IPB

Welcome Guest ( Log In | Register )

How does Opus compare to AAC and other codecs in general transparency?, [TOS #5: split from “Listening test using 2013-03-09 build”/TID:99841]
zerowalker
post Mar 11 2013, 19:33
Post #1





Group: Members
Posts: 239
Joined: 6-August 11
Member No.: 92828



I wonder, do you guys, who are following and testing Opus more closely.
Think that it will or it is, competing against AAC good in overall transparency?

I donīt know if AAC is one of the best in Transparency or not. But i really like it compared to MP3, which sounds very bad at 192 and below,
it sounds like it has some kind of noise filtering activated. But maybe itīs just me?

But how is Opus compared to AAC and similar in overall today?

I understand that there is no Correct answer, and that it also changes with time as Opus isnīt finished.

But just want to here what you think about it, is it a competitor, or is it just a speech codec in your ears?

Thanks:)

(I am NOT asking if itīs the Best, or what is the Best or anything like that. Just want to know how you feel about it in itīs current time)

This post has been edited by zerowalker: Mar 11 2013, 19:34
Go to the top of the page
+Quote Post
 
Start new topic
Replies (1 - 14)
wswartzendruber
post Mar 11 2013, 20:51
Post #2





Group: Members
Posts: 85
Joined: 11-December 06
Member No.: 38563



There was a double-blind test done a while back that placed Opus on top, ahead of even HE-AAC and Vorbis.

Given that the current alpha-stage encoder is showing noticeable improvement and that we're less than a year out from standardization, I see Opus improving greatly in the coming years.

Given the parties backing this codec (Google, Microsoft, Mozilla, Xiph, etc...) I see it becoming widely supported.
Go to the top of the page
+Quote Post
db1989
post Mar 11 2013, 21:01
Post #3





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



QUOTE (wswartzendruber @ Mar 11 2013, 19:51) *
There was a double-blind test done a while back that placed Opus on top, ahead of even HE-AAC and Vorbis.
Said test was clearly linked in the thread from which this post was split for being off-topic, and in which the OP zerowalker already asked other questions that could have been answered by simply following the link and acted apologetic when this was pointed out. Maybe some people just want others to spell everything out for them all the time without having to do any thinking of their own.

This post has been edited by db1989: Mar 11 2013, 21:02
Reason for edit: “the OP” was ambiguous
Go to the top of the page
+Quote Post
zerowalker
post Mar 11 2013, 21:05
Post #4





Group: Members
Posts: 239
Joined: 6-August 11
Member No.: 92828



QUOTE (db1989 @ Mar 11 2013, 21:01) *
QUOTE (wswartzendruber @ Mar 11 2013, 19:51) *
There was a double-blind test done a while back that placed Opus on top, ahead of even HE-AAC and Vorbis.
Said test was clearly linked in the thread from which this post was split for being off-topic, and in which the OP zerowalker already asked other questions that could have been answered by simply following the link and acted apologetic when this was pointed out. Maybe some people just want others to spell everything out for them all the time without having to do any thinking of their own.


Well, i think that General Transparency, is a bit different from specific Samples at certain bitrates.
Though i may be wrong there.
Go to the top of the page
+Quote Post
wswartzendruber
post Mar 11 2013, 21:07
Post #5





Group: Members
Posts: 85
Joined: 11-December 06
Member No.: 38563



I think that the concept of general transparency may be a fallacy. There will likely always be samples of sound that give Opus encoders trouble even at higher than normal bitrates.
Go to the top of the page
+Quote Post
zerowalker
post Mar 11 2013, 21:12
Post #6





Group: Members
Posts: 239
Joined: 6-August 11
Member No.: 92828



QUOTE (wswartzendruber @ Mar 11 2013, 21:07) *
I think that the concept of general transparency may be a fallacy. There will likely always be samples of sound that give Opus encoders trouble even at higher than normal bitrates.


Well thatīs surely true. The same goes for Encoding Video, nothing will always work the same at all content.

But with General Transparency, i wanted to know if people have used it as much to have a General feeling of when itīs normally Transparent.

Letīs say that MP3 is often Transparent at 256kbps.

But then it can also be Transparent at 128 at specific samples, or maybe need 320 at others.
But the General transparency would then be 256, as thatīs what people found working most of the time (This is just hypothesis).

So i was wondering if anyone have use Opus in such a way to know, where it usually goes transparent for their use, and how it compares to others, like AAC or maybe Vorbis, as those are the more capable ones i think.

I am not looking for a General Rule where it goes transparent, as that doesnīt exist, there will always be differences.
Go to the top of the page
+Quote Post
wswartzendruber
post Mar 11 2013, 21:17
Post #7





Group: Members
Posts: 85
Joined: 11-December 06
Member No.: 38563



I'm not quite sure what you're asking for.

QUOTE
So i was wondering if anyone have use Opus in such a way to know, where it usually goes transparent for their use...

QUOTE
I am not looking for a General Rule where it goes transparent
Go to the top of the page
+Quote Post
zerowalker
post Mar 11 2013, 21:25
Post #8





Group: Members
Posts: 239
Joined: 6-August 11
Member No.: 92828



QUOTE (wswartzendruber @ Mar 11 2013, 21:17) *
I'm not quite sure what you're asking for.

