IPB

Welcome Guest ( Log In | Register )

 
Reply to this topicStart new topic
Quality of current aoTuv Ogg Vorbis stereo
Neuron
post Jan 12 2013, 22:46
Post #1





Group: Members
Posts: 141
Joined: 14-December 12
Member No.: 105171



I've searched the forum for interesting information about Ogg Vorbis. I use it for Internet radio listening (the station I listen to uses aoTuv at -q2), but I've found many old (2002-2005) threads about the supposed stereo collapse vorbis has even at ~128 kbps? Does Vorbis still suffer from these issues or was this solved?

Thanks in advance for good answers.
Go to the top of the page
+Quote Post
Rescator
post Jan 12 2013, 23:18
Post #2





Group: Members
Posts: 73
Joined: 13-January 09
From: Trondheim
Member No.: 65515



QUOTE (Neuron @ Jan 12 2013, 22:46) *
I've searched the forum for interesting information about Ogg Vorbis. I use it for Internet radio listening (the station I listen to uses aoTuv at -q2), but I've found many old (2002-2005) threads about the supposed stereo collapse vorbis has even at ~128 kbps? Does Vorbis still suffer from these issues or was this solved?

Thanks in advance for good answers.


You might want to hint to the radio station about Ogg Opus.
Vorbis, AAC and MP3 are not that good at lower bitrates. (they where never designed for that)
There is a very cool guy doing lowbitrate tweaks on Vorbis though http://wiki.hydrogenaudio.org/index.php?title=AoTuV
other than that AAC+ was the king of low bitrates.
Ogg Opus is the the new kid on the block and and Chrome, Opera, Firefox, (and IE as well I assume) will support it under the WebRTC standard (Apple's Safari will hopefully tag along as well).

As to the stereo collapse, I've yet to see (or hear) any codec that handles really low bitrates well stereowise (it's in their design to collapse the stereo to save on bits). Although Opus is really impressing me so far.


--------------------
"Normality exist in the minds of others, not mine!" - Rescator
Go to the top of the page
+Quote Post
saratoga
post Jan 12 2013, 23:29
Post #3





Group: Members
Posts: 4715
Joined: 2-September 02
Member No.: 3264



QUOTE (Neuron @ Jan 12 2013, 16:46) *
Does Vorbis still suffer from these issues or was this solved?


At that bitrate I would not expect to have stereo problems. Threads from a decade ago aren't too relevant, Vorbis was quite new then.

QUOTE (Rescator @ Jan 12 2013, 17:18) *
Vorbis, AAC and MP3 are not that good at lower bitrates. (they where never designed for that)


q2 is ~100kbps. Thats not a particularly low bitrate, and in fact Vorbis and AAC are designed more or less exactly for that range. I doubt Opus has much if any advantage here, as all modern codecs should do quite well.

QUOTE (Rescator @ Jan 12 2013, 17:18) *
other than that AAC+ was the king of low bitrates.


At these bitrates, AAC+ is not recommended.
Go to the top of the page
+Quote Post
db1989
post Jan 12 2013, 23:46
Post #4





Group: Super Moderator
Posts: 5141
Joined: 23-June 06
Member No.: 32180



I agree that 128 kbps is not a “low bitrate” by the customary usage of the term, especially in the context of streaming.
Go to the top of the page
+Quote Post
Neuron
post Jan 12 2013, 23:56
Post #5





Group: Members
Posts: 141
Joined: 14-December 12
Member No.: 105171



QUOTE (db1989 @ Jan 12 2013, 23:46) *
I agree that 128 kbps is not a “low bitrate” by the customary usage of the term, especially in the context of streaming.


-q2 is 96-112 kbps generally through, although I saw the bitrate go up to 128 kbps on hard to encode music. And it is aoTuv vorbis, at least Winamp stream information says that.
So does ogg have a good stereo image at 96-112 kbps?
Go to the top of the page
+Quote Post
Neuron
post Jan 13 2013, 00:14
Post #6





Group: Members
Posts: 141
Joined: 14-December 12
Member No.: 105171



Also, at which bitrate would ogg be generally equivalent to -V5 LAME? -V2? 128 kbps CBR LAME? I am asking because I am considering Ogg Vorbis for encoding ripped CD music as well.

This post has been edited by Neuron: Jan 13 2013, 00:15
Go to the top of the page
+Quote Post
saratoga
post Jan 13 2013, 00:45
Post #7





Group: Members
Posts: 4715
Joined: 2-September 02
Member No.: 3264



Vorbis is a little more efficient the MP3, so I would expect it to require a slightly to somewhat lower bitrate as compared to -V5 LAME.
Go to the top of the page
+Quote Post
Neuron
post Jan 13 2013, 00:58
Post #8





Group: Members
Posts: 141
Joined: 14-December 12
Member No.: 105171



Thanks, but what quality setting would that be? Is the Ogg Vorbis point stereo really as lossy as FhG mp3 low bitrate intensity stereo?

