The greatest trick the Devil ever pulled, was convincing the world he didn't exist."

--scale vs. mp3gain

Reply #7 –
Quote
I think it was Precisionist who said that when you apply -1,5 db to a file which is near full scale the result is about 70% of the volume AFAIK, so MP3Gain is not so precise...
[a href="index.php?act=findpost&pid=261194"][{POST_SNAPBACK}][/a]

70% seems to be a lot to me...1,5 db is hardly noticeable.
A quote from mp3Gain's FAQ:
Quote
The bad news: MP3Gain can only adjust the volume of your mp3 files in steps of 1.5 dB.

The good news: 1.5 dB is a small enough step for most practical purposes. Most humans can just barely hear a volume change of 1 dB.

I can't see why anyone would use --scale instead of mp3Gain. There are way too many advantages that mp3Gain offers and clearly makes up for that little 1.5 db tradeoff, which I don't consider a problem at all. The only purpose of --scale that I can see is for people that can't use mp3Gain or people that want to apply the same volume to all their files - no matter if it still clipps or not, and are not interested in AlbumGain or getting the max volume without clipping, and will never change the volume again...
--alt-presets are there for a reason! These other switches DO NOT work better than it, trust me on this.
LAME + Joint Stereo doesn't destroy 'Stereo'

--scale vs. mp3gain

Reply #8 –
Quote
Quote
--scale doesn't actually "normalize" MP3's per se, but rather it attenuates them.
What I mean is, say you use "--scale 0.97." This will not normalize the MP3's you're creating to 97%, but rather it will make them all 3% quieter than their input files.
[a href="index.php?act=findpost&pid=261165"][{POST_SNAPBACK}][/a]

...I know it!
Quote
mp3Gain is better since you'll never know what --scale you have to use to prevent clipping. In addition, mp3Gain allows you to undo your changes. If you want to change the volume later you have to use mp3Gain anyway (unless you want to encode the song again  ). Btw. mp3Gain is lossless...
[a href="index.php?act=findpost&pid=261181"][{POST_SNAPBACK}][/a]

This is true, but one step down to prevent clipping is a way quieter...
I think it was Precisionist who said that when you apply -1,5 db to a file which is near full scale the result is about 70% of the volume AFAIK, so MP3Gain is not so precise...
PS: I know that I'm exchanging normalization and ReplayGain values (because I'm not educated in this territory) but I hope you understand what I mean...
[a href="index.php?act=findpost&pid=261194"][{POST_SNAPBACK}][/a]


Human hearing is on a roughly exponetial scale covering ~12 order of magnitude.  1.5dB is barely perceptable, and as I recall its actually less then the MP3 encoder itself can round the volume by, so I don't see the point in even mentioning it.

--scale vs. mp3gain

Reply #9 –
Quote
Human hearing is on a roughly exponetial scale covering ~12 order of magnitude.  1.5dB is barely perceptable, and as I recall its actually less then the MP3 encoder itself can round the volume by, so I don't see the point in even mentioning it.
[a href="index.php?act=findpost&pid=261234"][{POST_SNAPBACK}][/a]

Well guys that's a pity I haven't found the right Precisionst's post but it's very easy...
Just encode a track, make a copy and on the copy apply MP3Gain -1,5 db...
I'm not blind, it's clearly hearable! Judge yourself... Sorry, but that's my observation.
I've read the article about -1,5 db not to be perceptable too, but my ears says something different!
PS: Do not want from me ABX proof because it's so evident!

Edit: Typos & other mistakes...
Sorry for my poor English, I'm trying to get better... ;)
"The greatest trick the Devil ever pulled, was convincing the world he didn't exist."

--scale vs. mp3gain

Reply #10 –
Quote
Quote
Human hearing is on a roughly exponetial scale covering ~12 order of magnitude.  1.5dB is barely perceptable, and as I recall its actually less then the MP3 encoder itself can round the volume by, so I don't see the point in even mentioning it.
[a href="index.php?act=findpost&pid=261234"][{POST_SNAPBACK}][/a]

Well guys that's a pity I haven't found the right Precisionst's post but it's very easy...
Just encode a track, make a copy and on the copy apply MP3Gain -1,5 db...
I'm not blind, it's clearly hearable! Judge yourself... Sorry, but that's my observation.
I've read the article about -1,5 db not to be perceptable too, but my ears says something different!
PS: Do not want from me ABX proof because it's so evident!

Edit: Typos & other mistakes...
[a href="index.php?act=findpost&pid=261251"][{POST_SNAPBACK}][/a]


Did you read my post?  You quoted it, which implies you did, but I have my doubts.  What with how you assume I said the opposite of what I actually said.

