Dynamic Spectral Analysis of Jagged Mechanical Signatures of a Brittle Puffed Snack

The instrumental evaluation of crispiness and crunchiness of dry and wet cellular foods is challenging. Available texture analysis methods do not always reliably predict sensory analysis results. Temporal sensory integration is suspected to be a key factor in the perception of crispiness and crunchiness. Thus, short‐time Fourier transform, continuous wavelet transform and Hilbert–Huang transform are proposed and applied as dynamic alternatives for analyzing multifracture events. The resulting time–frequency–magnitude spectra graphically show the degree of similarity between the samples. These representations contribute to an understanding of the dynamics of airy foods' jagged mechanical signatures, as demonstrated on corn starch extrudates. In most cases, they finally permit the recognition or discrimination of similarities and differences in the degree of brittleness, corresponding to a specific production process and water content. The analytical techniques should help to determine relevant and objective characteristics that correlate with sensory studies. PRACTICAL APPLICATIONS: Short‐time Fourier transform, continuous wavelet transform and Hilbert–Huang transform help to understand the physical processes and the temporal evolution of the breakage behavior of foods. They are powerful tools for analyzing jagged mechanical and acoustic food signatures. Each method offers a different perspective, thereby enabling the exploration of unforeseen characteristics that could lead to better predictions of sensory‐felt crispiness and crunchiness. A food's dynamic fingerprint helps in recognition of similar products that belong to the same family, despite natural individuality, and aids the discrimination between different products. After relating classified food samples to their structure and consumer preferences, food structure design and quality control can be improved. The methods can be applied to multidisciplinary food texture studies that examine air‐conducted crushing sounds, bone‐conducted vibrations, dampening effects of muscles and fatty tissues, or chewing muscle and neural activities. This manuscript intends to make modern techniques of signal analysis more accessible to food scientists.