In recent years there has been an increase in the use of portable digital recording devices such as, smart phones, tablets, dictaphones, and other portable hand-held recorders for making informal or in-situ recordings. Often it is not possible to connect a clocking signal to these devices as such recordings are affected by the deviations of the actual clocking rate of the device from the expected rate. This variation causes problems in the synchronization of signals from multiple recording devices and can prevent the use of some signal processing algorithms. This paper presents a novel methodology for determining the actual clock rate of digital recording devices based upon optimizing the correlation between a recording and a ground truth signal with varying degrees of temporal stretching. The paper further discusses the effects of sample frequency variation on typical applications. The sampling rates of a range of commonly used mobile audio recording devices was found to deviate from the nominal 48 kHz, with a standard deviation of 0.8172 Hz. The standard deviation of sampling rates for a single device type, used for long term logging of bio-acoustic signals, was found to be 0.1983 Hz (at a sampling rate of 48 kHz).
(See document for exact affiliation information.)
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.
