Version 1.7
===========
* Settings: Rename source
* Setup Assistent added
* Signal Processor: Down-Mix implemented
* Signal Processor: The downmix headroom is computed from the power sum of the mixed channels, a 5.1 downmix no longer attenuates the front channels by about 10 dB below the level of stereo content.
* Signal Processor: Bus compressor added at the end of the downmix menu (linked detector over all channels, no panning shifts), it reduces the loudness difference between movie sound and music, a ratio of 100 operates as a limiter. The auto mode derives the parameters from the downmix configuration, frames without downmix action (e.g. stereo content) skip the compressor.
* Signal Processor: The controls of inactive features (downmix attenuation, bass management tuning, compressor parameters) are grayed out and reject value changes, the user sees which settings are effective.
* Signal Processor: Support for Satellite/Subwoofer-Speakers added
* The UI has been improved to help users better understand which signal is being output on which output and channel.
* Input status: The complete status (format, bits, channels, channel layout) is updated as soon as a signal is detected, also on streams that transfer zero samples (digital silence). Streams with 24bit content are displayed with 24 bits.
* Input status: AC3/DCA streams are decoded and played at the end of each repetition frame. Streams with a payload size that is not a multiple of 4 bytes (e.g. 2.0 AC3 with an odd frame size) were never decoded (no playback start), and the PCM frame that is sent between the format detections crashed with an empty channel layout.
* Signal Processor: Parametric equalizer added: up to 9 bands per output channel (4 on the LFE), peaking, low shelf, high shelf and notch filters, 20 Hz ... 20 kHz, gain -18 ... +18 dB. The bands are folded into the bass management kernel (no additional latency stages), the maximum boost of the resulting kernel is shown per position. The equalizer is off by default: a global enable switch at the top of the equalizer menu switches the whole equalizer in and out of the signal path for a direct A/B comparison.
* Signal Processor: Night Mode added: three preset stages (light, medium, strong) and a custom stage attenuate the bass with a high-pass folded into the bass management kernel and compress the dynamics with a bus compressor. Automatic loudness compensation follows the gain reduction of the compressor, so the perceived loudness stays at the level of the "off" stage in every channel layout and program level: loud passages are tamed and brought back over time, quiet passages stay quiet. Dialog focus lowers all channels except the center (effective with center channel in the layout).
* Signal Processor: Listening position added: distance per output channel (30 ... 2000 cm) derives the speaker delay (shown in ms) and a distance falloff gain per channel; the intensity control (0 ... 100 %) scales the gain, the LFE level is excluded from the gain.
* Signal Processor: Sample rate limit added: streams above the limit are downsampled with a linear-phase anti-alias filter, e.g. 192 kHz to 96 kHz or 44.1 kHz. The sample rate family is always preserved: 44.1 kHz family streams (44.1/88.2/176.4 kHz) never become 48 kHz family rates, the pitch never changes.
* Internal input added (with the feature package dsc8-control-desktop): a new source "Internal" (position 0, name editable) captures the audio of the snd-aloop loopback card, e.g. streams of players on this machine (direct ALSA or via PipeWire). The capture adopts rate, channels and format of the running stream, and it detects the access class of the peer (interleaved/non-interleaved, e.g. PipeWire's MMAP_NONINTERLEAVED) and opens the capture device with the matching class, so the kernel's format check (loopback_check_format) is satisfied without any client configuration. The input status shows the real format, rate and channels of the stream.
* Backup and restore added: The settings and the control states are exported to a JSON file (menu item "Backup" in Settings/System) and imported again (menu item "Restore"); the content is applied immediately, controls that are missing in the file keep their current value and the response reports the controls that could not be applied. Both functions are REST endpoints (GET /application/backup, PUT /application/restore) and require the connected sound card; the menu items are hidden on the device display (kiosk) and on phones and tablets.

Version 1.6
===========
* Initial public version. 
* Implements the API via REST.
* Implements the controls for switching the inputs.
* Implements the controls for volume und mute.
* Implements an input decoder for AC3 and DCA. Status controls are included.
* Implements a signal generator. You can use it to test if your speakers are correctly connected.
* Implements a phase inverter for the output.
* Implements controls to change the channel layout for the output.
* The settings of the ALSA controls a saved at program start and restored at program stop.
* The settings of all controls are stored after change. These stored settings are loaded at the next program start.
* Changes on controls are notified over a WebSocket.    