Study Shows "Non-Audiophiles" Can't Hear Difference Between 64 and 256 kbps MP3 Files

[Sonnie]

According to a study released by Cognitive Daily, the average listener struggles to tell the difference in audio quality between the lowest-resolution MP3 files at 64 kbps and the so-called high-resolution 256 kbps files that are sold in limited quantities as “high definition” from the likes of Apple’s iTunes and Amazon’s new download site.

The study looked at three different resolutions of audio, ranging from Good (64 kbps) to Better (128 kbps) and Best (256 kbps), for two selections of music, including a Santana track and a classical piece by composer Aaron Copeland. The study also asked the group to self-define themselves on a scale of one to nine in terms of how much they consider themselves to be audiophiles (nine being the most critical of listeners and one being a listener who isn’t concerned with audio quality at all). The results show that self-proclaimed audiophiles could better hear the differences between the three grades of audio better than non-audiophile listeners. Fascinatingly, listeners with decades of musical training on various instruments showed no specific gift for hearing the difference between the levels of these low-resolution MP3 files.

Where one could find fault with the study is in its methodology. A compact disc, the most successful audio format in the history of the history of the world, packs 16-bit resolution and 1411 kbps data rate. This is many times higher than that of the tested MP3s that are currently being pushed on the market today. Much higher-resolution downloads from less mainstream music websites like Music Giants and iTrax start at 1100 kbps and increase to over 9000 kbps, thus getting to the resolutions that audiophiles heard on truly high-resolution formats like SACD and DVD-Audio. These vastly higher-resolution files have many times more data, allowing for a musical experience that is much closer to what is captured on the master tape simply because of the higher resolution, data rate and sampling rate. Complicated audio events like a cymbal crash require tremendous amounts of data for an audio system to reproduce it in the same way the human ear would hear the sound live. Lower data rate and resolution MP3 files often are criticized for sounding “bright” or “shrill” by audio professionals and mainstream listeners alike when compared to true HD music formats that extend far past the performance of low-resolution, “lossy” (meaning compressed) MP3 files. One noted record producer suggested the difference between the 256 kbps MP3 format in this study is like testing the 0-60 on a Prius vs. a Camry, when true HD resolutions are more like testing a Ferrari. The amount of data and resolution is many times higher and more able to bridge the gap between the physical limitations of audio playback and an actual musical event.

The sad commentary here is that the four major record labels are unwilling to market their content at any level of real high definition. Simply put, one-fourth the resolution of a 25-year-old compact disc (256 kbps) format is not “high-resolution” anything, despite what Apple says. The majors fought over the differences between SACD and DVD-Audio, leaving discerning consumers feeling ambivalent about investing in either format because of a lack of titles, complications in system set-up, a lack of video content and overall cost. Yet video games costing $60 to $90 per title sell in volumes that are five times higher than today’s best-selling CDs. High-definition movies on HD DVD and Blu-ray sell in increasingly strong volumes, while over 3,000,000 HDTV sets are sold per month in the United States. The ship has likely sailed for SACD and DVD-Audio, but both the new HD DVD and Blu-ray formats allow huge storage capacity for an audio and video experience that can expand the “album” concept far beyond one or two low-resolution files. But the major labels, as their sales spiral down the toilet bowl, simply fail to release their music in any compelling format that improves on the value proposition or audio quality, thus turning to very low-resolution downloads because they represent the lowest-hanging fruit for a business that simply cannot market its way out of a paper bag. The art of emotionally charged high-resolution music has been left to smaller players, but if you want to hear what the potential of a real download sounds like, listen to the Super HD downloads from Music Giants. There is no comparison between the power of what they are doing and a mere 256 kbps file.

Source: AVRev News

[mrstampe]

Before you read too much into this, chech out the original article -- link.

This Cognative Daily "journal" is a non-peer reviewed, blog-style web site. The source article for the avreview piece is really no more than a bubblegum quiz. Here are just a few of the article's weaknesses:
1) Subject were self selected from website viewers and self rated on the audiophile scale without objective scoring or frames of reference for attempting to differentiate the two groups. How many Bose aficionados have you known to claim audiophile status? Or, examples of Mr. Moneybucks who buys a $50k system thinking that will buy him a discriminating ear or audiophile status?
2) Subjects were asked to listen to a Santana track on their own headphones, and a Copeland classical clip over their computer speakers!! No controlled environment or speaker standards for reproduction.
3) All conclusions were made based on two 30sec clips. No respectable study would make claims based on such limited data of so little quality.

Jerry Del Colliano of avrev.com, makes some interested observations, but don't make too much of the pseudo-study he references.

[yourgrandma]

I could see not being able to tell the differance between 256 and 128 under those conditions, but 64!? Thats like what cellphone ringers use. It is very frustrating to see this going on. So called studies like this will only encourage this trend towards lower and lower quality reproduction. I cant claim to have a really good ear, especially since my insident with my home made black powder cannon, but I cringe any time I hear a file with anything below 256kbps resolution. And thats with no referance copy to compare to, I havent had the guts to actually put myself to the test on this, because Im slill trying to convince myself i have decent hearing.

[phaseshift]

As another member indicated earlier- I would not put a lot of weight in the findings there. Not that I do not believe that they are on the right track, but the way that the test is administered appears to lack appropriate controls and as I far as I can tell, no control group (where nothing changed etc...)

Another point is that the speaker being undefined may leave a lot on the table in terms of how faithfully the speaker reproduces the input material. While you can do a relative test in a manner something like they have done in the study, ideally, you would want to have the reviewers listen thought a set of high quality / high fidelity loudspeakers (or headphones- whichever you prefer). Edit- In a controled environment


In a controlled environment with descent speakers, you can pretty easily pick out the difference between the lower bit rate formats and the higher ones. I bet that if you had all the people in the same room, doing an instantly switched test, the results would show the average listeners to be significantly better than they are now. My bet is that the experienced listeners are more aware of the details to look for and to some degree, have a bit sharper auditory memory skill set. In other words, they may not have better ears, but they are more careful listeners thus better equipped to distinguish between the differences they hear.

[tonyvdb]

I do find it very hard to believe that one cold not tell the difference between a 64kbs file and a 256
I can see if the equipment used to do the test was a cheap system but even my kids can hear the difference. I wonder why people don't use variable bit rates this gives you better quality but still a smaller file size.

[avaserfi]

I too find this very hard to believe and fall inline with everyone who says there was little control to this study. I recently completed a study on preference between dynamic compression and uncompressed version of the same song which has less distortion and change than the drastic compression spoken of here. In the end I got very strong results showing that overall the uncompressed version was liked more - only one person couldn't tell the difference.

Perhaps this was because I was using quality equipment in a more controlled environment especially comparing to this report.

More can be seen here: http://www.hometheatershack.com/foru...07-post11.html

[drf]

This article has missed the point.

People can tell the difference between 64 and 256. Why does it appear that only audiophiles can tell the difference? because most audiophiles know what to listen for so they know which is which. It can be likened to a mechanic listening to a car while tuning it up, your average joe isn't going to know when the car is running better but if you tell them what to listen for, even though they know nothing about cars, they will start to hear the difference.

Another thing this study seems to fail to comprehend is that not only are there different bitrates but there are also different qualities of compression. A poor codec or quality 256K rip has the potential to sound worse than a good 64K rip. Unfortunately they are not comparing apples to apples and for the most part once you are told what to listen for, you will realise that you don't need to be an audiophile to hear the difference.

[trust7]

Does not seem true at all.