Dynamic Range Compression (Night Mode)

What dynamic range compression is

Dynamic range is the distance between the loudest and quietest sounds in a mix. Dolby's Hybrik documentation describes dynamic range control as boosting low-level signals, leaving mid-level signals untouched and compressing high-level signals, which reduces the overall range. In Dolby codecs it is carried as gain words in the bitstream that "do not alter the source audio," and higher-end decoders can defeat it completely or the consumer can apply it selectively, which is what a "late night mode" does. Sound On Sound puts the use case plainly: in a quiet room the full range can be used, but late at night, or with sleeping neighbors, a greatly restricted range is more appropriate, while the producer decides how the DRC works through the metadata.

How the metadata works

ATSC A/52, the Dolby Digital standard, defines two control signals. The first is dialnorm, a 5-bit value from 1 to 31 read as -1 dB to -31 dB below digital full scale, which tells the system where average dialogue sits. The second, dynrng, gives a gain change of plus or minus 24 dB, updated per audio block of 5.3 ms in steps as fine as 0.25 dB. A third, compr, has twice the range (plus or minus 48 dB) at half the resolution (0.5 dB) and updates once per 32 ms syncframe; it exists for links with very limited headroom, such as an RF modulator feeding a low-cost TV, or hotel and airline playback. A/52 says a decoder shall by default apply dynrng, but may let the listener ignore it or use only a fraction of boosts versus cuts.

Dolby's authoring tools define how the gain words are generated. Hybrik lists five profiles (film_standard, film_light, music_standard, music_light and speech), and Sound On Sound adds a sixth, None. The default Film Standard profile boosts up to 6 dB below -43 dB, boosts at 2:1 from -43 to -31 dB, leaves a 5 dB null band from -31 to -26 dB, cuts at 2:1 from -26 to -16 dB and at 20:1 from -16 to +4 dB. Speech boosts by up to 15 dB at a 5:1 ratio. Dialnorm anchors the null band, so content at dialogue level is untouched. Two forms exist: Line mode for decoders with discrete outputs, and the more restrictive RF mode, which Sound On Sound says adds an 11 dB boost to a nominal -20 dBFS dialogue level.

Worked example from the standard

A/52 gives numbers. Dialogue sits at -25 dB, explosions reach 0 dB and quiet sounds are 50 dB below dialogue. The encoder writes -15 dB of gain for the loudest explosions and +20 dB for the quietest. A listener sets dialogue to 60 dB SPL and takes full compression: explosions reach 70 dB SPL (25 dB above dialogue, minus 15 dB), quiet sounds play at 30 dB SPL (50 dB below dialogue, plus 20 dB), and the range is 70 - 30 = 40 dB. At 70 dB SPL dialogue with boosts halved the explosions hit 80 dB SPL and quiet sounds sit at 30 dB SPL, a 50 dB range. With DRC ignored and dialogue at 75 dB SPL the explosions reach 100 dB SPL and quiet sounds 25 dB SPL, a 75 dB range, exactly as mixed.

Night mode in real receivers

Each brand wraps the metadata differently. Sony's STR-DH770 offers D. RANGE with COMP. ON (compression as intended by the engineer), COMP. AUTO (only when the source flags it) and COMP. OFF, but only for Dolby Digital and Dolby TrueHD, and only from the front panel. Yamaha's RX-V685 has Maximum (no adjustment), Standard (regular home use) and Minimum/Auto (clear sound at night or low volumes), applies to Dolby Digital and DTS bitstreams, and adjusts TrueHD automatically from the stream. Denon's AVR-X3800H separates two controls: Loudness Management, on by default, applies the Dynamic Compression setting plus dialogue normalization, while Dynamic Compression offers Auto, Low, Medium, High and Off, with Off the default. It works on Dolby Digital, Dolby Digital Plus, TrueHD, Atmos and compatible DTS signals, and Auto is unavailable for DTS. The same page lists a Dialog Control of 0 to 6 for compatible DTS:X.

Night mode vs other level controls

Compression is not the same as lowering the volume, and it is not dialogue enhancement. Volume moves everything, so turning down also sinks quiet passages toward the room's noise floor. DRC reshapes the range around dialogue level, which is why a 6 dB boost to quiet material and a 20:1 limit on peaks lets you watch at a low level. Dialogue enhancement targets speech directly (see dialogue enhancement), and downmixing changes channel balance (see Lt/Rt and Lo/Ro). Loudness targets sit upstream: Hybrik lists ATSC A/85 at -24 LKFS plus or minus 2 dB, EBU R128 at -23 LUFS plus or minus 1 dB and Netflix at -23 LKFS. Read dialnorm and reference level for the calibration context.

Common confusions

Night mode does not work on every source, and its name is not standard. Sony's setting ignores DTS, Yamaha's TrueHD handling is automatic, and Denon's Auto position is unavailable for DTS. Not every program carries a DRC flag either; Sony's COMP. AUTO acts only when the flag is present. The manuals checked here document night-style controls that act on Dolby (and for some brands DTS) metadata; whether another brand's "night mode" instead uses its own signal processing is not documented in them. Finally, DRC is part of the bitstream vs PCM question: whichever device decodes the stream is the one that applies or ignores the metadata, so when a TV decodes and sends PCM, the setting that matters is the one on the TV, not on the receiver. For Dolby Digital Plus see Dolby Digital and the AVR spec guide.