Home Page Company Clients Partners Reports CallWeb SampleCalc EPlanCalc French

Doubt is not a pleasant condition, but certainty is an absurd one.

APRIL 4, 2003

Is music in compressed file format really of lower sound quality?

Thanks to the Canadian Private Copying Collective for allowing us to distribute our report on the sound quality of compressed music files. This report is available on our Web site. Here is the report summary.

This study is based on the results of a Web survey conducted in December 2002. The respondents were 258 Canadians age 12 to 49 who had a multimedia computer with Internet access in their homes at the time of the survey. Each respondent compared the sound quality of two versions of the same selections of music: the original version and a version that had been compressed using the MP3 algorithm. The following conclusions were drawn from the responses.

In 14% of the cases, the respondents correctly identified the original track of music as being of better quality.

The respondents who did perceive a difference in the sound quality of the tracks they listed to assigned it a monetary value of 20%.


This study is based on data gathered from 258 respondents in an on-line survey conducted from December 6 to 18, 2002. These respondents were recruited by Echo Sondage telephone, on December 3 and 4, from a list of the respondents to Circum Network Inc.'s Music Monitor survey who had reported that they had a multimedia computer and Internet access in their homes and had agreed to be contacted again. Only persons who were age 12 to 49 at the time they were originally interviewed and whose interviews were completed between January and September 2002 were included in this study.

In the recruitment phase of this study, the refusal rate and the response rate, as calculated by industry- standard methods, were 23% and 44%, respectively. In the data-collection phase, the response rate was 59%.

A random sample of the size used in this study produces a margin of sampling error no greater than ± 4.7 percentage points for proportion estimates for the entire sample, taking the effects associated with the sampling plan into account. The margins of sampling error for sub-groups within the sample are larger.

The estimates produced from these data were adjusted for gender, age, region of residence, and mother tongue, so as to match the distributions calculated in the Music Monitor survey for the population segment used in this study.

The survey questionnaire was designed to meet two research objectives. It comprised six sections, and the number of questions asked varied with the responses provided. The minimum was 6 questions, and the maximum was 23. A music CD prepared especially for this study was sent to each respondent's home. This CD contained five selections of five different genres of music, each recorded in two different formats. The respondents were asked to listen to both versions of each selection on their own audio equipment and to indicate whether either one had better sound quality than the other.

 Return to the home page
 List all news

To reach us:

General address : service@circum.com
Benoît Gauthier : gauthier@circum.com, @BGauthierCEEQ
Tel. : +1 819 775-2620, Fax : (no more fax; sorry)
238 Fleming Road, Cantley, Québec J8V 3B4