IPB

Welcome Guest ( Log In | Register )

 
Reply to this topicStart new topic
Your opinion on MP3 CBR 320kbps, [TOS #6: moved from MP3 - Tech]
tiago_fernandes
post Mar 13 2013, 19:00
Post #1





Group: Members
Posts: 4
Joined: 13-March 13
Member No.: 107203



Hello everyone! I'm new here so go easy on me, I don't know much about the subject.

So, I'm a music lover who's trying to select his all-time favorite songs and I've come to the conclusion that I should use .mp3 files because this is the most common format for music files.

The problem is, there are lots of different bit-rates, etc, and that really confuses me. I would like to know if CBR 320kbps is the best possible quality for an .mp3 file and why/if 44.1Khz/Joint Stereo should be chosen.

Please disregard everything else like file size, etc.

Thank you in advance, and pay no mind to grammar errors.

Regards from Portugal.
Go to the top of the page
+Quote Post
db1989
post Mar 13 2013, 19:08
Post #2





Group: Super Moderator
Posts: 5275
Joined: 23-June 06
Member No.: 32180



QUOTE (tiago_fernandes @ Mar 13 2013, 18:00) *
I'm new here so go easy on me, I don't know much about the subject.
Neither of these statements renders you unable to search for answers to these questions. The exact same ones have been asked here tens if not hundreds of times.

I could have just left it at that, but some (masochistic? procrastinating?) part of me pushed me begrudgingly to offer brief answers. I have no problem if other users prefer to avoid repeating themselves by allocating you some responsibility to search for yourself; plenty of reading material is available here and elsewhere.

QUOTE
I would like to know if CBR 320kbps is the best possible quality for an .mp3 file
Not necessarily. Firstly, quality is relevant only in terms of whether or not a setting is generally transparent to you; above that threshold, no setting can be said to be of any higher quality than any other. Secondly, VBR is a much more sane way to distribute bitrate and may – theoretically, again not relevant if transparency is attained – enable more efficient allocation of bits on an instantaneous basis than would forcing the encoder to use the maximal number all the time.

QUOTE
if 44.1Khz/Joint Stereo should be chosen.
44.1 kHz is the sampling rate of CDs, if that’s your implication here. And yes, joint stereo is superior. Again, please search for the even more numerous questions about the latter rather than expecting us to repeat the answers yet again.

And this does not belong in MP3 - Tech. As that subforum’s subtitle states, it’s for discussions of technical details of the underlying format: asking which settings to use does not qualify as technical by a long shot.
Go to the top of the page
+Quote Post
tiago_fernandes
post Mar 13 2013, 19:14
Post #3





Group: Members
Posts: 4
Joined: 13-March 13
Member No.: 107203



db1989, I am sorry for posting this in the wrong section, I thought it was the right one at the time.

I have been searching the forum but everyone participating in this kind of discussion (is 320kbps good, etc) seems to post these very elaborate answers and it makes it hard for a foreigner to easily understand things.

Regardless, I apologize.
Go to the top of the page
+Quote Post
db1989
post Mar 13 2013, 19:27
Post #4





Group: Super Moderator
Posts: 5275
Joined: 23-June 06
Member No.: 32180



OK, no big problem. If the technical language makes it hard to understand, that’s something that should be worthwhile for us to try to help. Which things did you find hard to understand?

For instance, transparency just means that a signal processed by a lossy encoder still sounds the same as the original, lossless signal – to an individual listener. Joint stereo is a way of representing two channels as (mid+side) rather than (left [and separate] right), which often makes sections of audio cheaper to compress and thus consume less space; intelligent encoders such as LAME (which I guess you know is a famous encoder of MP3 around here) can vary between M/S and L/R modes on a frame-by-frame basis so that every frame (a small section of the input audio) is encoded in the way that is most efficient.

VBR may be more efficient (A) because it does not use the full bitrate all the time, including when the encoder deems that such a high bandwidth is not necessary, and (B) because it can use the bit-reservoir, a way to save unused bits for subsequent frames, and thus may be able to allocate especially complex frames more than 320 kbps.

