Talk:Four-channel Compact Disc Digital Audio

Is this article factual?
We don't need articles about rumors. You are supposed to find references before posting, not post rumors and then vainly hope that someone else can provide the references to support it (more likely, the rumor will become someone else's reference). Please rethink your approach. Mirror Vax 01:02, 25 May 2005 (UTC)


 * Hi Mirror Vax. I agree.  Rumors belong on the talk pages &mdash; to be investigated.  (That's why we're here.)  Sorry if you perceived the article to be about rumor, but it's not. It's about a very real feature of a very widely used standard.  This feature is obviously not well known and untapped for a variety of reasons.  Lots of underutilized and dead standards are documented on Wikipedia.  I personally find this one interesting.  Thus, my call for contribution.  :-)  --Ds13 05:08, 2005 May 25 (UTC)


 * If it's a "very real feature of a very widely used standard", then why isn't there any evidence for it? Why are you unwilling to cite your sources? (I don't believe the Red Book standard was your source, since it's not available on the web and costs a small fortune to buy). Mirror Vax 18:01, 25 May 2005 (UTC)


 * Hi, Mirror Vax. I'm not sure how to respond... Four Channel Audio is in the Red Book, so citing that as my "evidence" seems quite sufficient.  Please take the time to follow the reference in the article to Cdrdao and you'll a reference to four channel audio on the very front page of that particular application's website.  The only other thing I can suggest for someone still doubting (I promise, this isn't a conspiracy!) is to grep the source code of Cdrdao (conveniently, it's GPL'd), and its man page, and at least one of its example TOCs to find this feature in hiding (along with other lesser used features, such as ISRC, the copy flag, and pre-emphasis).  It's there, and Cdrdao is a mature, established CD burning app.  This is just one app that I'm familiar with myself &mdash; there may be other apps which support the feature also &mdash; thus, my call for contributions in the first place, which brings us full circle!  Thanks for your concern, and happy editing! --Ds13 20:38, 2005 May 25 (UTC)


 * You can't cite as a source a document that you haven't even read. It may be a source for someone else, but it isn't one of your sources. The actual sources you cite do not support all the claims made in this article, not by a long shot. All you have is the existence of a "4-channel" bit (which I was, of course, aware of). Where is the source that says that CD players implement this bit in the manner you describe? All I'm asking is: where did you get this information? It's a simple question, easy to answer. Mirror Vax 21:14, 25 May 2005 (UTC)


 * Mirror Vax, if you wish to tag this article as a stub or factually-disputed, I'm fine with that. In the mean time, I'm satisfied with the factual basis of such a young article (less than two days old).  Googling for four channel CD audio turns up references to this feature, in addition to the compact disc article itself claiming that this is specified in the Red Book. Adding to that the existence of the Cdrdao app, docs, and code referring to this feature should put factual questions to rest, in my opinion. However, if you wish to change the article intro to read "... is allegedly specified by the Red Book..." or any other bold edits, I'm also fine with that.  I hope you can assume good faith in my intention to create factual, non-biased edits throughout Wikipedia.  I will assume good faith in your desire to scrutinize citations and if I find more details, specifications, use, or publications of this feature, rest assured that this article will reflect it! Cheers. --Ds13 00:06, 2005 May 26 (UTC)


 * The links you provide do not contain the information in this article, so they cannot possibly be your sources. Where did the information come from? Why can't you answer that simple, basic question? I'm not asking for vague "references" to a bit, I'm asking for sources of the specific information in this article. Citing sources is recommended practice (see WP:CITE), so I'm not making any strange or unusual demands. I'm not going to edit this article. If I did edit it, it would end up blank, so I don't think you want me to be bold. Mirror Vax 01:23, 26 May 2005 (UTC)


