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: How to exlpain that VBR > 192 CBR (Read 10026 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

How to exlpain that VBR > 192 CBR

Hi guys,

this might be a stupid post but I am trying to explain to a fried that

1) 192 CBR mp3 is not equivalent to a CD quality

2) that a VBR encoded (-V 1 or even -V 2) mp3 has better quality than 192 CBR

How to exlpain that VBR > 192 CBR

Reply #1 –
1.  Well, if your friend perceives 192kbps CBR to be CD quality then there is no convincing.  192kbps is a high enough bitrate that many people may perceive it to be CD quality even though it is CBR.  There is no need to convince if they can't ABX between 192kbps CBR and the source audio CD.

2.  VBR is always better than CBR encoding.  I use this analogy when explaining it to friends: Music is like people, it comes in all shapes and sizes.  You have tall skinny people, short round people, tall big people, normal sized people, and so on.  One song can be comprised of different sized people given that one song will express a certain frequency range and it can go from being complex to being simple.  When encoding to lossy, you are trying to squeeze this music through a door.  With CBR, the door remains the same size no matter what the shape of the person entering it is.  Some people will fit through fine will other people will end up having their left and right sides cut off, their heads cut off, or their shoulders on up cut off.  With VBR, the door adjusts to the sizes of these people.  So the tall skinny people fit just as well as the short round people, tall big people, normal sized people, or small people.  I know it may sound simple but my friends always seem to understand when I use that analogy.

Just keep in mind that Lame VBR gives you a constant quality level rather than a constant bitrate.  The encoder adjusts itself according to the complexity of the music.  That being said, many people may perceive -V 5 (~130kbps VBR) to be CD quality.  So in all practical terms, -V 5 could technically be better than 192kpbs CBR as it yields CD quality results in much smaller file sizes.  What you should do is encode music using various Lame mp3 settings.  I would rip a couple of songs to a lossless format then to -V 5, -V 4, -V 3, and -V 2 Lame mp3.  Sit your friend down and conduct a blind ABX test using foobar2000 and a decent pair of headphones.  That will be the only way to tell what your friend perceives to be CD quality.  Remember that your ears are not my ears, they aren't your friend's ears, and they are different from everyone else.  So what you perceive to be CD quality is different than everyone else.  Many people are fine with 128kbps VBR songs and recent listening tests here on HA show that 128kbps VBR is really not that bad at all.  It is a whole lot better than it was back in 2000 and it exceeds the negativity that many people give it.

How to exlpain that VBR > 192 CBR

Reply #2 –
Nice way to explain it!  Thanks for the suggestion.  Also, is there something written or screen shots of something that visually shows the difference in ... i guess frequencies captured -- the metaphore you use tall ppl short ppl, etc.  I feel like if  i can show him visually maybe like a waveform or graph that could be a good start.  Also, tomorrow I am "battling" his 192 with my vbr encoded mp3 in my car ( I have decent setup with a nice 12" sub woofer.  I am hoping we can hear flatter response or weaker bass from the 192 for a rap/pop song.

How to exlpain that VBR > 192 CBR

Reply #3 –
On classical, trip-hop and some genre CBR will waste lots of space. Also on mono or near mono stuff the bitrate will be nearly twice than what it should be. 3 minute silence will be encoded @ 192k instead of 32k . In the event of a difficult sample , the psymodel might not know what to do and tell VBR to throw 300 k bit or so regardless of the quality setting used. CBR would worsen an already bad situation by fixing the bitrate. The opposite can also happen in a rare occasion - vbr being too aggressive, then cbr / abr has some limited advantage.

VBR is always the winner for efficient encoding. This is something that the CBR hardliners need to understand , Why else are they in the lossy game ?? To save space and maintain quality.

How to exlpain that VBR > 192 CBR

Reply #4 –
There isn't a way to visually show how VBR encoding is more efficient/superior to CBR encoding.  Frequency plots won't show you anything other than that Lame mp3 uses a low-pass frequency cutoff.  You can only explain to him using analogies, you can't visually express it (unless you were to draw pictures of people going through doors).

As for the battle in your car, I don't think it is going to prove much.  What you should do is hook a notebook up to your car (if it has an auxiliary input) then conduct a blind ABX test using software.  Keep in mind that blind ABX tests are really the only way to test the audio quality of files.  Simply switching between tracks won't do it, you need to use software and conduct everything blindly.

How to exlpain that VBR > 192 CBR

Reply #5 –
Short version:

- There are two kinds of "quality" - technical/theoretical, and perceived/practical. The first describes how accurate a signal is stored on a media, before it even reaches your ears. The second describes how indistinguishable to the original it sounds to you, in your mind.

- The term "CD Quality" describes a certain "technical qualitylevel" - it is not about "perceived quality". Thus, even if you perceive something as "indistinguishable", it may still on the media (i.e. harddrive) be stored at below CD-Quality. Why is that possible? Well, because: your hearing-abilities < CD-Quality

- Why does technical quality matter anyways? Well, it matters whenever what you do with audio is NOT listening, but something else. MP3 is good enough for listening, but it does not have reserves if you do significant additional processing on the audio - this applies to reencoding as well as intensive DSPs.