Edit:  To be more constructive, let me restate that.  No one has said 1.5dB is inaudiable, it obviously is, so please don't make up and beat down useless strawmen.  Instead stick to the dicussion at hand.

--scale vs. mp3gain

Reply #11 –
Quote
Did you read my post?  You quoted it, which implies you did, but I have my doubts.  What with how you assume I said the opposite of what I actually said.

Edit:  To be more constructive, let me restate that.  No one has said 1.5dB is inaudiable, it obviously is, so please don't make up and beat down useless strawmen.  Instead stick to the dicussion at hand.
[a href="index.php?act=findpost&pid=261286"][{POST_SNAPBACK}][/a]

Don't get me wrong, English is not my native language so as it happened before now now it happened once again... I sorry if I misunderstand you.
Sorry for my poor English, I'm trying to get better... ;)
"The greatest trick the Devil ever pulled, was convincing the world he didn't exist."

--scale vs. mp3gain

Reply #12 –
Quote
I can't see why anyone would use --scale instead of mp3Gain. There are way too many advantages that mp3Gain offers and clearly makes up for that little 1.5 db tradeoff, which I don't consider a problem at all. The only purpose of --scale that I can see is for people that can't use mp3Gain or people that want to apply the same volume to all their files - no matter if it still clipps or not, and are not interested in AlbumGain or getting the max volume without clipping, and will never change the volume again...
[a href="index.php?act=findpost&pid=261197"][{POST_SNAPBACK}][/a]


Because by using --scale i get a 5 to 10% savings in bitrate (depending on how loud the source was) and more precise attenuation (1.5dB really is a pretty grainy stepping).

If I am concerned about clipping, I can run my mp3s all through an mp3gain clipping analysis afterwards... for your information, attenuating roughly 350 albums this way from all different eras and genres, I did not have a SINGLE clipping mp3 (using the default 89db recommendation).

--scale vs. mp3gain

Reply #13 –
Quote
Because by using --scale i get a 5 to 10% savings in bitrate (depending on how loud the source was) and more precise attenuation

ok, you got a point here...I didn't think of that

Quote
(1.5dB really is a pretty grainy stepping).

didn't someone just say, that --scale won't allow a more precise volume change than 1.5 dB either?

Quote
for your information, attenuating roughly 350 albums this way from all different eras and genres, I did not have a SINGLE clipping mp3 (using the default 89db recommendation).

yes, I can confirm that. I use 89dB myself...(mp3Gain though) However, your method doesn't always work. Just think of those Albums that are basically one single Track without any break or anything. If you change the volume for all tracks, you'll most likely notice a huge difference when a new Track starts (even though it is technically the same song)...in the transition to the next track you know...but as I said in my previous post, there is a use for --scale:

Quote
for people that can't use mp3Gain or people that want to apply the same volume to all their files - no matter if it still clipps or not, and are not interested in AlbumGain or getting the max volume without clipping, and will never change the volume again...
--alt-presets are there for a reason! These other switches DO NOT work better than it, trust me on this.
LAME + Joint Stereo doesn't destroy 'Stereo'

--scale vs. mp3gain

Reply #14 –
Quote
Quote
Because by using --scale i get a 5 to 10% savings in bitrate (depending on how loud the source was) and more precise attenuation

ok, you got a point here...I didn't think of that
[a href="index.php?act=findpost&pid=261442"][{POST_SNAPBACK}][/a]

Quote
By using --scale your resulting MP3's should be a bit smaller because you can only lower the --scale switch AFAIK.
[a href="index.php?act=findpost&pid=260722"][{POST_SNAPBACK}][/a]

Quote
Quote
(1.5dB really is a pretty grainy stepping).

didn't someone just say, that --scale won't allow a more precise volume change than 1.5 dB either?
[a href="index.php?act=findpost&pid=261442"][{POST_SNAPBACK}][/a]

Quote
By using MP3Gain your MP3's will actually sound a way quiter than the original if you apply gain of -1,5db or more so I can say that --scale is more precise.
[a href="index.php?act=findpost&pid=260722"][{POST_SNAPBACK}][/a]

   
Sorry for my poor English, I'm trying to get better... ;)
"The greatest trick the Devil ever pulled, was convincing the world he didn't exist."

--scale vs. mp3gain

Reply #15 –
Quote
Quote
Because by using --scale i get a 5 to 10% savings in bitrate (depending on how loud the source was) and more precise attenuation

ok, you got a point here...I didn't think of that

Quote
(1.5dB really is a pretty grainy stepping).

didn't someone just say, that --scale won't allow a more precise volume change than 1.5 dB either?