 * I'd just like to say what Ds13 and others in this discussion seem to have been too polite to say thus far: Stop being a dick. It's a thing that's known to exist, from many sources, not least the official CD standard, and it's something I've personally been aware of since sometime in the 90s. You want to argue with us next that the sky isn't blue on a sunny summers day, and that healthy grass isn't green? Sure, the details of it are up for endless debate, speculation and just plain fantasising (I fancy 4 discrete channels at 29400Hz, using 12-bit A-law (or 32072/11 bit,35280/10 if the effect of reducing headroom proved perceptually worse than losing bits), if we have to stick to the same transfer rate and avoid overly complex pre/post-processing)... we'll just never know what it could have been given that it was never finalised and possibly never really developed much beyond an idea of something that would be nice to have. However, we do know it existed. And the fact of their being a debate or some evidence of what MIGHT have been is still something that can be included, as it's a significant part of the mystery. 193.63.174.10 (talk) 15:49, 18 November 2009 (UTC)


 * This article mentions the 3rd bit in the Q subcode channel being used to switch between 2-channel and 4-channel audio: http://www.ee.washington.edu/conselec/CE/kuhn/cdaudio2/95x7.htm (scroll down to the section on subcodes). The article also excerpts some diagrams from the IEC908 standard, which leads me to believe it was written with reference to the standard.  Mirror Vax: given that there are multiple agreeing secondary sources, do you really doubt the facts in this article?
 * The only complaint I could think of is that the topic probably isn't important enough for its own article — it might fit better in the main Compact Disc article (of course, with information on a few 4-channel audio releases it might be enough for its own article) --James 29 June 2005 01:33 (UTC)


 * The entire relevant content in the article you cited is, "Bit 3 controls the number of channels (2 or 4)". That trivial and well-known tidbit in no way confirms the claims of this article. I can find no evidence that the bit works (or was ever intended to work) as described in this article, and besides that the description is implausible. None of the sources cited by the author contain the information in this article. I have repeatedly asked for the author's actual source(s) but he won't say. This leads me to the conclusion that the article is either a fabrication or the sources are so bad that the author is embarrassed to reveal them. Mirror Vax 29 June 2005 03:06 (UTC)


 * To make it easier for Ds13 to fix the article, it would be helpful if you could identify which claims you dispute.
 * That the Red Book CD standard contains a provision for 4-channel audio?
 * That the number of channels are controlled by a bit in the "Q" subcode channel?
 * That doubling the number of channels requires halving the data-rate for each channel?
 * That each channel in a 4-channel stream is encoded at 16-bit 22.05kHz PCM?
 * That commercial 4-channel audio releases are uncommon? (following this up is already listed in the todo, but seems plausible).
 * That all CD recorders can record 4-channel audio? (this should probably be changed to "most", but I'd hazard a guess that > 99% of PC CD recorders can, since the subchannel data is written along with the rest of the audio data).
 * That CD recording software capable of 4-channel recording is uncommon but exists? (the docs for cdrdao do cover it so there is at least one, and the Apple CD writing APIs support it to)
 * That software developers are sometimes discouraged from using 4-channel recording in their software? (this one includes a reference to the Apple docs)
 * Just saying that the article is factually incorrect without specifying which sections you dispute does not help anyone. --James 29 June 2005 14:08 (UTC)


 * Sorry if I wasn't clear - I dispute ALL the claims in this article. That's why I said earlier that if I were to be "bold" and edit it, it would end up as a blank page (but that would be rude so I'm not going to do it). Here's what I believe to be factual:
 * There is a vestigial "4-channel" bit defined, but it never did anything and is not useful.
 * The original intent behind the "4-channel" bit is unknown, but probably was not as described in this article. If I were to hazard a guess, it would be that it was intended to signify some sort of encoded form, not four discrete channels.
 * Mirror Vax 29 June 2005 16:43 (UTC)


 * If you say you are disputing all claims, then I don't believe you have thought things through. Some of the claims may be speculation, but some of the ones I mentioned above are properly referenced.  I don't have any CD player capable of handling CD-TEXT or CD+G, but I don't find the two standards implausible.
 * I'd be very surprised if the Red Book standard didn't define what the audio track should contain if the 4 channel audio bit is set. If the content was not defined then they would not have defined the flag bit either, leaving it reserved for use in future standards documents.
 * Now, I did a little experimentation with cdrdao, and burnt a disc with the same data written in 2-channel and 4-channel mode into separate tracks, then another track which interleaved 1 2-channel sample from one file with 1 2-channel sample from another and wrote it as a 4-channel track. Cdparanoia reported the right number of channels for each track.
 * two players would not play the disc (one a DVD player, the other a CD player)
 * two players played the disc, but did not appear to treat the 4-channel tracks differently to the 2-channel tracks (one was an LG DVD player connected to a digital receiver/amplifier, the other was a fairly standard Windows computer).
 * These results seem consistent with this being a rarely used piece of the standard that is no longer recommended for use (if a particular case is not tested in some code, it is usually buggy).
 * Out of the claims in the article that I listed above, the only two that I can't see any evidence of are (3) and (4). If those two claims were removed, which specific claims would you still dispute? --James 30 June 2005 10:41 (UTC)


