Stem (audio)

A stem in audio refers to a discrete, self-contained group of audio tracks that are mixed together to form a subcomponent of a larger audio project, such as a song or soundtrack. Stems are often created as a deliverable for mixing, mastering, remixing, or broadcast purposes.

Unlike individual tracks, which represent single instruments or vocals, stems typically represent broader categories of sound. For example, a song might have stems for drums, bass, vocals, and instruments (guitars, keyboards, etc.). Within a film soundtrack, stems might include dialogue, music, and sound effects.

The key characteristic of a stem is that it is a premixed subset of the final audio. This provides flexibility to the recipient, allowing them to adjust the balance of these broad elements without needing to manipulate each individual track. A mastering engineer, for instance, might subtly adjust the level of the vocal stem relative to the instrumental stem to achieve a better overall balance. A remixer might use the instrumental stem of a song to create a new version with different drums and vocals.

Stems are typically delivered as high-quality audio files, often in WAV or AIFF format, and at the same sample rate and bit depth of the original project. The stem creation process usually involves carefully balancing and processing the individual tracks within the stem before rendering it as a single file. Good stem creation practices aim to avoid introducing unwanted artifacts or imbalances in the final product.