This post has been edited by Neuron: Jan 13 2013, 00:59
Go to the top of the page
+Quote Post
saratoga
post Jan 13 2013, 01:30
Post #9





Group: Members
Posts: 4715
Joined: 2-September 02
Member No.: 3264



QUOTE (Neuron @ Jan 12 2013, 18:58) *
Thanks, but what quality setting would that be? Is the Ogg Vorbis point stereo really as lossy as FhG mp3 low bitrate intensity stereo?


Are you asking if Vorbis is comparable to MP3 in terms of stereo compression? If so, I would say that its a little better.
Go to the top of the page
+Quote Post
Neuron
post Jan 13 2013, 01:56
Post #10





Group: Members
Posts: 141
Joined: 14-December 12
Member No.: 105171



QUOTE (saratoga @ Jan 13 2013, 01:30) *
QUOTE (Neuron @ Jan 12 2013, 18:58) *
Thanks, but what quality setting would that be? Is the Ogg Vorbis point stereo really as lossy as FhG mp3 low bitrate intensity stereo?


Are you asking if Vorbis is comparable to MP3 in terms of stereo compression? If so, I would say that its a little better.


How come? Because LAME has lossless m/s joint stereo while Ogg Vorbis has lossy point stereo. Or am I wrong? Or do you mean compared to the old FhG intensity stereo used at 112 kbps and lower?
Go to the top of the page
+Quote Post
saratoga
post Jan 13 2013, 01:59
Post #11





Group: Members
Posts: 4715
Joined: 2-September 02
Member No.: 3264



QUOTE (Neuron @ Jan 12 2013, 19:56) *
Because LAME has lossless m/s joint stereo while Ogg Vorbis has lossy point stereo. Or am I wrong?


You're wrong. Vorbis can do joint stereo as well as more complex methods.
Go to the top of the page
+Quote Post
yourlord
post Jan 13 2013, 02:13
Post #12





Group: Members
Posts: 172
Joined: 1-March 11
Member No.: 88621



Let me just start by stating that Vorbis at q2 is my go to codec and quality for all my portable music encodes.

Now, might I ask about your application? Are you looking for "good enough" quality to load music on a digital audio player or are you looking for a codec and setting to use for your primary library/archive?

I've been using Vorbis at q2 for quite some time now and have had no regrets. I don't archive my music in Vorbis as any lossy codec is a poor choice for that application. My archive is FLAC, which allows me to encode to the most efficient/convenient/supported codec for the target device (largely Vorbis in my case as I tend to make that support a requirement in the products I buy).

Go to the top of the page
+Quote Post
Neuron
post Jan 13 2013, 12:10
Post #13





Group: Members
Posts: 141
Joined: 14-December 12
Member No.: 105171



QUOTE (saratoga @ Jan 13 2013, 01:59) *
QUOTE (Neuron @ Jan 12 2013, 19:56) *
Because LAME has lossless m/s joint stereo while Ogg Vorbis has lossy point stereo. Or am I wrong?


You're wrong. Vorbis can do joint stereo as well as more complex methods.


What are the main stereo methods for Vorbis? Because all those old threads are like "zomg Vorbis uses intensity stereo, stereo image collapse!". I know they are not accurate today, just asking for clarification so I can encode my music with Vorbis without regrets.

This post has been edited by Neuron: Jan 13 2013, 12:12
Go to the top of the page
+Quote Post
Mach-X
post Jan 13 2013, 15:40
Post #14





Group: Members
Posts: 236
Joined: 29-July 12
Member No.: 101859



It doesn't matter what the 'main stereo methods' are. The entire point of lossy encoding is to use psychoaccoustics to 'fool' the brain into thinking it's hearing a full bandwith signal in all its glory. Now if you are asking 'which vorbis quality setting will sound transparent to me?', the only person who can answer that is you with your equipment and your ears. I can offer you what *I* believe is transparent, but somebody will quickly come along with different music/ears/equipment and tell me I'm wrong or that THEY can hear artifacts at that setting. You need to test encode a few tracks at different settings and find what's comfortable for you. Now, all that being said if you still want a 'guideline' on where to start, I used to use vorbis at q5 after hearing it was 'transparent'. But I was curious and started hard testing for myself, and found that as low as vorbis q2 I couldn't reliably blind test against original files. Yay for me much smaller collection. For lame mp3 encoding, I find v4 is my transparency threshold if you truly insist on making apples to oranges comparisons since vorbis and mp3 use different methods to accomplish the same goal.
Go to the top of the page
+Quote Post
db1989
post Jan 13 2013, 16:25
Post #15