Quote
for your information, attenuating roughly 350 albums this way from all different eras and genres, I did not have a SINGLE clipping mp3 (using the default 89db recommendation).

yes, I can confirm that. I use 89dB myself...(mp3Gain though) However, your method doesn't always work. Just think of those Albums that are basically one single Track without any break or anything. If you change the volume for all tracks, you'll most likely notice a huge difference when a new Track starts (even though it is technically the same song)...in the transition to the next track you know...but as I said in my previous post, there is a use for --scale:

Quote
for people that can't use mp3Gain or people that want to apply the same volume to all their files - no matter if it still clipps or not, and are not interested in AlbumGain or getting the max volume without clipping, and will never change the volume again...

[a href="index.php?act=findpost&pid=261442"][{POST_SNAPBACK}][/a]


First of all, indeed --scale does give you very fine-grained control over the scalefactor not possible by adjusting the global gain field in 1.5db steps.

Second of all, you argue that mp3gain is better because --scale uses track gain not album gain? WTF? You might want to do some reading before posting misinformation, no offense. I apply the same --scale to every track on an album.

--scale vs. mp3gain

Reply #16 –
Quote
Second of all, you argue that mp3gain is better because --scale uses track gain not album gain? WTF? I apply the same --scale to every track on an album.
[a href="index.php?act=findpost&pid=261511"][{POST_SNAPBACK}][/a]

sorry, but you didn't get what I was saying...I was talking about those Soundtracks for instance that don't have any silence between the tracks...so it is basically one 60 minute track with 12 chapters or so to make it easy to jump to a certain part. Just imagine a very loud track1 which dB level is 99dB followed by a rather quite track2 with a dB level of  85 dB. These values are the *average* dB level of a track. Anyway, when you change the volume of both tracks to 89dB, track1 would be 10dB quieter than before and track2 would be 4dB louder than before...however, since there is no silence between both tracks you would notice a huge volume difference when track2 starts...

All I'm saying is, that on Albums with no silence between the tracks --scale wouldn't be a good idea...unless you don't mind the huge volume difference all of a sudden...
--alt-presets are there for a reason! These other switches DO NOT work better than it, trust me on this.
LAME + Joint Stereo doesn't destroy 'Stereo'

--scale vs. mp3gain

Reply #17 –
Quote
sorry, but you didn't get what I was saying...I was talking about those Soundtracks for instance that don't have any silence between the tracks...so it is basically one 60 minute track with 12 chapters or so to make it easy to jump to a certain part. Just imagine a very loud track1 which dB level is 99dB followed by a rather quite track2 with a dB level of  85 dB. These values are the *average* dB level of a track. Anyway, when you change the volume of both tracks to 89dB, track1 would be 10dB quieter than before and track2 would be 4dB louder than before...however, since there is no silence between both tracks you would notice a huge volume difference when track2 starts...

All I'm saying is, that on Albums with no silence between the tracks --scale wouldn't be a good idea...unless you don't mind the huge volume difference all of a sudden...
[a href="index.php?act=findpost&pid=261515"][{POST_SNAPBACK}][/a]

1) Great to see you ignoring my posts!
2) You still didn't get the purpose of Album Gain which takes care of the volume balance of an album and works right either it is with or without gaps, dynamic changes in volume, whatever...
3) --scale also works the right way because you never alter the volume balance on album if you're using the same scale factor on all of the songs!

I suggest you to read more carefully because some the answes were in the previous posts
Sorry for my poor English, I'm trying to get better... ;)
"The greatest trick the Devil ever pulled, was convincing the world he didn't exist."

--scale vs. mp3gain

Reply #18 –
Quote
1) Great to see you ignoring my posts!

sorry, but this wasn't my intention 

Quote
3) --scale also works the right way because you never alter the volume balance on album if you're using the same scale factor on all of the songs!

ok, wait a minute. Are you actually saying that I could make every file 10% quieter? In that case there wouldn't be any problem. However, how would you change a file to 89dB then? I mean, you don't know the volume of a file...so if I understood you right, I can either use volume relatively to the original volume in % or use a absolute dB level like 89dB? Still, how would you know what % you have to choose to prevent clipping?

I was under the impression that you can only choose a absolute value like 89dB...
--alt-presets are there for a reason! These other switches DO NOT work better than it, trust me on this.
LAME + Joint Stereo doesn't destroy 'Stereo'

--scale vs. mp3gain

Reply #19 –
Quote
Quote
3) --scale also works the right way because you never alter the volume balance on album if you're using the same scale factor on all of the songs!

ok, wait a minute. Are you actually saying that I could make every file 10% quieter?[{POST_SNAPBACK}][/a]

Yes, I am.
Quote
In that case there wouldn't be any problem. However, how would you change a file to 89dB then? I mean, you don't know the volume of a file...
[a href="index.php?act=findpost&pid=261531"][{POST_SNAPBACK}][/a]

By using analyze function of WaveGain and than LAME & --scale switch.
Quote
...so if I understood you right, I can either use volume relatively to the original volume in % or use a absolute dB level like 89dB?
quieter?