There are other things that may be relevant to you, but you’d need to ask about them. Sorry if I seemed harsh: I didn’t even consider that you weren’t a native speaker of English, which I guess is a compliment in a way. smile.gif I was probably excessively quick because other users haven’t searched in the past, but obviously you shouldn’t be blamed for that. If you’ve had a look around but are still unsure about some things, let us know what we can try to clarify.
Go to the top of the page
+Quote Post
julf
post Mar 13 2013, 19:28
Post #5





Group: Members
Posts: 361
Joined: 14-January 12
Member No.: 96426



QUOTE (tiago_fernandes @ Mar 13 2013, 19:14) *
I have been searching the forum but everyone participating in this kind of discussion (is 320kbps good, etc) seems to post these very elaborate answers and it makes it hard for a foreigner to easily understand things.


320 kbps is 'good', for whatever reasonable definition, except it takes up extra space if a lower bitrate would work for you. So what is it that you want to do?
Go to the top of the page
+Quote Post
tiago_fernandes
post Mar 13 2013, 19:42
Post #6





Group: Members
Posts: 4
Joined: 13-March 13
Member No.: 107203



Thanks for your replies, db1989 and julf.

I'm considering buying an iPod Shuffle to store my favorite songs. It's lightweight and small, which makes it good to carry around and use whenever I'm on the bus or train. I understand that by using the default Apple Earphones the sound quality won't be that great in noisy environments, but since my collection will be pretty limited (therefore, file size isn't that relevant) I am trying to understand how I can obtain the best possible quality in the .mp3 files themselves.

I can't understand how some users say that VBRs are better than CBR 320kbps. How can it be better if the only way of going as high as 320kbps is using CBR? And if your file constantly plays at the highest possible bit-rate, it makes no sense (to me) when people claim to prefer VBRs quality-wise.

Also, according to this article (http://www.richardfarrar.com/what-is-joint-stereo/), when you reach bit-rates higher than 256kbps, Full Stereo is better than Joint Stereo. This is another thing that confuses me, because every .mp3 file that I have is a Joint Stereo file even though they are CBR 320kbps.

Please note that the files I have at the moment were not self-encoded.

Thanks for trying to help me out, guys.

This post has been edited by tiago_fernandes: Mar 13 2013, 19:48
Go to the top of the page
+Quote Post
DVDdoug
post Mar 13 2013, 19:48
Post #7





Group: Members
Posts: 2712
Joined: 24-August 07
From: Silicon Valley
Member No.: 46454



You CAN say that 320CBR is the least-lossy or least-compressed MP3 option (and you get the biggest files).

ALL MP3 compression is lossy, so it gets rather tricky when you say "best" or "better" unless you can actually hear a difference. If the 320 and a lower bitrate sound identical, or the lower bitrate and 320 both sound identical to the uncompressed original, you cannot say 320 is better. No matter what bitrate you use, none of the bytes in the de-compressed audio data are identical to the original uncompressed file.

If you are not concerned with disc space, there's no harm in using 320. I use V0 (the least-lossy variable bitrate option) for that reason. I wanted good-quality MP3 and I didn't have any disc-space concerns.

QUOTE
I've come to the conclusion that I should use .mp3 files because this is the most common format for music files.
You might also consider AAC. It's almost as universal as MP3 and I don't know of any portable players that cannot play AAC ...I'm thinking I may have made the "wrong" choice myself. There are issues with silence at the beginning (or end?) of MP3s that make it tricky to get gapless playback. As far as I know this is not an issue with other formats. If you want gapless playback, it's probably best just to avoid MP3. Gapless playback not a big concern to everyone and there are several work-arounds, but there is not one-solution that works on every player.