 * I have thought things through and I reiterate that I dispute all the claims, every one. None make any sense if the facts are as I believe them to be (see above), and nobody has provided any references (not even bad ones) for the claims of this article. I have looked and found no evidence whatsoever that the "4-channel" bit does anything, or ever did anything. Your test results are, not surprisingly, consistent with that. Mirror Vax 30 June 2005 11:12 (UTC)


 * The article does contain references for some of the claims (not all of them — as I said, I have no way to verify what encoding was intended for 4-channel CD audio). Multiple independent secondary sources for claim (1) (cdrdao documentation, Apple's CD writing API docs).  The same for (2) (e.g. the main Compact disc page).  For claim (5) it is fairly obvious that the releases are uncommon.  With respect to claim (6), I recorded some tracks with the bit set using my CD writer, and it was using the generic MMC protocol that every modern CD writer supports.  For claim (7), an example of such software has been pointed out above (again, cdrdao).  A reference for claim (8) is included in the article.
 * So again, which of these remaining specific claims are you disputing on a factual basis? It is not exactly uncommon for sections of a standard to end up being unused, and end up unimplemented as time goes on. --James 30 June 2005 13:53 (UTC)


 * James, thanks for explicitly identifying the claims made in the article. I think this is the path to progress.  Even if that progress is deletion, as Mirror Vax suggests, we can now do it rationally, claim by claim as each claim is found to have no support.
 * I'm still looking for more evidence for claims 3 and 4, but for now, they might be the ones to remove. I'm content with the support for the other six claims, but we'll see what Mirror Vax has to say about your observations on those.  Incidentally, those two claims came from searches for this topic on recording and audio discussion boards.
 * I've looked at the Cdrdao source where the FOUR_CHANNEL_AUDIO, four_channel_audio, and 4CH flags are used and it's clear that the four channel audio flag is being checked and set (along with pre-emphasis, etc.) One of the other bits, SCMS is explicitly unsupported and will result in an error if set, however right next to that, the four channel audio bit appears to be used without incident (if transparently/simply).  If you find code interesting, grab the source to Cdrdao-1.2.0 and check line 484 of utils/cue2toc.c &mdash; one of several references.  There is reasonable math behind the half-rate theory (i.e. media limits, CLV, and only one head/laser), but at this point I agree it should be considered speculation.  On a side note, what sample size and rate did you use for the raw data given to Cdrdao in your tests, James?  16/44.1 16/22.05? --Ds13 June 30, 2005 16:37 (UTC)


 * Thank you for finally admitting that the article is speculation. Once it is in doubt that the bit does ANYTHING, I don't see how the other claims have any meaning. What sense does it make to talk about a "4-channel" feature when that "feature" consists of a vestigial bit that does nothing, never did anything, and never will do anything? Now if you can document the history of it, what the original intent was, then you would have a good little article. Until that information surfaces, there's really not much to say. Mirror Vax 30 June 2005 21:38 (UTC)


 * Mirror Vax, you misread what I meant. To repeat, I'm still looking for more evidence for claims 3 and 4, but for now, they might be the ones to remove.  There is reasonable math behind the half-rate theory (i.e. media limits, CLV, and only one head/laser), but at this point I agree it should be considered speculation.  So that's claims 3 and 4 and they are gone.  The reason why there is speculation about the sample size & rate at all is because there is clearly a feature and it's not clear what the standard allows.  Yes, I said "standard", as stated in multiple versions of Apple OS developer documentation about 4 channel audio, which references the Red Book standard.  (Cited.)  Mature software not only refers to the existence of this feature, but claims to implement it.  (Cited.)  Etc.  Now, speculation about the sample size and sample rate in no way contaminate the factual basis of claims 1, 2, 5, 6, 7, and 8 which have references and support, as James observes.  What do you have to say about these claims?  Feel free to add to that list of disputed claim.


