Last year, SoundCloud upgraded its AAC encoder for the first time in over a decade. The new libfdk_aac delivers higher perceptual audio quality across every metric we track. Yet, on a spectrogram, a hard shelf appears at 17 k Hz, where energy used to extend to 20 k Hz.

Lossless files (WAV, FLAC) keep every sample intact—what goes in comes out bit for bit. Lossy compression (AAC) forces the encoder to discard data to hit a target file size. The question is what it chooses to lose.

When SoundCloud compresses to growing standard, the encoder receives a fixed number of bits per chunk. It allocates more bits to bands where the ear is most sensitive. Dropping the top frequencies lets the rest of the signal get sharper—a trade‑off that favors audible quality over a pristine spectrum.

At 256 kbps there are nearly enough bits for everything, so you’d expect full bandwidth. The encoder still rolls off around 17 k Hz, a small but measurable improvement. Fraunhofer validated these cutoffs with listening tests; trained listeners couldn’t reliably detect the loss.