Quantcast

The MP3 gain scandal!

elementalist2

Standard Member
Bah! I made a big long rant and IE lost it :mad: let's rant again...

This may not be news to a lot of informed folks here, but I only recently discovered MP3Gain and am still in shock.

I've used various MP3/WMA software and hardware products over the years but had never worked out an issue I had noticed more and more until recently.

I think we've all noticed tracks and sometimes whole albums that had nasty distortion at high volumes. Noticing it a LOT in the last few months (having got some real good headphones and a nice player) I made a concerted effort to research the issue.

So, to cut a long story medium-sized, I discovered the whole gain issue and the MP3Gain software and spent days and days of effort converting all my WMAs to MP3 and then adjusting them all for gain problems.

So here I am with tracks I thought were just badly produced or encoded now sounding lovely and am I happy? Well, yes, but I'm also MAD! I might never have known about it and I'm sure that the majority of people won't ever, but will just keep on putting up with the bass on their new Hed Kandi iTunes download sounding like their speakers are damaged.

What I want to know is, how are 'they' getting away with it?! How are iTunes, etc. selling sub-standard files and how are MS etc. selling sub-standard rippers without someone in the know saying, "hey, hold on! If you don't adjust the gain on some of these, they sound crap!"

Tell me if I'm misunderstanding something, but it seems like something that should be an industry wide shame! Sure, I'll bet a lot of people don't notice, but that's no excuse!

I just don't understand why noone has sorted this out already? Is there a conspiracy? Is EMI, Sony, etc. paying to gag folks so they can perpetuate the CD-is-better-than-MP3 to eak out the last of the market while they belatedly gear up to bleed the MP3 market?

I haven't been so angry since 'they' started intentionally mangling CD files to stop 'mobster pirates' from using PCs to dupe music (i.e. irritate legit customers who rip to MP3)

Gah! Let's storm the EMI offices! Who's with me?!?! :mad: ;)
 
L

lukey

Guest
I have to agree with your comment on substandard downloads, ive only used itunes a couple of times and cant believe the sound difference to a high quality mp3 file of actual cd. Naturally the downloads are at a low bitrate so you cant expect super quality. personnally i just like to buy cd's and rip them in either 320kbs vbr or in lossless which gives as close to the original as can be for a compressed file. As for Mp3gain, i have found just as good as a result by using the sound check function in itunes :confused:
 

elementalist2

Standard Member
Ah! Haven't read about this 'soundcheck' function. I'll ask an iPodder and see what they say - it's not gonna calm me down totally though! hehe

Also, the gain problem is totally independent of encoding sample rate - I understand it's a setting on top of the sampling that sets the relative output volume of the track. It's just crazy that ripping software doesn't analyse and alter this as standard... gah - getting angry again!
 
L

lukey

Guest
the sound check function in itunes and on your ipod simply levels the volume of all mp3 or other music files on your ipod so they are all the same volume so you dont get the jump in volume between different types of music, in essence it reduces the gain and thus distortion....well thats what ive found with my experience anyway. Still doesnt make the itunes downloads sound good though!!
 

Cloysterpeteuk

Well-known Member
What volume do people use for mp3gain, I whack it upto 100 but I don't really notice any distortion.
 

elementalist2

Standard Member
Cloysterpeteuk said:
What volume do people use for mp3gain, I whack it upto 100 but I don't really notice any distortion.
Eek! That's making files potentially loads worse! I'm real surprised you haven't noticed lots of distortion - maybe your player software compensates (Im aware that some can and does).

Generally, I use the 'maximise' function - it analyses the whole album checking each individual file for what the maximum gain setting can be without clipping (the clipping of waveforms due to excessive gain setting is what causes the distortion) and then can either adjust the individual tracks or can check what safe gain setting for the whole album can be used - I use the whole album maximise so that relative track volumes are maintained.

Try picking a track you know has 'big bass' beats like a modern dance/rap track or a heavy metal track - listen for the bass beats going 'crunch-crunch' instead of 'thud-thud'. Now reset the gain using maximise and the 'crunch' goes away leaving a lovely meaty thump without distortion.

P.S. the maximise function is 'hidden' away (not sure why - it's the best feature) look for the checkbox under Options-Advanced.
 

Cloysterpeteuk

Well-known Member
I'm gonna go and test two files, one untempered with at whatever volume the default download volume is and comparing it to a 100db version of the same track, i'm thinking ACDC Highway to Hell?. I don't think the ipod defaults earphones are sensitive to notice, maybe I sh ould try them with my more expensive Sony's?.
 

Cloysterpeteuk

Well-known Member
I just undid the gain chages I made to the 100db version of Highway to Hell and it was still at 94db with the clipping Y so maybe I wouldn't notice any difference

Do you think I should remove all gain changes are give the itunes sound check a go?. I will go and have a look for a maximise indivual tracks, but serely that would end up giving you differing volume levels again?.
 

elementalist2

Standard Member
I guess if you've gone from clipped to more-clipped it would sound pretty much the same.

Try the "Apply Max No-Clip Gain for Album" option. It will drop the whole album volume to a level where all the files are not clipping.