 * If you're going to be the one to label an article as "factually inaccurate", you should be prepared to point out one or more specific unfactual claims (e.g. a sentence or two). --Ds13 July 1, 2005 00:08 (UTC)


 * I think I've explained by position adaquately, but I'll try one more time. You are leaping to the conclusion that there must be a real working feature since there is a bit. But standards often define bits that end up being unimplemented, meaningless, and useless. It's not unusual at all. I believe this is such a case, given the complete lack of evidence that the "4-channel" bit does anything useful. If that's true, then each of these statements is false or misleading:
 * "Four Channel Compact Disc Digital Audio is a four-channel audio format specified under the Compact Disc Digital Audio Red Book standard."
 * "Storing four simultaneous channels of audio instead of the usual two requires appropriately set control bits in CD Subcode channel Q."
 * "Even though the Red Book standard allows for four-channel CDs, there are very few mass-marketed releases in this format."
 * "All CD recorders can write discs with the necessary subcode channel data for four-channel audio. CD authoring software which can generate four-channel audio data is less common. One such application is cdrdao."
 * "While four-channel audio is part of the Red Book standard, software developers are often discouraged from implementing this feature in CD burning software due to newer and higher fidelity media formats being available, such as Super Audio CD and DVD-Audio."
 * - Mirror Vax 1 July 2005 01:32 (UTC)


 * The input files I passed to Cdrdao were standard 44.1kHz 16-bit stereo wav files. So for one of the tests, I was writing exactly the same data to two tracks, with and without the flag set. The program did not like it when I passed it a 22.05kHz quad wav file (it said the file was corrupt).
 * I agree that the details you give about how 4 channel audio data is encoded on the disc are plausible. However, I haven't seen any sources for this (Cdrdao isn't much help here, since it is just writing a sequence of 16-bit samples in either case.  Cdparanoia isn't either, since it hard codes the wave file output parameters).  I suggested removing the claims as a way to move forward, but Mirror Vax seems a little stubborn. --James 1 July 2005 00:53 (UTC)


 * Agreed. I have removed the statements regarding sample size and sample rate from the article. The remaining statements are referenced to my satisfaction but I'll get in touch with the developers of Cdrdao to see if they have other sources for their information. In the mean time, maybe Mirror Vax can raise specific objections to the specific statements you identified. --Ds13 July 1, 2005 01:15 (UTC)

Released CDs?
I've left a citation from Apple's Developer Connection documentation (in addition to the Cdrdao project and Compact Disc parent article references), so hopefully we can agree that this feature exists now. On to more interesting research... In several places on the net (e.g. in recording enthusiast discussion groups, programmer forums, etc.) I see it said that "very few" CDs have been released with four channel audio. (Not surprising, since consumers have rejected quite a few multichannel audio-without-video standards over the decades.) Anyways, I've added a wiki to-do to determine what these "very few" CDs actually are, or if in fact there have been none. --Ds13 18:17, 2005 May 26 (UTC)


 * Anyone aware of commercially released four channel CDs? I read rumors here and there of old quad recordings having been re-released on CD, but I can't find any definitive catalog entries.
 * Here's one such rumor... Four of Harry Nilsson's albums were originally recorded and mixed for quadraphonic sound; someone close to Harry has remastered the quad mixes into four-channel CDs that decode very nicely on a home theater surround system. Sorry,they're not commercially available, and probably never will be, and that's a shame, because in surround sound, these albums are incredible. I was given my copies as a gift by someone who knew how big a Harryhead I am; http://www.gerrold.com/soup/2003_12_28_archive.htm  --Ds13 19:43, 2005 May 24 (UTC)


 * Can you also provide proof of a Home Sound System ever produced that is capable of decoding these so called 4 channel CDs? (as these are distinctly different to Dobly Surround and DTS encoding methods. --User:203.4.129.6


