Downmix merges the center, LFE, surround and back channels of a multichannel recording into the front stereo pair. Use it when only two speakers (or headphones at the outputs) are connected, so that no information of the extra channels is lost.
Each group adds one channel family to the channels in front of it: center, LFE and surround are mixed into the stereo front pair, the rear channels into the surround channels:
Every group has its own Enabled switch. The Attenuation slider sets the level at which the channel is added to the mixture: a higher value attenuates it more, so you control how dominant dialog, bass or ambience remain in the stereo image.
Mixing channels together reduces the perceived loudness of the audio signal. The compressor compensates for this effect, at the price of slightly reduced dynamics. The Mode list switches the compressor off, configures it automatically in Auto mode, depending on which downmix options are enabled, or leaves the configuration to you in Manual mode.
The compressor only works together with the downmix: if you disable all downmix groups, the compressor group disappears and the signal is passed unmodified.
The graphic shows the compression at various ratios. The gray line represents a ratio of 1.0:1. This means that the incoming signal is output uncompressed and therefore unchanged. The green curves show the response at different ratios. Compression begins only after the threshold, which is drawn in the graphic as a dashed line; the graphic shows a threshold of -36 dB. The compression does not start abruptly after the threshold. It flows smoothly into the configured ratio.
The gray curve shows the level of the input signal over time. The green curve shows the compressed output signal for the respective attack setting.
With a short attack, the compressor engages almost immediately. This catches loud peaks, but can make the sound less punchy.
With a long attack, the compressor waits a moment before it attenuates more. The start of a sound, for example on drums, comes through better and sounds punchier.
The gray curve shows the level of the input signal over time. The green curve shows the compressed output signal for the respective release setting. The input falls to the quiet level at the dashed line.
With a short release, the compressor stops attenuating almost immediately after the level drops. Loud and quiet passages keep their contrast, but the volume can start pumping.
With a long release, the compressor keeps attenuating for a while after the level drops. Level changes are smoothed, but quiet passages stay quieter for longer.
The gray curve shows the response of a compression with a ratio of 6:1. The resulting signal at the output is audibly quieter. The make-up gain raises the loudness back up. In the graphic, the green curve represents a boost of 9 dB.