Group: Super Moderator
Posts: 5141
Joined: 23-June 06
Member No.: 32180



You have many, many questions. A good proportion of them could, I’m sure, be answered with reference to the official documentation…
[T]he Vorbis spec provides for, and Vorbis release 1.0 rc1 and later implement a coupled channel strategy. Vorbis has two specific mechanisms that may be used alone or in conjunction to implement channel coupling. The first is channel interleaving via residue backend type 2, and the second is square polar mapping. These two general mechanisms are particularly well suited to coupling due to the structure of Vorbis encoding, as we'll explore below, and using both we can implement both totally lossless stereo image coupling [bit-for-bit decode-identical to uncoupled modes], as well as various lossy models that seek to eliminate inaudible or unimportant aspects of the stereo image in order to enhance bitrate. The exact coupling implementation is generalized to allow the encoder a great deal of flexibility in implementation of a stereo or surround model without requiring any significant complexity increase over the combinatorially simpler mid/side joint stereo of mp3 and other current audio codecs.
…and to the changelog for aoTuV with reference to stereo-related changes, the threads discussing aoTuV, and so on.

That will take care of most of the theory. In practice, Mach-X is, of course, correct in saying that no one else can tell you what will sound good to your ears.
Go to the top of the page
+Quote Post
greynol
post Jan 13 2013, 19:01
Post #16





Group: Super Moderator
Posts: 10000
Joined: 1-April 04
From: San Francisco
Member No.: 13167



One needs to know how to conduct proper DBT tests and understand that something like 8 out of 13 does not exactly constitute a compelling score. wink.gif


--------------------
Your eyes cannot hear.
Go to the top of the page
+Quote Post
Mach-X
post Jan 14 2013, 05:11
Post #17





Group: Members
Posts: 236
Joined: 29-July 12
Member No.: 101859



If I could honestly differentiate between vorbis q2 and flac 8 out of 13 times I'd be waaaaaaayyyyyyyyy down the rabbit hole shouting about how golden cable tips 'vastly improve spatiality and high frequency response'. That being said, I don't bother with dbt when deciding on a codec setting to use. I just encode at increasingly higher levels until the obvious smearing/artifacts are gone and be done with it. Why bother abxing with fighter beat or castanets? Neither of those are in my music library. And abxing always ends up with me chasing my tail with not so wonderfully produced (but still great music!) albums where I discover what I *think* are artifacts which actually turn out to be part of the recording! Not that abxing isn't important, it surely is when it comes to dispelling audio myths, but it can also give new users false ideas when they get 8 out of 13 guesses 'correct'. For now I listen with my sansa clip, my sennheiser iems, and my sanity intact!
Go to the top of the page
+Quote Post
Neuron
post Jan 14 2013, 16:19
Post #18





Group: Members
Posts: 141
Joined: 14-December 12
Member No.: 105171



QUOTE (Mach-X @ Jan 14 2013, 05:11) *
If I could honestly differentiate between vorbis q2 and flac 8 out of 13 times I'd be waaaaaaayyyyyyyyy down the rabbit hole shouting about how golden cable tips 'vastly improve spatiality and high frequency response'. That being said, I don't bother with dbt when deciding on a codec setting to use. I just encode at increasingly higher levels until the obvious smearing/artifacts are gone and be done with it. Why bother abxing with fighter beat or castanets? Neither of those are in my music library. And abxing always ends up with me chasing my tail with not so wonderfully produced (but still great music!) albums where I discover what I *think* are artifacts which actually turn out to be part of the recording! Not that abxing isn't important, it surely is when it comes to dispelling audio myths, but it can also give new users false ideas when they get 8 out of 13 guesses 'correct'. For now I listen with my sansa clip, my sennheiser iems, and my sanity intact!


ABX testing lossy vs lossless is not in the same league as nonsense like gold cables or green markers on CDs.

Thanks for all the input through. So the conclusion is that current Vorbis does not destroy stereo and that I should determine my required quality by ABX testing. I will do that when I'll have time.
Go to the top of the page
+Quote Post
greynol
post Jan 14 2013, 16:41
Post #19





Group: Super Moderator
Posts: 10000
Joined: 1-April 04
From: San Francisco
Member No.: 13167



Lossy artifacts or interconnects, the problem is still the same: a meaningful discussion cannot be progressed without there being a basis of facts upon which all parties can agree.

It is usually best to research before posting, especially when there is the risk of passing off incorrect notions as fact.


--------------------
Your eyes cannot hear.
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: 17th April 2014 - 23:46