This post has been edited by DVDdoug: Mar 13 2013, 19:50
Go to the top of the page
+Quote Post
pdq
post Mar 13 2013, 19:48
Post #8





Group: Members
Posts: 3450
Joined: 1-September 05
From: SE Pennsylvania
Member No.: 24233



You may ignore that article, because it is wrong, at least for a good mp3 encoder like lame. Joint stereo simply means that the encoder is able to select the best means of encoding the material. So-called "full stereo" takes away the encoder's ability to make that choice. At the highest bitrates it probably doesn't matter that much, but for those rare cases where there might be an audible difference, joint stereo should be the better choice.
Go to the top of the page
+Quote Post
db1989
post Mar 13 2013, 19:56
Post #9





Group: Super Moderator
Posts: 5275
Joined: 23-June 06
Member No.: 32180



QUOTE (tiago_fernandes @ Mar 13 2013, 18:42) *
I am trying to understand how I can obtain the best possible quality in the .mp3 files themselves.
Again, “best possible quality” = transparency, which is by definition the same quality as the input file, so once your personal threshold of transparency has been passed, talk of relative quality is meaningless. Anything beyond that merely increases the probability that the encoder will be able to represent the input signal in a way that is transparent to you, which may already have passed 100% at a lower setting.

How to determine what’s transparent to you? Test using ABX double-blind testing. Or trust people here’s recommendations. Most will probably recommend VBR -V2 or maybe -V0. Which brings us to…

QUOTE
I can't understand how some users say that VBRs are better than CBR 320kbps. How can it be better if the only way of going as high as 320kbps is using CBR? And if your file constantly plays at the highest possible bit-rate, it makes no sense (to me) when people claim to prefer VBRs quality-wise.
Please note that a lack of ability to understand is in no way evidence against what we have told you and will probably continue to tell you. 320 kbps is overkill for many input signals and represents a waste of space in many cases. VBR enables the encoder to use only the bits it needs once it has judged the complexity of the input signal and allocate bits accordingly. Are either of these concepts hard to understand, really? Not to mention that 320 kbps CBR may not always provide the highest probability of transparency (because, again, quality is not a very useful term here) for reasons such as the one that I already introduced.

QUOTE
Also, according to this article (http://www.richardfarrar.com/what-is-joint-stereo/), when you reach bit-rates higher than 256kbps, Full Stereo is better than Joint Stereo. This is another thing that confuses me, because every .mp3 file that I have is a Joint Stereo file even though they are CBR 320kbps.
According to someone somewhere, something. So what? You asked us whether joint-stereo is advisable, and our answer is an overwhelming yes. Perhaps compare the emphasis that Hydrogenaudio places on objectively derived advice, as opposed to subjective and unsubstantiated fluff, and ask yourself whether that website follows similarly high standards.

Enabling joint-stereo simply gives the encoder the ability to choose intelligently between mid-side and left-right (‘simple’) stereo on a per-frame basis. Given that mid-side coding is a simple but highly effective, and procedurally lossless, way to transform a pair of inputs in a way that may enable them to be stored using fewer bits, how could the ability to use it in relevant cases be a bad thing?

QUOTE
Please note that the files I have at the moment were not self-encoded.
Is this relevant? Are they lossless? If not, I presume/hope you don’t intend to transcode them to 320 kbps.

Finally, not that there is any reason not to trust an encoder and setting that you have tested or for which you are willing to accept advice, but if you want the best possible chance of transparency (a.k.a. ‘quality’), there is no choice but to use a lossless format. Otherwise, use a lossy encoder; choose a setting that represents a sane balance between quality and bitrate, rather than piling on the full 320 kbps for no real reason; and make efficient use of your disk-space.

One last time for emphasis: If there is a setting (i.e. VBR level, mean bitrate, etc.) at which an encoder is likely to produce (generally) transparent results for you, there is no need to go higher. The only possible excuse is a concern that something eventually might exhibit an audible artefact, and if this is a concern that you hold, you have two options: do some testing to reassure yourself that a lossy encoder and setting [whatever] are fine, or go fully paranoid and use a lossless format ‘just in case’. Landing somewhere in-between is suboptimal and likely only to consume storage capacity unnecessarily.

