October 2008


Does MP4 = CD Quality?
Article By Barry Fox (England)
From BAS September 2008

  We can now expect more advertisements claiming that compressed audio at 128kbps is "CD Quality."

The landmark ruling comes from the UK's Advertising Standards Authority, which considers complaints about misleading advertisements. The CD Quality judgment followed a complaint made against Nokia (www.ASA.org.uk...).

A series of posters, for the Nokia 5300 XpressMusic mobile phone, claimed "up to 1500 tracks, CD-quality sound." A consumer challenged the "CD quality" claim. Nokia argued that the XpressMusic supported Advanced Audio Coding (AAC) and the phone comes with software that encodes CD tracks using MPEG-4 AAC Low Complexity (LC) at 160kbps.

In support, Nokia cited an ISO (International Standards Organization) report "on the MPEG-2 AAC Stereo Verification Tests." Listening tests had been carried out to determine whether listeners could distinguish between original and compressed samples and Nokia argued that the results showed that MPEG-2 AAC LC at 128kbps gave sound quality which sounded as good as the original CD, even to expert listeners. Listening panels had been unable to distinguish between compressed AAC files encoded at 128kbps and CD sound.

Nokia argued that MPEG-4 AAC LC has exactly the same performance as MPEG-2 AAC LC.

The ASA also looked at the results of controlled listening tests carried out by the Communications Research Centre, which compared audio codecs with CD sound, and showed that AAC encoding at 128kbps was indistinguishable from CD sound.

This satisfied the ASA. "We considered the tests proved that most listeners were unable to distinguish between compressed AAC files encoded at 128kbps and CD sound," says the ASA. "We concluded that Nokia had substantiated the claim ‘CD quality sound' and it was unlikely to mislead."

So, with this precedent set, advertisers will no longer need to use fudges like "near-CD quality" or "digital quality." We can expect a flood of ads that push the boundaries further, with lower bit rate "CD quality." But how closely did the ASA look at the ISO report it relies on ( www.ASA.org.uk...)?

The report dates back to February 1998 and was written by David Meares (then with the BBC R&D Department at Kingswood Warren); Kaoru Watanabe of NHK in Tokyo; and Eric Scheirer, from MIT Media Labs in the US.

David Meares is now a broadcast engineering consultant. He says: "The ASA ruling makes interesting reading, both for what it says and for what it does not say. The reference (source) recordings were taken from CDs and transferred to DAT for the tests. So all source items were 16-bit linear, with the exception of the harpsichord item which was judged to be too loud and was attenuated by 6dB and re-quantized."

"As far as the detailed results for AAC LC at 128kbps are concerned, overall the results are statistically distinguishable from the source material, but only just so. This is because two items, harpsichord and pitch pipe, were difficult for the codec."

"What is not covered, either by the ASA ruling or the ISO report, is the possibility of adverse results where upstream compression has taken place. If there are any residual coding footprints on the source material, which by themselves are inaudible, it can make the job of the final delivery codec more difficult and the overall result can be judged to be impaired. If the final delivery mechanism is CD, this problem is avoided." So when the ASA blesses AAC at 128kbps as CD quality, it is really blessing AAC 128 as CD quality only if it has been coded directly from a 16-bit linear original such as CD.

It's very likely that some companies which now claim CD quality from 128 AAC will be sourcing from material that has already been compressed; for instance, for storing on a server, or sending down a line or over the air. Any anomalies from one compression codec will then daisy chain with any anomalies from the other coder — much as codecs have difficulty compressing hissy audio from disc or tape, because they waste bits trying to compress the hiss accurately.

It looks as though the ASA has just robbed hifi of the audio benchmark we have enjoyed since CD was launched 25 years ago. No-play Station?

The instruction manual that comes with the PS3 game machine (which is a terrific value Blu-ray, CD, DVD player that also accesses the Internet) carries several warnings.

"Recently some record companies have started marketing music discs encoded with copy protection technologies. Some of these discs do not conform to the CD standard and may not be playable on this system." It was Sony Music that used XCP copy protection on music CDs and exposed PCs to viruses while trying to stop them from copying music. Replacing the discs and paying penalties has cost Sony tens of millions — which the company can ill afford because it is heavily subsidizing the PS3.

"A Dual Disc is a dual-sided disc that combines a DVD side and an audio side. Note that the audio side is not guaranteed to play because this type of disc does not conform to the specifications required for an audio CD."

Sony Music makes and sells Dual Discs.
Recent Sony DVD releases including Casino Royale, The Pursuit of Happyness, and Stranger Than Fiction would not play on some DVD players. Sony blamed its own ARccOS copy-protection system. "Recently, an update that was installed on approximately 20 titles was found to cause an incompatibility issue with a very small number of DVD players," the company said in a statement. "Since then, the ARccOS system has once again been updated, and there are no longer any playability problems."

Players were freezing or shutting off in the menu section of recentreleases from the studio, so Sony is providing replacement discs to customers who have had problems.

Some of the players affected were made by Sony.
Recent patents filed by Sony's DADC pressing plant in Austria reveal some of the ways Sony's music and movie divisions are disregarding the CD and DVD standards which Sony's hardware division helped produce. All the patented ideas are based on the same basic principle: a DVD player uses navigation codes on the disc to play scenes in the correct order, and disc copying software cuts corners by assuming that the scenes will play in natural order. So DADC will put scenes in the wrong order, or add obstacles like corrupt data, and rely on the navigation codes to steer the player around the obstacles and into playing scenes in the correct order. So a simple copy either stops playing halfway through or skips scenes or plays the movie with scenes in the wrong order — for instance, so that the secret of a whodunit comes up early on. Some of the obstacles may be messages which tell the user that they shouldn't be watching an illegal copy, or punishes them with advertisements for products. The opportunities for problems, such as players playing original disc scenes in the wrong order and infuriating innocent viewers, seem to be endless.

Another patent from DADC, this time from the US plant in Terre Haute, IN, seems even more risky to put on the open market. In this case Sony has joined forces with US company Veriloc to develop a double whammy way of preventing disc theft from shops and disc copying at home. A CD or DVD is pressed with one part of the disc covered by a material which permanently changes color and transparency when hit with a laser beam of carefully tuned frequency. So until the disc has been legitimately purchased, and scanned with the correct light at the checkout, it will not play.

Other parts of the disc are coated with material which changes transparency when exposed to too much light from a player laser. This can happen when the disc is spun at high speed for copying. The change in transparency can hinder readout or trigger software which has been automatically loaded into a PC when you play the disc. So in theory the disc plays normally but stops reading when copied in a PC. It can also stop playing if played too many times. Once again there seems to be an awful lot that could go wrong once discs escape the lab and reach the consumer.

You can read more in US patent filings 20060239155 and 20070036356 and 20070050585; links:
Link 1,

Enjoy the Music.com highly encourages our readers to join the Boston Audio Society by clicking here).

This article is copyrighted © by the author or the Boston Audio Society.
It is posted on Enjoy the Music.com with their permission,
and with all rights reserved.