Headroom and Clipping
Digital full scale is a hard wall; headroom is what lets transients and later processing fit.
Related
The problem
Digital audio tops out at 0 dBFS, and anything beyond is clipped, with the tops of the waveform sliced flat and harsh high frequencies left behind. Every plug-in, extra track and automation move can raise the level, so both recording and mixing need spare room. Headroom is not wasted dynamics; it is space for every step that follows.
How it works
Fixed-point formats cut hard at 0 dBFS and discard what exceeds it. Floating point carries enormous internal range and does not clip immediately above zero, but it still clips when handed to a fixed-point device or encoder. Clipping generates high harmonics that sound like breakup, and brief heavy clipping is sometimes used as a loudness trick at the cost of distortion. Summing tracks does not add peaks arithmetically, but when phases align they can stack very high, which is why buses need more room.
Settings and practice
Record with peaks around minus 12 to minus 6 dBFS to cover performance surprises. Mix with bus peaks near minus 6 dB and leave 3 to 6 dB before mastering. Work in floating point and avoid repeated fixed-point conversions in the chain. Check level before and after every plug-in, since some add obvious gain. Use a bus limiter to catch transients, not to push the whole mix louder. When ear training, separate clipping from compression: one slices the waveform, the other narrows the dynamics.
What to listen for
On the hip-hop kit, whether kick and snare together break up; on the live ballade, whether heavy chords plus applause turn harsh; on the string crescendo, whether the layers stay separate. Genuine clipping does not sound louder, only dirtier.
Listen in the library
All tracks from midicn-lib — inline badge is the license tier (C1 commercial / C2 non-commercial / C3 study only).
Sources
- 浮点与定点满刻度的差别，依数字音频通识核对，未抄原文
- 增益结构的分级做法，依录音工程公开资料核对，未抄文本
Used for fact-checking only; all prose is original.