I did this across my whole collection (took hours and hours) and haven't found any problems and have re-tried many that used to sound cruddy and now are lovely.
 

elementalist2

Standard Member
Cloysterpeteuk said:
I'm gonna go and test two files, one untempered with at whatever volume the default download volume is and comparing it to a 100db version of the same track, i'm thinking ACDC Highway to Hell?. I don't think the ipod defaults earphones are sensitive to notice, maybe I sh ould try them with my more expensive Sony's?.
I've heard the iPod ones are poopy, yes. The Sony in-ear buds tend to be much better. I recommend the ones with the rubber sound-excluding fittings (Sony MDR-EX71SLB Fontopia - £25ish). I recommend even more Shure E4C earphones with triple flange fittings, but they are, like, £150!

Better earphones will definitely make the problem show up. The reason I recently researched this whole issue was because I bought some Shure earphones and my music sounded worse! It did until I gain-adjusted it, anyway.
 

Cloysterpeteuk

Well-known Member
elementalist2 said:
I guess if you've gone from clipped to more-clipped it would sound pretty much the same.

Try the "Apply Max No-Clip Gain for Album" option. It will drop the whole album volume to a level where all the files are not clipping.

I did this across my whole collection (took hours and hours) and haven't found any problems and have re-tried many that used to sound cruddy and now are lovely.
Can't really do album gain as I don't really have any albums, I used the individual maximum vol without clipping but there was a load that it didn't do, the vast majority of those were AAC files which it ignored even though they usually fine. Now the volumes range from 89 - 94 can you notice that volume differance with your ears?.

I'll slowly get all the collection in better quality because everynow and then i'll go and download a higher bitrate version of some of my songs as a lot are 128.
 
S

shadowritten

Guest
Am I missing something here? Are we suggesting that MP3s (or other formats) are being deliberately sold without first having their gain 'corrected' to remove distortion? If so, then I'm not surprised.

The gain of any particular track or album is 'fixed', if you will, according to the way the engineers set it for any given CD. So, when an online retailer rips/encodes CDs in readiness for resale, they're simply doing just that: ripping and encoding. Imagine how much extra time - and, therefore, how much more they'd charge you for this - if they then had to use a program like MP3Gain to modify gain by track or album.

I've used MP3Gain with great success, as echoed in the posts here. But I'd rather I got to choose how loud or quiet (relatively) my MP3s were than have some retailer do it for me, then have to redo it all myself!

Word of advice, incidentally. I ran MP3Gain for nearly 48 hours constantly last year, and completely knackered the read/write arm of my HDD! But then, I was analysing and correcting gain on several thousand MP3s. Wouldn't recommend being this stupid!
 

Cloysterpeteuk

Well-known Member
What vol level do you mp3gain at shadow?, i'm started to think my ears are too crap to notice distortions and other audio faults.
 
S

shadowritten

Guest
Cloysterpeteuk said:
What vol level do you mp3gain at shadow?, i'm started to think my ears are too crap to notice distortions and other audio faults.
Like elementalist2, I used to use the Max no-clip gain. But I did it by track, not album. Which makes sense when you think about it. Okay, so it's nice to maintain gain within an album ... but what happens when you go shuffle? You've still got those big jumps in gain!

However, even this method wasn't ideal. So I analysed everything, kept adjusting the normal target till I noticed the fewest number of tracks clipping, then went for a uniform track-by-track gain correction. Think I ended up with all tracks coming in at between 89.9dB-92.7dB. Any louder, and a much greater number of my tracks would've distorted horribly.
 

Uridium

Well-known Member
took at look at the mp3gain site earlier and noticed it mentioned that some players had a tendency to display garbled Id3 tags after gain had been changed.

After many months of pain all my CD's are ripped, tagged and filed to perfection and would be near suicidal if I had to go through that again.

so I just wondered if anyone with a HD3 and a Netgear MP101 had used MP3gain and can vouch for it leaving the tags displaying ok on these two players?
 

elementalist2

Standard Member
uridium said:
took at look at the mp3gain site earlier and noticed it mentioned that some players had a tendency to display garbled Id3 tags after gain had been changed.

After many months of pain all my CD's are ripped, tagged and filed to perfection and would be near suicidal if I had to go through that again.

so I just wondered if anyone with a HD3 and a Netgear MP101 had used MP3gain and can vouch for it leaving the tags displaying ok on these two players?
Back-up is your friend. I would advise backing up some or all and trynig them out.
 
S

shadowritten

Guest
uridium said:
took at look at the mp3gain site earlier and noticed it mentioned that some players had a tendency to display garbled Id3 tags after gain had been changed.

After many months of pain all my CD's are ripped, tagged and filed to perfection and would be near suicidal if I had to go through that again.

so I just wondered if anyone with a HD3 and a Netgear MP101 had used MP3gain and can vouch for it leaving the tags displaying ok on these two players?
Hi mate

No worries with tags on the 3000+ MP3s I did - everything A okay (though can't speak for HD3/Netgear situation, just generally).