 * Nope, I'm not aware of any. Supporting this would seem to require adding external jacks and internal circuitry to CD players (read: spend money) and since consumers of the era had just finished rejecting several quadraphonic audio standards, manufacturers (& publishers) really wouldn't have a strong reason to get behind this (read: spend money). I believe similar reasons apply to CD+G decoders and outputs &mdash; something you rarely find on CD players, despite it being an interesting idea. --Ds13 07:16, 2005 Jun 15 (UTC)


 * According to the web page The Vingage Knob TVK there was a Pioneer model back in 1981, the model 1516, which supported quad output. Check the Pioneer #1516 under the section TVKº11 Invisibilia. In some place I read that it used 32 kHz sampling frequency rather than the (stereo) CD standard of 44,1 KHz; the bits/sample figure was 12 or 14 rather than 16. But I don't keep the source I read this info from so you can happily remove this last comment or file it as "another urband legend". Well, OK, I'll check it thoughtfully later on. --88.29.17.61 (talk) 12:48, 7 March 2009 (UTC)

The preceding might be thoroughly confusing to anyone not familiar with surround-sound formats.

Anyone is free to release a surround-sound CD that encodes a plurality of channels into stereo (two channels). Almost every Nimbus CD has been Ambisonically encoded, and there were a fair number of CDs using Dolby MP and Circle Surround. Such usage is not part of the Red Book standard, nor does it need to be.

The Red Book standard was for four "discrete" channels. The assumption was that playing time would be halved to accommodate the added channels. This automatically renders the disk incompatible with the basic two-channel CD format.

This article rather misses the point that four-channel CDs failed because of the lack of playback compatibility. For full compatibility, the "main" channels would have to be Lf+Lr and Rf+Rr, with the "surround" channels Lf-Lr and Rf-Rr, as with CD-4 phonograph records. This would let the two-channel listener hear "everything", if all two-channel players recognized the disk as a surround recording, and extracted only the "main" channels. As the proposed four-channel format was fundamentally incompatible from the get-go, and the specs didn't require stereo players to be compatible with multi-channel recordings, four-channel Compact Disk flopped. WilliamSommerwerck (talk) 12:50, 26 December 2011 (UTC)

Getting closure on factual nature of the article?
I checked three versions of the official Red Book (all in paper version, so no 'searching' possible). The June 1980, the Sept. 1983, and the Nov. 1991 issue give the same remark on page 1, "Main parameters", namely -Number of channels: 2 and/or 4 simultaneously* sampled. In the footnote: *In the case of more than two channels the encoder and decoderdiagram has to be adapted. (I literally copied the Dutch/Japanese version of the English language). That is all that I could find in the above three official texts. I do not understand at all what the CD founding fathers meant with the term 2 and/or 4 simultaneously sampled, but that is a different story. Also I could not find any 'adaption' of the encoder and decoder diagram or subcode channel assignment. So in conclusion: the Red book mentions the possibility of a 4-channel version, but it has never been implemented. Also I have never seen or heard of any later implementation. Dsc 14:11, 27 July 2005 (UTC)


 * Thank you Dsc. This is what I've been "stubbornly" saying. Perhaps Ds13 will finally be convinced? Mirror Vax 17:28, 27 July 2005 (UTC)

A few months ago I made a remark regarding the 4-channel issue in the wikipedia CD article. It read 'four-channel sound is an allowed option within the Red Book format'. Today I changed it into 'four-channel sound is an allowed option within the Red Book format, but it has never been implemented'. I think that that is all we can say. Dsc 20:01, 27 July 2005 (UTC)

Yes, thanks for the sleuthing, Dsc. I see it this way, now: we're dealing with a semi-official standard that didn't go far enough in spec and was left stillborn by the marketplace. I suggest (hopefully to Mirror Vax's pleasure) the removal of whatever needs to be to get the article back into "factual" shape. The remaining facts being 1) what the Red Book does say and what it doesn't say (e.g. sample size/rate, etc.), 2) current softare (Cdrdao) and docs (Apple Developer Network) that do go as far as the Red Book prescribes (subcode flag), realizing that setting a bit in the burned data alone isn't enough. I'm hoping that anyone else in the future searching for the history or state of 4CH audio on CD might profit from our effort and debate. --Ds13 00:11, September 12, 2005 (UTC)

Would the "four-channel" bit have indicated four discrete channels?
This article is about the quadraphonic bit in the CD-DA spec. It is not about a four-channel CD created using matrixing. Creating these is trivial; burn a conventional stereo CD-R with a Dolby MP encoded soundtrack and there you have it. This article is about CDs with four discrete channels.

