Talk:Vorbis

From Wikipedia, the free encyclopedia

To-do list for Vorbis: edit · history · watch · refresh


Here are some tasks you can do:
  • Requests: Create a Spoken Version of the article
  • Cleanup: Cleanup in an attempt to get it as FA
Peer review A request has been made for this article to be peer reviewed to receive a broader perspective on how it may be improved. Please make any edits you see fit to improve the quality of this article.

Contents

[edit] Tutorial: How to play and/or use Ogg Vorbis files

Please check Help with Ogg Vorbis for advice on playing sound files on Wikipedia and other places.

If you would like to see a list of all Ogg Vorbis files available at Wikipedia (most of them under the public domain), go to Wikipedia:Sound/list.

--Saoshyant talk / contribs (I don't like Wikipedophiles) 10:45, 25 October 2006 (UTC)

[edit] Use in video games

To me, it seems pointless to list nothing except the most popular video games using Vorbis since the Xiph.org wiki has a full list anyway. Samuella 17:23, 1 April 2006 (UTC)

[edit] License

Is there any evidence backing up the claim that Vorbis is unpatened? After reading [1], I have my doubts.

Samuella 00:32, 26 February 2006 (UTC)

Sure. Xiph.org has not applied for any patents, nor do they intend to. Therefore Vorbis is totally unpatented. Perhaps you mean to ask: Does Vorbis infringe upon any existing patents? That's a difficult question to answer, given the vast number of patents and the often obscure and ambiguous scope of them. The article from 2002 you reference (which was written while Vorbis was still in beta) mostly complains about the lack of a formalized Vorbis spec and so is irrelevant to the patent issue. Xiph.org was issued a private legal opinion presumably saying that Vorbis is as safe as can be presently determined. That's as good as it gets under the US patent system. --Bk0 (Talk) 05:16, 26 February 2006 (UTC)
The only point that is sure is that Xiph does not intend to apply for any patent regarding Vorbis. There is no evidence that Vorbis is not covered partially by some external patents, although Xiph claims to have conduced some research about it.--Gabriel Bouvigne 11:56, 25 August 2006 (UTC)

[edit] Windows media player

The artical states that microsoft has a plugin for windows media player, does anyone have a link to back that up? I cant find it.

Ralph Sleigh 15:13, 20 February 2006 (UTC)


Neither can I. KH I cannot put the four tildes (202.79.62.14 00:13, 23 February 2006 (UTC))because I am not registered. (I do not have permanent IP address.)


The Directshow filters allow WMP to play Vorbis.

Samuella 22:49, 25 February 2006 (UTC)

[edit] Radio

I mentioned the use of Ogg Vorbis on CBC radio and Radio France and got reverted on the grounds that these streams didn't exist, much to my surprise since I listen to these regularly. See http://www.radiofrance.fr/services/aide/difflive.php#ogg and http://www.cbc.ca/listen/ogg.html . Is the argument that no mention should be made of Ogg Vorbis radio unless it is used exclusively, or am I missing something? --Marlow4 11:30, 17 May 2005 (UTC)

Sorry, but I really could not find the streams in the first place. I added the links above to the corresponding pages and rewrote your point back to the main page. Feel free to reformulate the sentence keeping Wikipedia:NPOV in mind. --Hhielscher 13:29, 17 May 2005 (UTC)

[edit] Non-neutral statement

The line "It may be some time before one sees more Ogg format files than MP3 files" seems a little out of place in this article. Why is it assumed that Ogg Vorbis will overtake MP3? Just because I, or anyone else, would like to see it happen does not mean it will. For that reason, I am removing this line; people can change it back if they like, but I would like to see some reasoning if so. Thanks :) 11 May 2005

I don't think it was meant to be interpreted as "Vorbis will overtake MP3, but it may take awhile" but rather "MP3 has enormous coverage and Vorbis will probably not be overtaking it in the near future." There is an unwritten "if at all" in that sentence. --Kerohazel 23:43, 28 May 2005 (UTC)

If I could I would throw my weight after Kerohazel. He has a point. 202.79.62.14 00:16, 23 February 2006 (UTC) (I do not have a permanent IP address.)

[edit] Quality

"Quality settings run from -1 to 10 and are an arbitrary metric; files encoded at -q5, for example, should have the same quality of sound in all versions of the encoder, but newer versions should be able to achieve that quality with a lower bitrate."

