Article
Aircraft Engine Speed Estimation from Flyover Audio Recordings: an Open Access Solution? (en)
* Presenting author
Abstract:
One of the main challenges in auralization is generating high quality sound source synthesis signals and models. The quality of these models is often dependent on faithful real-world source data. This information can be particularly difficult to access for aircraft sources: cockpit data is typically protected information and not publicly available. The speed of the engines is a particularly important parameter, which is not easily accessible. This work describes a method to estimate the engine speeds of selected aircraft trajectories from audio recordings of flyovers. We apply time-domain processing and convolution-based filters over the recordings’ spectra, in order to find the blade-passing frequency (BPF) of the aircraft engines. Aircraft model and respective engine specifications are openly available. By relating such specifications with the BPF vector, an engine fan speed estimation can be generated over time. We make use of open-access information only and discuss the limitations of our approach. The validation of this method remains a challenge due to lack of open access engine speed data.