I raised the topic of quad CDs on the rec.audio.high-end newsgroup seven times between August 1995 and January 1998. Despite the fact that many pro audio people frequent this group, nobody ever mentioned the existence of a quad CD or player. The topic was also raised on the sursound e-mail discussion list in January and March 2005. Again, none of the many audio pros on the list knew of a quad CD or player which had been commercially released.

It is impossible to prove a negative, so actual proof of the non-existance of a four-channel CD or player will always be elusive. However, given the above, I feel very confident that they never existed as commercial releases.

Martin.leese 19:43, 24 March 2007 (UTC)


 * I think I understand our disagreement. In my view, this article is not about "CDs with four discrete channels." It's about the underspecified "four-channel" feature hinted at by the Red Book (and denied by IEC 908, which is at least as authoritative as the proprietary Red Book specification). The Red Book did not specify the meaning of the "four-channel" bit. We don't know whether the Red Book's "four-channel" bit was intended to signify four DISCRETE channels or merely signal the presence of some sort of matrix encoding.


 * So saying "This mode would have included four separate channels" is too strong. We have no idea whether the mode would have included four separate channels or not. I think you're making the assumption that the purpose of the bit was to signify four discrete channels, but that just goes beyond the available evidence.


 * On the other hand, you're right that matrixed four-channel CDs do exist. We're not trying to deny their existence -- I'm just saying that the purpose of the Red Book's "four-channel" bit may or may not have been to signal the player to enter a matrix-decoding mode, or any other crazy scheme. The point is it was never specified, and was dropped from the IEC version of the specification.


 * KeithWinstein 06:08, 25 March 2007 (UTC)


 * But what you are suggesting does not make any sense. There are an infinite number of possible ways to matrix-encode four channels down into two. Techniques that were actually released into the marketplace include SQ, QS, Electro Voice, UMX, Dynaquad, Matrix H. A single "four-channel" state indicating all of these would have been useless. This is because they all needed to be handled differently with different decoding parameters. Your suggestion that this whole group was being referred to as a single entity is not sensible. The only interpretation that makes sense is that the single state referred to four discrete channels.


 * Given that your interpretation does not make sense, I am forced to revert your changes. I will do this only after giving you time to respond.


 * Martin.leese 20:54, 25 March 2007 (UTC)


 * This does not quite follow. That there are many ways to perform matrixing does not mean that the bit, if specified, could not have referred to a single method (I was going to cite a cinema theatrical example here, but I think it is too confusing). If matrixing were to be used, certainly a standard form would be selected. Ambiguity in potential types of matrixing used is not a legitimate demonstration that the channels must have been discrete. Why don't you guys contact the principal authors of the red book standard and ask them? It's not like they're dead (are they?) jhawkinson 13:14, 26 March 2007 (UTC)


 * Martin, indeed, there are several possible matrix methods that could have been chosen, if the meaning of that bit had ever been specified, which it never was. There are also several possible discrete-encoding methods that could have been chosen! The designers could have chosen a four-channel discrete mode with 22.05 kHz sampling and 16 bits per channel. Or they could have chosen a four-channel discrete mode with 44.1 kHz sampling and 8 bits per channel. Or they could have chosen a four-channel discrete mode with two 16-bit 33.075 kHz channels and two 8-bit 22.05 kHz channels. Or they could have chosen a four-channel discrete mode with 44.1 kHz sampling, 16 bits per channel, and 2x rotation speed. The "four-channel" bit no more indicates a particular discrete encoding (of the many that could have been chosen) than it does a particular non-discrete, matrix mode (of the many that could have been chosen). Does that clarify why I think we need to include the possibility of a non-discrete encoding? Thanks for giving me time to respond.


