Standard configurations are all the typical configurations, such as stereo and 5.1, where each channel has a name giving a hint about the placement of its corresponding speaker. For example, "Height Back Left" is placed to the back and to the left of the listener, above the ground.
The LFE (.1) is never downmixed. If both the input and output configurations have an LFE channel, the signal of the LFE channel of the input is copied over to that of the output. If the output does not have an LFE channel, the signal of the LFE channel of the input is dropped.
With 4.x and 5.x configurations, there is only one pair of speakers that is not in the front, so there is no distinction between "side" or "back" channels. They are both synonyms of "surround." For these configurations, we chose to consistently call surround channels SL and SR, where "S" stands for "surround." This led to a somewhat arbitrary decision whereby the surround channels SL and SR of 4.x and 5.x configurations are mapped to the side channels SL and SR of 6.x and 7.x configurations, instead of the back channels BL and BR. This preference is justified by the fact that the preferred angles of surround speakers in 5.1 setups, as advocated by Dolby, matches those of the side speakers in 7.1 setups.
At the moment, there is no way to configure Wwise to route surround channels of 4.x or 5.x configurations into the back channels of 6.x or 7.x configurations.
The Wwise downmix performs no normalization, given the arbitrary nature of normalization coefficients. It is, therefore, common that downmixing from configurations with large numbers of channels to configurations with few channels results in loudness gains. Designers should take care of normalizing according to their taste using States or RTPCs on Bus volume(s).
An ambisonic configuration routed to an ambisonic configuration of different order drops input channels or fills output channels with silence. Ambisonic configurations routed to standard configurations are said to be "decoded". Standard configurations routed to ambisonic configurations are said to be "encoded".
Anonymous configurations do not implement downmixing. Routing behaves like the LFE channel in standard configurations. If a corresponding channel exists in the output, it is copied over; otherwise, it is dropped.