During the past two decades, the capacity of the personal computer to capture, store, deliver, and play sound has revolutionized audio services in libraries. The computer audio technology of the 1980s allowed librarians to begin transferring sound from deteriorating or obsolete media to more stable formats, and more recently, advances in network speed, audio compression, and streaming technology have offered libraries opportunities to extend access to their sound recording collections in ways that were barely imaginable a decade ago. Users are now able to listen to recordings remotely, they can listen at different points of the same recording simultaneously, and they have easy access to recordings that were once restricted because of their condition or format. In describing these new collections of digitized sound, music librarians have used a number of terms, the most common of which is music library, a natural extension of library, widely used to describe digitization projects in libraries. The libraries comprising the Digital Library Federation have arrived at the following definition of Organizations that provide the resources, including the specialized staff, to select, structure, offer intellectual access to, interpret, distribute, preserve the integrity of, and ensure the persistence over time of collections of digital works so that they are readily and economically available for use by a defined community or set of communities. (1) In 2000, Amanda Maple and Tona Henderson described the issues that must be confronted by a librarian planning a digital music library project, and explained the decisions made for their own project at Pennsylvania State University. (2) The issues fall into three broad categories: infrastructure (including the selection of hardware, software, streaming technology, and method of access); collections (including decisions on what to digitize and why, and related questions of copyright); and staffing (including who does what, who employs them, how the work is funded, and who provides training and public service). In 1999, when Maple and Henderson wrote their article, no more than fifteen libraries were digitizing audio, but since then, dozens of libraries have mounted digitization projects, and the number continues to grow. Now that a substantial base of digital music library projects is in place, we can assess how librarians have dealt with the issues identified by Maple and Henderson. In order to collect information on the projects, I distributed a note on MLA-L (3) on 10 January 2002 asking librarians engaged in digital audio projects to participate in a survey. I sent questionnaires to the fifty respondents, and thirty-five were completed and returned. In preparation for the present article, I sent a note to these initial thirty-five respondents in July 2002 asking for updated information. At the same time, I issued a second call on MLA-L and received eight more responses, yielding a total of forty-three responses. The respondents represent forty-two libraries: thirty-seven university libraries and five college li braries (see table 1). PRESERVATION AND ACCESS The principal work of the projects undertaken by these forty-two libraries falls into two broad categories: reformatting rare recordings or recordings in obsolete formats, and making high-demand recordings more easily available to the public--in other words, projects that provide preservation and access, activities that have been closely linked during the past two decades. Historically, preservation and access have been seen in opposition: materials are preserved by restricting access to them, and in turn, providing access to materials endangers their preservation. During the last quarter of the twentieth century, initiatives to reformat materials--by transferring the sound from a wax cylinder to reel-to-reel tape, for example--allowed librarians to preserve materials by shifting access away from the original to a surrogate. …