 * There's so many ways they could have implemented it... maybe it was even just "reserved for future use", and they hadn't yet figured out how to do it. It may have involved some variation on A-/mu-law or ADPCM encoding (such as the playstation's low-tech videos make good use of - reduce sample rate to about 4/5ths, encode as ADPCM, and now you can stuff non-MPG compressed audio into a 2x rate stream alongside MJPEG-esque video to be decoded by a 33mhz CPU with little difficulty). Or a mix of compromises. How about 4x 25.2khz at 14-bit depth (harking back to the original putative specs)? 29.4khz at 12-bit? (The genius choice of the multi-divisable 44100 as the base sample rate means all these figures are absolutes, btw, not rounded) ... Or even 33.6khz (allowing for a decent 16khz frequency response plus below-nyquist filter rolloff), 14 bit with a slight hint of A/mu intensity scaling (for perceptually better than 16 bit linear quality), using 3 "real" channels, two being discrete stereo with the third one having simpler and more robust (and therefore better sounding) matrix encoding to drive the third and fourth speaker outputs? Who knows... Remember that Pro-Logic actually had left, right, rear and centre as its 4 channels, not FL, FR, RL, RR... so it might not have been entirely suitable for "real" quadraphonic, though good enough for vocal pop/rock music with a bit of extra ambience ;) 193.63.174.10 (talk) 14:27, 18 November 2009 (UTC)


 * KeithWinstein 22:27, 26 March 2007 (UTC)

Merge article
I think this page should be merged in Red Book (CD standard) — Preceding unsigned comment added by 93.56.118.181 (talk) 19:38, 3 September 2011 (UTC)

How to create a working 4 ch CDDA
For 4 channel audio it has to be 22/16 instead of 44.1/16. There is no standard software I know of to create these. This is what worked for my tests on many players, but since the Redbook never specified this it is hard to say if it would work on all players. I have never read any documention that actually specified this, actually found it on some forum and tried it and it worked.

I created a 4 channel .wav at 22/16 and then a cue sheet w/ the 4 ch flag set and used CDRWIN to write it (Nero and others will convert it to 44.1 so only CDRWIN worked). Another tricj was to write it as binary instead of audio as CDRWIN didn't like audio - but as I said it works. — Preceding unsigned comment added by Tyros1972 (talk • contribs) 19:52, 26 December 2011 (UTC)

Hi Tyros1972, could you please go into more detail about how you tested this on the Emerson NR303TT and verified that it played four independent channels of 16-bit, 22.05 kHz audio? I'm surprised by the idea that this (not-particularly-notable) CD player responds to the "broadcasting use" / ex-"four-channel" bit in this manner -- how do you even get four independent outputs out of the device? If you just listened to the L and R channels and it sounded reasonable, I don't think that's enough. After all, ANY CD player will produce reasonable-sounding output if you sample four channels at 22.05 kHz and write them to a CD-R in 16 bits at C1C2C3C4 order. The left channel will get C1C3C1C3..., with each original channel sampled at 22.05 kHz, and the right channel will get C2C4C2C4..., again sampling the original channels at 22.05 kHz. So from a dumb CD player that ignores the "broadcasting" / "four-channel" bit, you will hear a mixture of C1 and C3 out of the left channel and C2 and C4 out of the right channel. If that's what you heard from your Emerson, that doesn't mean it's respecting the bit or decoding four channels of audio. That's what any CD player will do. KeithWinstein (talk) 08:29, 9 January 2012 (UTC)

Hi Keith,

No the Emerson and the Magnavox DVD players only had ST output, so can't verify the 4 ch. Why would they play at 22.05? If I stick a 82000/16 PCM in there the player will play it at 44100 (be very slow), so if the player is playing the 2205 at the right speed doesn't that prove it?

I do have a 5.1 setup w/ a DVD player so I can test to see if it actually plays 4 independent channels and that should say something. — Preceding unsigned comment added by Tyros1972 (talk • contribs) 13:52, 9 January 2012 (UTC)

EDIT: ah I got what you're saying since it's 2205(x2) for each channel, yep so it is playing 44100. Yep you are probably right but I will test the 5.1 player for the heck of it and see if it works. I don't have high hopes :) — Preceding unsigned comment added by Tyros1972 (talk • contribs) 14:01, 9 January 2012 (UTC)