This post has been edited by db1989: Mar 13 2013, 20:11
Go to the top of the page
+Quote Post
tiago_fernandes
post Mar 13 2013, 20:10
Post #10





Group: Members
Posts: 4
Joined: 13-March 13
Member No.: 107203



Thanks everyone! Everything is clear now.

I gotta give it to you all, now that I think of it, CBR 320kbps may be a bit overkill. If I want great quality, might as well go Lossless, instead of high quality lossy.

You guys have been great!
Go to the top of the page
+Quote Post
db1989
post Mar 13 2013, 20:13
Post #11





Group: Super Moderator
Posts: 5275
Joined: 23-June 06
Member No.: 32180



Glad you found the replies helpful.

QUOTE (tiago_fernandes @ Mar 13 2013, 19:10) *
I gotta give it to you all, now that I think of it, CBR 320kbps may be a bit overkill. If I want great quality, might as well go Lossless, instead of high quality lossy.
Sure, but also, don’t presume that mid-quality lossy wouldn’t sound exactly the same to your ears! Many people are happy to pick a generally ‘safe’ setting such as -V2 and not worry about it, but you can always run some tests to check for yourself.



QUOTE (DVDdoug @ Mar 13 2013, 18:48) *
There are issues with silence at the beginning (or end?) of MP3s that make it tricky to get gapless playback. As far as I know this is not an issue with other formats. If you want gapless playback, it's probably best just to avoid MP3. Gapless playback not a big concern to everyone and there are several work-arounds, but there is not one-solution that works on every player.
The OP mentioned the possibility of buying an iPod. As far as I know, gapless playback of MP3s is supported just fine in products by Apple, encoded either by iTunes itself (=Fraunhofer) or by LAME. Certainly, FWIW, I don’t recall noticing any audible gaps in the few specifically gapless transitions between songs that I’ve listened to on my (somewhat involuntary) iPod Nano, and that’s in stark contrast to the Creative Zen X-Fi Style that I had previously (before it got too wet to go on…)

This post has been edited by db1989: Mar 13 2013, 20:18
Go to the top of the page
+Quote Post
benski
post Mar 13 2013, 21:23
Post #12


Winamp Developer


Group: Developer
Posts: 670
Joined: 17-July 05
From: Brooklyn, NY
Member No.: 23375



With LAME, VBR -v 0 is better than CBR 320kbps simply because the implementation of the encoder differs slightly between CBR and VBR.

This is only a theoretical difference between the two encoder settings. In practice, it is impossible to test the difference between VBR -v 0 and CBR because the quality is already past the limits of human hearing.
Go to the top of the page
+Quote Post
krabapple
post Mar 14 2013, 19:59
Post #13





Group: Members
Posts: 2518
Joined: 18-December 03
Member No.: 10538



QUOTE (benski @ Mar 13 2013, 16:23) *
With LAME, VBR -v 0 is better than CBR 320kbps simply because the implementation of the encoder differs slightly between CBR and VBR.

This is only a theoretical difference between the two encoder settings. In practice, it is impossible to test the difference between VBR -v 0 and CBR because the quality is already past the limits of human hearing.


quality difference , you mean?


Because even as good as 320 kbps is, there are a few people on HA who have shown (to HA standards) that they can tell a 320kpbs mp3 from source, at least sometimes.

This post has been edited by krabapple: Mar 14 2013, 20:00
Go to the top of the page
+Quote Post
db1989
post Apr 12 2013, 10:43
Post #14





Group: Super Moderator
Posts: 5275
Joined: 23-June 06
Member No.: 32180



Threadjack split: Is joint-stereo definitely better in iTunes, LAME? Should I re-encode?
Go to the top of the page
+Quote Post

Reply to this topicStart new topic
1 User(s) are reading this topic (1 Guests and 0 Anonymous Users)
0 Members:

 



RSS Lo-Fi Version Time is now: 21st December 2014 - 16:15