I was under the impression that you can only choose a absolute value like 89dB...
[a href="index.php?act=findpost&pid=261531"][{POST_SNAPBACK}][/a]

I don't know... I'm not so edicated and my English is not at the right level...
Quote
Still, how would you know what % you have to choose to prevent clipping?
[a href="index.php?act=findpost&pid=261531"][{POST_SNAPBACK}][/a]

AFAIK, I wouldn't...

You can find an extensive answers to your questions at [a href="http://www.hydrogenaudio.org/forums/index.php?showtopic=10637&view=findpost&p=248330]Wavegain vs. MP3Gain[/url] thread. Highly recommended!
Sorry for my poor English, I'm trying to get better... ;)
"The greatest trick the Devil ever pulled, was convincing the world he didn't exist."

--scale vs. mp3gain

Reply #20 –
@ k.eight.a

all right, thank you
--alt-presets are there for a reason! These other switches DO NOT work better than it, trust me on this.
LAME + Joint Stereo doesn't destroy 'Stereo'

--scale vs. mp3gain

Reply #21 –
About the little debate you guys have been having above...

In simple terms which we can all understand and agree upon, I think:
- MP3Gain is useful for albums of MP3's to be made the same perceived volume.
- --scale gives you more precise control, but unlike MP3Gain, it can not perform normalization but rather it only "scales down" the volume of MP3's by a certain amount.

--scale vs. mp3gain

Reply #22 –
Quote
- --scale gives you more precise control, but unlike MP3Gain, it can not perform normalization but rather it only "scales down" the volume of MP3's by a certain amount.
[a href="index.php?act=findpost&pid=261708"][{POST_SNAPBACK}][/a]


I'm sorry I still don't understand where you're coming from on this point. You can apply a scale of 1.200 for instance if you want to raise the volume... this is "normalization" in exactly the same sense that mp3gain does. Same algorithms and everything if you are running a wavegain analysis to calculate the scalefactor. Just more precision and the adjustment takes place at the time of encoding rather than afterwards.

--scale vs. mp3gain

Reply #23 –
Quote
Quote
- --scale gives you more precise control, but unlike MP3Gain, it can not perform normalization but rather it only "scales down" the volume of MP3's by a certain amount.
[a href="index.php?act=findpost&pid=261708"][{POST_SNAPBACK}][/a]


I'm sorry I still don't understand where you're coming from on this point. You can apply a scale of 1.200 for instance if you want to raise the volume... this is "normalization" in exactly the same sense that mp3gain does. Same algorithms and everything if you are running a wavegain analysis to calculate the scalefactor. Just more precision and the adjustment takes place at the time of encoding rather than afterwards.
[a href="index.php?act=findpost&pid=261751"][{POST_SNAPBACK}][/a]

Yes, exactly!
Can we call the adjustments MP3Gain & --scale switch does normalization? 
Quote
By using --scale your resulting MP3's should be a bit smaller because you can only lower the --scale switch AFAIK.
[a href="index.php?act=findpost&pid=260722"][{POST_SNAPBACK}][/a]

My bad, sorry for possible confusion...
Sorry for my poor English, I'm trying to get better... ;)
"The greatest trick the Devil ever pulled, was convincing the world he didn't exist."

--scale vs. mp3gain

Reply #24 –
Quote
Quote
- --scale gives you more precise control, but unlike MP3Gain, it can not perform normalization but rather it only "scales down" the volume of MP3's by a certain amount.
[a href="index.php?act=findpost&pid=261708"][{POST_SNAPBACK}][/a]


I'm sorry I still don't understand where you're coming from on this point. You can apply a scale of 1.200 for instance if you want to raise the volume... this is "normalization" in exactly the same sense that mp3gain does. Same algorithms and everything if you are running a wavegain analysis to calculate the scalefactor. Just more precision and the adjustment takes place at the time of encoding rather than afterwards.
[a href="index.php?act=findpost&pid=261751"][{POST_SNAPBACK}][/a]
No, it is not normalization — normalization is when you make the gain or the max peak value of a collection of songs the same; meaning you're applying a different gain to each file.

Using the --scale switch does not normalize MP3's, it just makes the MP3 louder or quieter than the source file. And yes I KNOW that it can make them louder. I didn't think it was necessary to mention since with the ridiculous dynamic compression of modern music nobody would ever want to raise the volume.