The aim of this paper is learning the shortest possible lengths of audio fades and crossfades that are not audible as audio artifacts. The determined lengths can be utilized in adaptive streaming scenarios during pauses and content switches. Subjective and objective tests were performed, utilizing speech and music signals with various fade-out and fade-in lengths. Subjective evaluation was performed by critical listening tests where listeners were asked to grade the quality of the fade-out or fade-in and listen for unwanted audio artifacts. Basing on the subjective test results, optimal ranges of fade-out and fade-in times were selected—50 to 100 ms for fade-in, and 100 to 200 ms for fade-out. Objective tests were conducted using optimal times chosen by the listening tests. The results confirm that the selected ranges of fade-in and fade-out lengths do not introduce significant harmonic distortion and noise into the signal.
(See document for exact affiliation information.)
https://aes2.org/publications/elibrary-page/?id=19517
(125KB)
Click to purchase paper as a non-member or login as an AES member. If your company or school subscribes to the E-Library then switch to the institutional version. If you are not an AES member Join the AES. If you need to check your member status, login to the Member Portal.