I would disagree with this. I think the developers have made some poor choices of late and the latest encoders are worse than earler ones.

Basically they allocate much more of bandwidth to the very high treble than they used to which shortchanges the midrange - even at the highest bitrates they never allocate very much bandwidth to the midrange. Yes the highs are a little crisper, but I don't think it was worth losing definition in the midrange where our hearing is more accute.

Unfortunately it takes much more bit-bandwidth to improve the treble than it does to improve th midrange and base (consider that the highest octave has the same amount total entropy as the entire rest of the spectrum combined).

"Many users feel that Vorbis reaches transparency (sound quality that is indistinguishable from the original source recording) at a quality setting of -q5, approximately 160 kbit/s. For comparison, it is commonly felt that MP3 reaches transparency at around 192 kbit/s, resulting in larger file sizes for the same sound quality."

Transparancy at -q5 or 192 MP3... My God some people have insensitive ears (or the tests are done on extremely poor quality equipement). I would say that in comparing music that I'm familiar with, on high quality equipement, MP3 NEVER reaches full transparency, and Vorbis only reaches it at the highest quality, and only some of the time at that.

Joshua Scholar

There have been several scientific studies using very high quality equipment and experienced listeners, and they all concluded that MP3@256kbps was generally indistinguishable from original recordings, which makes 192kbps a good estimate for average listeners and equipment. One such test was done by Heise in Germany, no idea if it's linked on the web. The most common cause of people claiming other results is not understanding (double) blind testing, and consequently grading with their imagination than with their ears, or severe hearing damage, which invalidates the assumptions the codecs make about the human hearing.

As of 2005/2006: There are some reasonable and recent tests that may tell about the current state of audio encoding. In August, 2005, guruboolez (an audiophile) posted at Hydrogen Audio the double-blind listening tests that he'd made, for classical music at 180 kbps (link at main page). For each piece, he describes the possible problems that may happen if the encoding is not good. He describes possible pre-echoes, out of tune tremolos, ringing, loss of details, etc. In the conclusion, it becomes clear that no encoder was really transparent, but also states that at least (aoTuV beta 4) Vorbis and MPC are pretty good, with one third of Vorbis samples being transparent (he states that his own limits are close to be reached). Lame, which also encodes with high quality, comes next, tied to MPC, although he states that quality is not homogeneous between samples, like Vorbis and MPC were. A much more recent test, however, went public at January, 14, 2006. It was a public (~30 listeners per sample) listening test at 135~145 kbps (128 kbps nominal), which may be considered closer to normal people for reference purposes. Although there's no transparency at this rate, as user comments show (and, surely, as they expected), the grades given to all the encoders (including here Vorbis, MP3 and AAC and WMA) are reaching the top, which means that lack of transparency at those rates are not making the sound annonyng anymore. In particular, if you are worried about being sad because you lost some brilliant part of your music by encoder's fault, forget it. Those are old times. Concluding here, when Joshua Scholar says that MP3 never reaches full transparency, he is partially right, because, altough this may not be true with, say, LAME 3.97+ encoding at --preset insane, as highly mature as it is right now, it completely depends on which encoder one is willing to use. Using an old MP3 encoder may give you a bad encoded song at 320 kbps. --Hdante 03:47, 22 January 2006 (UTC)
The most noticeable effect with low (<192) bitrate MP3 is a muddiness to the sound. This is not audible on the low quality speakers most people use with their PCs, but glaringly obvious on decent headphones or speakers. 128k MP3 sounds quite similar to cassette tape. I find high bitrate MP3 to sound quite good, but Ogg always manages to be just a little bit better while creating smaller files because I get the same quality at lower bitrates.

Could you link to specification ? Taw

See link to documentation in article -- The Anome

Vorbis and (Nanny) Ogg are both characters in Terry Prattchett books (Small Gods, and a bunch of them, respectively). I've wondered if there's a connection -- can anyone shed light on this? -- user:Tarquin

There's a FAQ about this on the ogg vorbis site. Apparently, the name Ogg is not related to pratchett, but Vorbis is. No explanation is given.
Yup. That's a really old question of mine, I dug up the FAQ on the Vorbis site & explained in the article. Thanks for checking it out. (I really must clean up my old talk!) -- Tarquin 13:37 Jan 2, 2003 (UTC)

I'm making a list of Vorbis-supporting non-PC hardware (portable MP3 players), does anyone think it would be a good or bad idea to put it here? Good, because if anyone else finds something they can add it, bad because it would be mostly links and paragraph-long descriptions. It may eventually go onto [vorbis.com]. -- User:Tcfelker

---

Any comparision between Ogg / mp3 / mp3pro / etc?