(Odd, but I'm sure I posted this answer last night, only it disappeared off the forum?)
 

elementalist2

Standard Member
shadowritten said:
Am I missing something here? Are we suggesting that MP3s (or other formats) are being deliberately sold without first having their gain 'corrected' to remove distortion? If so, then I'm not surprised.
Almost every modern CD I rip these days and every album I download has clipping issues in MP3Gain. Some of them are horribly distorted (to my ears) and some only just noticable.

I think it's the same syndrome as with TV adverts, without trying to look like too much of an old fart, over the last few years I've noticed them, relative to TV programs, get louder and louder in an effort to compete for our attention. CD producers are doing the same thing - making their tracks louder on the radio by upping the gain, which CD players cope with better than the more precise computer-based MP3 player model.

I believe the MP3Gain help files suggest 89 as a good volume setting for everything. I would agree that I've never seen a track where 89 would have caused clipping and it's more than loud enough to not cause it to be 'quiet'.

The only reason I chose to use Max No-Clip Gain instead is because you can't be 100% certain 89 is fine for all, whereas Max No-Clip will always stop clipping.

A setting of 89 or Max No-Clip, whichever is the lowest, would have been ideal :)
 
S

shadowritten

Guest
elementalist2 said:
I believe the MP3Gain help files suggest 89 as a good volume setting for everything. I would agree that I've never seen a track where 89 would have caused clipping and it's more than loud enough to not cause it to be 'quiet'.
Of course, things can go the other way and become too quiet.

Can I assume you're not a classical fan? Ripping some 400 classical CDs and passing these through MP3Gain resulted in roughly 25% of the tracks coming in at between 77dB and 85dB! Which isn't a problem if the amp in your DAP is man enough to push these tracks out at a decent rate without you cranking your volume to the max and thus consigning your battery to a much shorter overall life.

Oh, and I agree about engineers upping gain all over the place. This was my point, really. A professional musician friend says it gets on his wick! All I'm saying is that it's not a scandal in terms of the MP3s or whatever being sold by online retailers. They can only rip/encode what they're given; and if they themselves aren't even doing this process, then they've still less control. It rests with the engineers. And, as I say, to expect retailers to correct gain before resale is also to expect an according price hike per track or album.
 

GETanner

Active Member
I mp3Gained (AlbumGain at 89db) over 500 ripped CDs on a PIII 600 machine over the Summer. Took days but the results were nothing but perfect.

No longer do I have to constantly fiddle in my pocket to adjust the volume of my player.

Alas, I no longer have the excuse of having to constantly change the volume when I want to fiddle in my pocket.

G
 

elementalist2

Standard Member
shadowritten said:
Of course, things can go the other way and become too quiet.

Can I assume you're not a classical fan? Ripping some 400 classical CDs and passing these through MP3Gain resulted in roughly 25% of the tracks coming in at between 77dB and 85dB! Which isn't a problem if the amp in your DAP is man enough to push these tracks out at a decent rate without you cranking your volume to the max and thus consigning your battery to a much shorter overall life.

Oh, and I agree about engineers upping gain all over the place. This was my point, really. A professional musician friend says it gets on his wick! All I'm saying is that it's not a scandal in terms of the MP3s or whatever being sold by online retailers. They can only rip/encode what they're given; and if they themselves aren't even doing this process, then they've still less control. It rests with the engineers. And, as I say, to expect retailers to correct gain before resale is also to expect an according price hike per track or album.
Not a big classical fan :) But what I have got, yes, tend not to be 'ramped up' with gain. I can well imagine sound engineers being annoyed to hell! It must be like a chef having all his dishes covered in MSG by some restaurant marketing guy before they get to the customer!

I still say there is a big responsibility on the part of rippers. The techinical mysteries of how a ripper works is beyond 90% of people using them, I'm sure. They just check an option that says "CD Quality" and click "Rip". To then take something that plays fine on a CD player (aparently, they are 'used to' handling high gain of CD tracks) and produce what probably sounds awful on an MP3 player without any attempt to 'fix' it when it's a very easy process is wrong, IMHO.

I can guarantee that every company out there that have produced a ripper know about the gain problem. To not add some kind "Gain Check" function to their rippers is crazy/negligent.

It's like knowing that a little salt will make a meal taste better, but not bothering because your customers don't know that. They don't even put salt on the table! ;)
 
S

shadowritten

Guest
elementalist2 said:
I can guarantee that every company out there that have produced a ripper know about the gain problem. To not add some kind "Gain Check" function to their rippers is crazy/negligent.
I'm inclined to agree here. Perhaps it should fall to the makers of rippers/encoders to include something a little more 'precise', shall will say, than a bog-standard normalisation function (normalisation being something of an oddity from ripper to ripper IMO, and not always producing the desired result).
 

Trending threads

Latest News

Panasonic TVs get BritBox streaming app
  • By Andy Bassett
  • Published
Clearaudio Concept turntable re-launched as Active edition
  • By Andy Bassett
  • Published
Teufel announces Cinebar Ultima and Cinedeck TV sound solutions
  • By Andy Bassett
  • Published
Sky adds 300 hours of Channel 5 box-sets
  • By Andy Bassett
  • Published
Samsung Smart TVs from 2012-2015 require update for BBC iPlayer
  • By Andy Bassett
  • Published
Top Bottom