- Not all music is the same. A track has more difficult to compress parts, and less difficult to compress parts. CBR cannot distinguish between the two - it always puts in the same effort, no matter if its difficult to encode parts, or easy to encode parts. VBR on the other hand can put in more effort for difficult parts, and save effort during easy parts. Thus, the soundquality of CBR goes "up and down", while with VBR, it is more constant. (CBR = constant bitrate, variable quality |  VBR = variable bitrate, constant quality).

- Lyx
I am arrogant and I can afford it because I deliver.

How to exlpain that VBR > 192 CBR

Reply #6 –
Ok. I showed him all your exaplanation guys and he is saying what you all are saying is wrong/inaccurate.  He is saying read this:

http://episteme.arstechnica.com/6/ubb.x?q=...mp;m=7310942341

BTW, we listened to the same song encoded in 192 and my VBR, and VBR had audibly more bass -- the 192 sounded more flat.  We even had a third objective person when we conducted the listen.  The third person spotted the VBR as being better (more bass).  My friend is still not convinced about VBR...

How to exlpain that VBR > 192 CBR

Reply #7 –
Oh dear.

EDIT: This is pretty funny: as I look at who's reading this I see a Winamp developer and a Nero MPEG4 developer.  Before too long there will be at least two people who will read this who have been involved in the development of Lame(*).  We may also see the developer of WavPack come along, and flac.  If they see anything wrong with what has been written above that will necessarily mislead people who don't have the ability to understand mp3 on a deeper level, I guarantee they will say something.

That site says nothing that contradicts what has been said above.  You should also be aware that a lot has changed in SEVEN YEARS!!!

(*)EDIT #2: In no less than 5 minutes of me saying this a LAME Developer looked at this thread.

How to exlpain that VBR > 192 CBR

Reply #8 –
So he is telling the people on the forum where the LAME developers "live" to read a 7 year old thread and ignore all of their own collective experience on LAME specific developments in the intervening 7 years?

Right! 

Cheers, Slipstreem. 

How to exlpain that VBR > 192 CBR

Reply #9 –
That thread is 7 years old.  VBR is only as good as the encoder algorithm.  The current version of LAME is exceptionally good and much better than it was in 2001.  Many people's opinion of VBR us tainted by older encoders which implemented it poorly.

How to exlpain that VBR > 192 CBR

Reply #10 –
Quote from: Unknown on
Ok. I showed him all your exaplanation guys and he is saying what you all are saying is wrong/inaccurate.  He is saying read this:

http://episteme.arstechnica.com/6/ubb.x?q=...mp;m=7310942341


I say that it is either time to slap him upside the head, ditch that friend, or just give up.  Some people have their ways set.  Its seems as if your friend has it set in his mind that CBR is better and there is no convincing him even though you showed him a thread (this thread no less) form 2008 versus his 2001 thread.  You would probably have better luck yelling at a brick wall.  I guess there is no convincing then since your friend is set in their ways.

How to exlpain that VBR > 192 CBR

Reply #11 –
Quote from: Unknown on
The third person spotted the VBR as being better (more bass).  My friend is still not convinced about VBR...


I'd be pretty amazed if the VBR had audibly more bass in reality, and that you're very likely to notice no such effect if you conduct a true blind randomized listening test (e.g. using ABX). It's almost certainly a placebo type of effect because you're expecting VBR to sound more bassy.

That said, the arguments in favour of VBR are correct, but it's very unlikely you'll notice the difference between LAME VBR -V3 or better and LAME CBR 192 normally. It's only in problem situations that CBR 192 is likely to fall down badly enough to notice and even then it's more likely you'll hear it on headphones (even cheap ones) than on an expensive speaker set-up.
Dynamic – the artist formerly known as DickD

How to exlpain that VBR > 192 CBR

Reply #12 –
Quote from: Dynamic on
It's only in problem situations that CBR 192 is likely to fall down badly enough to notice and even then it's more likely you'll hear it on headphones (even cheap ones) than on an expensive speaker set-up.
On that note I think it's worth bringing up that since (and including) 3.97, it appears that Lame has suffered some regression in CBR quality...
http://www.hydrogenaudio.org/forums/index....st&p=537861

@Unknown: You should read this and listen to the samples in order to cut your teeth on blind testing.  Artifacts don't reveal themselves as noticeable changes in frequency response (unless you are a bat).

How to exlpain that VBR > 192 CBR

Reply #13 –
Thanks for the help and suggestion, guys.  I agree with everything said in this thread.  I think, at least, I have made a point and a suggestion to my friend.  I think this is enough.  After all, not all people care or are audiophiles like me/us...  Case closed

How to exlpain that VBR > 192 CBR

Reply #14 –
kornchild2002, thank you for that analogy. Most helpful 
 

How to exlpain that VBR > 192 CBR

Reply #15 –
Just tell him that this is not theoretical question, and there are many peoples who already done audio tests and choose vbr over cbr.