This line looks like a bit of an over-zealous statement:

"In the commercial sector, Vorbis has already replaced MP3 as the de facto standard audio codec with many newer video game titles employing Vorbis as opposed to MP3."

Is there any proof for this? Can it really be shown that of all the new games released more are using ogg than mp3 - this would be required for it to be a standard. I'll edit the line in a week if no one can show any evidence Tompagenet 19:41, 11 Apr 2004 (UTC)

No information was placed here so I have changed the text in the article Tompagenet 13:48, 19 Apr 2004 (UTC)


The Unreal Engine uses ogg since UT2003. What about other major game engines -- Doom III for eg? -- Tarquin 14:25, 19 Apr 2004 (UTC).

I have just discovered that Far Cry also uses ogg vorbis. praetor_alpha 04:10, 7 January 2006 (UTC)
A lot of Game developers have taken a liking to Ogg Vorbis due to the fact the decoding libraries are easy to implement in games and they don't have to pay royalities. Vorbis is an official extension of the OpenAL extension library so it could very well be used by the developers of Doom 3 and many other multiplayer games that are using the libraries. After the big Creative lawsuit with John Carmack and Creative Labs we can assume we know his stance on the situation. -Hydrogenaudio Lurker.

Thanks for the response, Tarquin. I have updated the article to say that games do use the codec, but I have removed the statement that it is the "standard" in games. If it is in fact true that more games use ogg than any other particular codec then as soon as we have some proof we can change this back. Tompagenet 14:58, 19 Apr 2004 (UTC)

What about adding a link to this list? http://wiki.xiph.org/GamesThatUseVorbis User:apecat

[edit] Is Vorbis inherently VBR?

Is Vorbis inherently VBR? ✑whkoh | ✖ 11:12, Nov 1, 2004 (UTC)

yes, the standard quality settings are vbr, altough I think it is possible to specify min, max and nominal bitrates. I guess it you specify them all the same it would be cbr Robneild 17:37, 1 Nov 2004 (UTC)

No, Vorbis is not inherently VBR by itself. However, current Vorbis encoders are inherently VBR.--Gabriel Bouvigne 12:20, 25 August 2006 (UTC)
The vorbis format is intended to be inherently VBR by the Xiph designers. However, at a low level, the format is just as capable of being CBR (called 'managed' mode) as the standard VBR quality settings. --Bk0 (Talk) 00:39, 26 August 2006 (UTC)

[edit] Use on Wikipedia?

The Theora and Vorbis royalty-free, open source video and audio codecs could help to make it possible for Wikipedia to host narrated video clips for those articles that explain processes or mechanical devices which are difficult to comprehend using words and pictures alone.

If anyone can point me to some examples of Ogg Vorbis/Theora being used here on Wikipedia, I would be most grateful, as I would like to contribute video and audio clips if it is possible to do so. — DV 07:52, 22 Dec 2004 (UTC)

Hi, DV. This page might point you in the right direction: Wikipedia:Sound. There's a link of there pointing to a cache of sound files stored on Wikipedia. I believe Ogg is already the format of choice. HTH, --Chopchopwhitey 15:32, 22 Dec 2004 (UTC)

Why don't we use a format that people can, you know, actually listen to? I don't understand why wikipedia is riddled with .oggs instead of .mp3s. :p

Anybody can use Ogg Vorbis. There's a big list of software that supports it in the main article, including Windows Media Player and iTunes. --Bk0 (Talk) 12:45, 30 December 2005 (UTC)

[edit] Codec comparisons

"Listening tests [2] have attempted to find the best quality lossy audio codecs at certain bitrates. They have concluded that all modern audio codecs including Ogg Vorbis, mp3PRO, AC-3, Windows Media Audio, MPC and RealAudio perform significantly better than MP3 due to the superior technology used."

No they didn't. The tests clearly show mp3s to be in the middle of the pack of the codecs used. And what's up with the line about "superior technology"; Does someone have a thing against mp3s?

