The linewidth of longitudinal acoustic waves in densified silica glass is obtained by inelastic x-ray scattering. It increases with a high power α of the frequency up to a crossover where the waves experience strong scattering. We find that α is at least 4, and probably larger. Resonance and hybridization of acoustic waves with the boson-peak modes seems to be a more likely explanation for these findings than Rayleigh scattering from disorder. PACS numbers: 63.50+x, 78.35+c, 78.70.Ck A controversy regarding the fate of very high frequency acoustic-like excitations in glasses arose in recent years [1,2]. Acoustic waves propagate in glasses up to rather high frequencies, for example up to at least Ω/2π ≃ 0.4 THz in vitreous silica, v-SiO 2 [3,4]. However, silica shows a pronounced "plateau" in the temperature (T ) dependence of its thermal conductivity κ(T ) [5]. This feature was explained by postulating the existence of a rapid crossover of acoustic waves into a regime where they experience strong scattering [6]. In v-SiO 2 , this crossover is expected at a frequency Ω co /2π ≃ 1 THz [7]. For the observed horizontal plateau to occur, besides strong scattering of thermal phonons, phonons of lower Ω must also experience a sufficiently strong T -independent scattering [8]. Several models predict that the inhomogeneous width of acoustic waves can increase with a high power of Ω, typically in Ω 4 [9, 10], leading to this rapid crossover. However, the origin of the high power remained debated, whether Rayleigh scattering from structural defects [6], or resonance with local modes [10,11,12]. Rayleigh scattering seems generally too weak in materials as homogeneous as dense glasses to be able to account for the relatively low Ω co suggested by the position of the plateau [6,8]. The local modes are seen in the "boson peak" (BP) and produce an excess over the Debye specific heat [13]. Successful models based on resonance or on soft potentials were developed to describe the observed properties, e.g. in [10,14,15].It might seem, judging from some recent literature [2], that the above picture should be revised after it became possible to observe in glasses the spectrum of longitudinal acoustic (LA) waves at THz frequencies with x-ray Brillouin scattering (BS) [16]. This new experimental tool is important here, as it potentially allows to check some of the previous conjectures. It led to many publications claiming to prove that sound propagates at frequencies much above the early expectations for Ω co , e.g. [17,18]. In this respect, it must be emphasized that inelastic x-ray scattering (IXS) still is a difficult spectroscopy with severe limitations on resolution and intensity. Firstly, it is practically not possible with current instruments to investigate scattering vectors Q below ∼ 1 nm −1 , a value which happens to coincide with the expected wave vector q at crossover, q co , in v-SiO 2 [7] and in many other glasses. Secondly, the narrowest instrumental profile allowing for sufficient intensity [19] still has an energy w...