Case Study mp3 - Conclusion

Case Study mp3 Soundtest / Home

Conclusion

It is hardly possible to perform the ultimate mp3 test. Interesting aspects such as the performance of hardware players or a comparison to minidisks, for example, were not included here. It should be noted that in any case, a % rating in listening tests can never be representative. The relatively poor performance of many mp3 recordings could be revesed using worse HiFi equipment or a different music selection. It was surprising to discover that mp3 could be recognized consistently, especially after a little practice, by the presence of hissing or noise whenever pulses occurred. The term "near CD-quality" has lately been associated with mp3. This appears a bit unduly favorable to a number of test listeners, especially since there are actually are near CD-quality audio devices such as DAT. The different mp3 encoders produce different results [3 There are differences, however all of them are far removed from CD quality]. Increasing the data rate to 256kb/s does improve results [4 The blindfold test produced significantly fewer correct calls]. But this is often a theoretical exercise since the 128kb/s variation used in our test is the de facto standard. In the best case, one may not know until after downloading which encoder was used on a recording downloaded from the Internet. We were not successful in our search for really excellent mp3 recordings with which high-end speakers or even amplifiers could be tested with a clear conscience. For a large range of music, though, mp3 produces acceptable results and in tandem with the high compression rate, this proves that its developers did an excellent job. There is potential for improvement and this can surely be realized through increased computation. It would be very important to ensure a certain quality level. Good encoders already give an idea of what is possible. Unfortunately, the development of new formats (e.g. . Real Audio 8, Windows Media Audio or Advanced Audio Coding (AAC), [5]) appear to be going in the direction of faster coding or higher compression while maintaining acceptable quality (e.g. for UMTS mobiles or as part of video transmission) rather than in the direction of HiFi. The good results obtained for analog recording clearly show that the new, coded data formats will have a tough time, aside from acceptance problems and an unobtainable 100% copy protection. And even PC non-experts have a shot. After all, analog copies cannot be prevented, and if the losses are less than for compression!

[1] http://www.trail-n-error.de/mp3theorie1.htm
[2] http://hellgate.hs-bremen.de/mp3/artikel.htm
[3] http://www-stud.fht-esslingen.de/~alkoit00/mp3enc/
[4] http://www.raum.com/mpeg/reviews_quality.html
[5] C. Meyer, ct´ Heft 6/00, S. 92ff
[6] ct´ Heft 22/00
[7] http://www.mp3-tech.org

Tips an Audio, mp3 and PC

• mp3 is not always at fault if the tone quality is less than desired. If possible, look at the originals using a wave editor, or listen to them very closely.
• For high-end applications, use a higher data rate. 192kb/s are often a very good choice. Some encoders don’t deliver better quality even for improved data rates..
• Compare several encoders. The original FHG encoder is said to be the best (although there is fervent opposition). Unfortunately, it isn’t free. The most frequently used encoder in CDex is not the best, but definitely not the worst. The LAME-encoder is not only our [3] highest recommendation. As a DOS program, it is well suited for batch processing of multiple files.
• Use music containing lots of pulses for listening tests. These make it easier to detect mp3 errors. The two piano pieces and „Without You" were particularly critical in our opinion.
• When making analog recordings, be sure and pull the antenna plug and don’t use too long cables. Set the maximum level carefully to avoid overdrive.
• The quality of PC soundcards can vary widely. We got lucky with our attractively priced PCI card. Many old ISA cards or on-board cards are much worse. To play it safe, be sure and read the current product reviews, as the market changes very quickly.

Comments by listeners

- One can get fairly euphoric when one discovers a long-lost favorite tune in mp3 format on the Net. The tone quality is usually acceptable for an older tune. But these test results make me wonder if CD burners don’t present the far greater danger to the music industry.
- Mp3 isn’t perfect, but most people won’t notice the difference. What do you want for nothing? New compression methods will only establish themselves if they also are available for free.
- The future belongs to compression algorithms. UMTS, DAB and digital TV will further expand the application range.
- Funny that the tone quality of the original has relatively little influence on possible mp3 problems. The errors were not easier to detect in classical music.
- This is junk. It sounds funny. Something’s wrong.

To the Beginning

To Top - To Homepage - German Version