That's true. 2006 tests show that (LAME) MP3 is tied with every other format at around 135~145 kbps (see below or main page for link). And about "superior technology", well, yes, newer formats do have bits of superior technology revolving the main code. The Musepack article describes some of them. --Hdante 04:14, 22 January 2006 (UTC)
"Very generally, these codecs achieve the equivalent of MP3 128kbps at around 80kbps."

Since the tests showed mp3 to be average among the codecs tested, this is clearly false. Furthermore, I have yet to see any evidence for ANY codec achieving the equivalent of 128kps MP3 at 80kps. The listening tests linked on this page (by Roberto Amorim) show only a slight difference between codecs when compaired at 128kps. Interestingly, which codec was "better" varied depending on the song chosen. (Which suggests that different codecs are better at representing different types of sound dynamics.)

August 2005 tests show that aoTuV beta 4 at 80 kbps is comparable to a LAME 3.97a ABR 128 kbps for pop music. Also, 96 kbps tests show that aoTuV beta 4 is better than LAME ABR 128 kbps for pop music. This is pretty much close to stating that a codec achieves at 96 kbps the same quality as a 128 kbps MP3 (even though LAME VBR @ 140 kbps is tied with aoTuV @ 140 kbps). See main page for links. --Hdante 04:14, 22 January 2006 (UTC)

I'm revising these lines to reflect what is actually shown in the listening tests. I hope we find other nonbiased listening tests for comparison.

- Pioneer-12 10:59, 20 Apr 2005 (UTC)

I think we should use this listening test when comparing formats at ~128kbps. It's a very recent test, and it shows that LAME MP3 isn't far behind Vorbis at this bitrate. Of course, ~80kbps Vorbis is probably superior to 128kbps Shine MP3. However, if we're going to make a comparison, we should probably compare the best (or most common) encoder for each respective format. --MuncherOfSpleens 18:25, 21 January 2006 (UTC)

Don't rely upon GuruB tests he is overly zeaolous listener with "golden ears". In fact you get say he O.C.D on private listening tests. Sabastian mutliformat test is a great indication how the Psychoacoustics models are doing in terms of optimal performance. It doesn't have to do with transparency all of the time, Ogg Vorbis and AAC are clearly to superior to MP3 when it comes to objective technology within them. -Hydrogenaudio Lurker 16:30, 21 Jan 2005 (UTC).

I had already added the link to Sebastian's multiformat test. BTW, guruboolez participated in that test, and if you pull his individual results out from the group, you can see that he preferred aoTuV and iTunes over the others at this bitrate "Anonymous River" site (Japanese language). "Golden ears" and "overzealous" are denigrating terms. I would say rather that guruboolez is a very experienced and sensitive listener, although an individual's preferences may not represent a group's preferences. Ff123 14:41, 22 January 2006 (UTC)

Ah, I see that you did. I should have searched for a link to the main page instead of the results page, I guess. --MuncherOfSpleens 17:23, 22 January 2006 (UTC)
We should refrain from using the term golden ears as there is no such thing. If someone claims they have them, then go ahead and make fun of them.
I don't like the term golden ears either, but what makes you think there's no such thing? All you have to do is inspect the responses and the scores given by the various listeners of these blind codec comparisons to know that there is a wide range of individual abilities to hear problems. I.e., there are plenty of people who hear absolutely no problems (rate scores of all 5's on their reports), but there are lots others who do. Among those who hear flaws, there are a select few who seem to hear more than others. Golden ears is a disparaging term to describe these people, but it is accurate. Ff123 19:40, 13 May 2006 (UTC)

[edit] File extentions

I didn't see any discussion of this in the article, but I was under the impression that Vorbis files (at least on Windows computers) had the extention ".ogg" . I can't imagine they'd have the extention ".vorbis".

The article says "[Vorbis] is frequently used in conjunction with the Ogg container and is then called Ogg Vorbis. It is often mistakenly called just Ogg (or "OGG")."