QUOTE
So i was wondering if anyone have use Opus in such a way to know, where it usually goes transparent for their use...

QUOTE
I am not looking for a General Rule where it goes transparent



Sorry, i have a hard time explaining it. and my English is lacking..

Well, i guess what it wrote there me contradict each other.

But easily speaking, i guess what i want it.

If people have use Opus for encoding much of there music and other sound content.
What is the usually bitrate for transparency?

I just want peoples current knowledge on it, not some real scientific proof.

Sorry for not explaining to well.
Go to the top of the page
+Quote Post
wswartzendruber
post Mar 11 2013, 21:40
Post #9





Group: Members
Posts: 85
Joined: 11-December 06
Member No.: 38563



I think the best idea may be to wait to use Opus. The rate of improvement is still very high.
Go to the top of the page
+Quote Post
db1989
post Mar 11 2013, 21:42
Post #10





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



In an effort to help, I can post what I think you’re looking for, as framed by the two different sentences quoted by wswartzendruber. Of course, please correct me if I’m wrong. You want individual users to let you know which settings were, for them, generally transparent across the samples that they have tested. You don’t expect a one-size-fits-all answer about which settings are transparent to all users or ‘the average person’. Is that correct?

Edit: Of course, given enough data-points, examples of the former could be superimposed to make some kind of prediction about the latter… but I don’t know that we’ll obtain a convincing sample size in this particular thread. tongue.gif

And I would imagine that accounts from other people in isolation are not hugely useful to predicting your experiences as a separate individual. If you’re currently unsure, I can recommend two things: test Opus for yourself on a set of files that will be representative of your library, or wait until the codec has matured further and additional multi-user tests have been conducted and reported. This really depends upon how much work you’re willing to do or, in the absence of that, how much you trust generalised conclusions about other people to be applicable to you.

This post has been edited by db1989: Mar 11 2013, 21:46
Go to the top of the page
+Quote Post
me7
post Mar 11 2013, 21:55
Post #11





Group: Members
Posts: 177
Joined: 23-August 06
Member No.: 34375



I think zerowalker wants a setting for Opus comparable to "Lame -V2", a widely accepted as a one-fits-all solution for people who don't want to bother with details. Such recommendation are important for an encoder to gain wide acceptance among common users!

Answer: It is too early. The encoder is still under heavy development and it will take some time for it to mature. Then, listening tests on various bitrates need to be performed to determine what is generally accepted as "transparent".
Go to the top of the page
+Quote Post
zerowalker
post Mar 11 2013, 22:14
Post #12





Group: Members
Posts: 239
Joined: 6-August 11
Member No.: 92828



Well both are pretty much correct.

But as db1989 said.

If i get the general idea of what people used to achieve transparency for them, i could "Achieve" a general setting that "should" work, though that is not my intention.
I really just want to know how good it is in general.

But i will do as you guys said, and wait till it has matured a bit, as you said the improvement is still increasing at a steady pace.

Thanks:)
Go to the top of the page
+Quote Post
Dynamic
post Mar 11 2013, 23:13
Post #13





Group: Members
Posts: 793
Joined: 17-September 06
Member No.: 35307



I think waiting is wise.

The transparent settings for iTunes and Nero AAC, Vorbis and Lame MP3 have been pretty well tuned with attention paid to problem samples that have arisen. Once most problem samples have been solved, an encoder can be considered pretty mature and fairly resistant to badly artifacted samples.
Go to the top of the page
+Quote Post
darkbyte
post Mar 11 2013, 23:56
Post #14





Group: Members
Posts: 107
Joined: 14-June 11
Member No.: 91517



I think because the Opus encoder is still evolving the most interesting area is portable use where some artifacts are acceptable and you can test the codec's capabilities.

For me unconstrained VBR @80kbps seems like a very good option for portable music. I'm using such files on my Note2 and with the original earphones i can hear only subtle differences between that and q5 vorbis i was using before which is mind blowing considering that the filesize is nearly half of the original Vorbis. I've also gave it a try with my Beyer DT-770 while i was programming and although i can notice the differences, the artifacts are so less annoying compared to HE-AAC where i can immediately notice the presence of SBR that i'm still very impressed with the quality. I can store this way 2 weeks of music on 10GBs. Simply amazing.

This post has been edited by darkbyte: Mar 12 2013, 00:01
Go to the top of the page
+Quote Post
kabal4e
post Mar 12 2013, 02:30
Post #15





Group: Members
Posts: 8
Joined: 10-March 13
From: Waikato, NZ
Member No.: 107144



I am personally not trying to achieve any General Transparency with music encoding, because as mentioned above it depends on the overall conditions, e.g. noisy environment, equipment one's using etc. However, ABX-ing hard samples will help you during the development stage, but for general use a person would be happy using some lower bitrate than the one proven to be transparent for a variety of hard samples.

For instance, I had a party at my place the other day. There was loud music and a few tracks were encoded with Opus 1.1-alpha at 128kbps. Even though 128kbps is not believed to be transparent, I didn't hear any complaints from the guests saying that music was to any extent 'different'. Because the people were just listening to it, and for them it sounded as it should. So, I'd agree that there is a Generally Accepted Level of Transparency, e.g. lame -V2. And for me it's Opus at 128kbps.

All above is IMHO.
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: 18th April 2014 - 01:28