But audio formats have typically been known by their extension name - mp3, wav, wma, aac, etc. So is it simply people making a mistake, or is the situation more complex?... --4.244.180.14 05:16, 6 May 2005 (UTC)

Filename extensions are basically meaningless, although you're probably correct that in the early-mid 1990s the name MP3 originated from the extension being used on Windows machines. There are several extensions being used for MPEG-4 AAC, for example: *.aac, *.mp4, *.m4a, etc. There's no reason why people can't start using more descriptive extensions for Ogg Vorbis files (*.ogv, *.oggvorb, *.vorb) since the 3-character limitation is just historical convention. --Bk0 12:54, 6 May 2005 (UTC)
No, MP4 is a container bitstream also. I don't understand why people get entirely up and arms about this. Most of the time Vorbis files are placed within an Ogg container bitstream (by default that's what the encoder does). You can place Vorbis files in other container bitstreams like MKA that's allowed. People captalize it for some reason in upper case letters. -Hydrogenaudio Lurker 16:02, 21 Jan 2005 (UTC).
AVI is a container format too, and is certainly capable of holding MP3 or Vorbis streams -- but if you put an MP3 stream in an AVI all by itself, do you give it an .mp3 extension? No, you give it .avi, because it's an AVI containing an MP3 stream. Same goes for Vorbis. -Crazycomputers 7:33, 17 March 2006 (UTC).

[edit] id3?

Does anyone know if or what Ogg Vorbis has that is equivalent or comparable to id3 tags? —This unsigned comment was added by 89.138.121.82 (talk • contribs) .

They're called Vorbis tags and they're actually more flexible than ID3. --Bk0 (Talk) 11:52, 5 April 2006 (UTC)
Care to write something about it in the article? —This unsigned comment was added by 89.138.121.82 (talk • contribs) .
Done. --Bk0 (Talk) 01:51, 6 April 2006 (UTC)

It sounds like ID3v2 has the exact same capability. Personally, I only encode lossy in mp3 format and I use both ID3v1 and ID3v2 tags for everything, including arbitrary data such as disc number or replaygain.

[edit] Pictures

We need more pictures if this could be a featuerd article. Any ideas? X [Mac Davis] (SUPERDESK|Help me improve) 04:59, 27 October 2006 (UTC)

But what sort of pictures can we have, in an article about an audio codec? Spectrograms are out of the question. --Kjoonlee 05:08, 27 October 2006 (UTC)
I'll be taking some screenshots of software dealing with Vorbis, I suppose.--Saoshyant talk / contribs (I don't like Wikipedophiles) 10:00, 27 October 2006 (UTC)

[edit] File format infobox

Regarding the recently added infobox, should .ogg be the only file extension mentioned? As I understand it, Vorbis-encoded audio can be encapsulated in a variety of container formats and thus the extension could be .ogg or .mka or even .avi. Sam 00:32, 28 October 2006 (UTC)

The same is true of other file formats, but I doubt they list all of them in their infoboxes. I think .ogg should be OK on its own. --Kjoonlee 03:48, 28 October 2006 (UTC)
Ah. How about "Most commonly .ogg"? Sam 14:31, 28 October 2006 (UTC)
I think it would be safe to just mention '.ogg', as it is a reasonable "default." If you want to mention that Vorbis data can have different forms, I think the article's main text would be the better place.. Just my two cents. --Kjoonlee 18:31, 28 October 2006 (UTC)
I agree. Sorry, Sam, but encyclopedia-wise, it's what makes more sense.--Saoshyant talk / contribs (I don't like Wikipedophiles) 10:48, 30 October 2006 (UTC)

[edit] The Vorbis II merge discussion

I reckon that article shouldn't even exist, because there is no such thing as a Vorbis II. As a Xiph.Org volunteer I can tell you that Monty is working on a next generation audio codec called Ghost, some of the specification is already public. There's also Ogg2, which will replace the current container, but there's still work needed (BTW, if someone wants to help there, they are more than welcome). There's no Vorbis II. I'll remove the merge notice in two days and ask for the deletion of the article if no one offers proof of otherwise.--Saoshyant talk / contribs (I don't like Wikipedophiles) 10:48, 6 November 2006 (UTC)

[edit] Tremor

Some disappointing information about Tremor (that probably would get deleted again in the main article):

  • Tremor has a high RAM footprint compared to MP3 decoders (trunk: 100 kB, lowmem branch: 40 kB); for comparison the Helix MP3 decoder uses 27 kB, and there are still better ones available
  • Tremor uses dynamic memory allocation extensively, a no-no for many embedded applications
  • Development seems to have stalled. The last significant modifications were made years ago, and as of October 2006, the Tremor sourcecode still has memory leaks that were already discovered in 2003.

--131.188.23.186 17:55, 9 November 2006 (UTC)

Retrieved from "http://en.wikipedia.org../../../v/o/r/Talk%7EVorbis_070a.html"