Recognize all HID++ 2.0 device types: the x0005 DeviceName type table
stopped at receiver (7), so newer types resolved to no kind at all.
Add the remaining official types (the G560 reports 14 = speaker) with
matching HID++ 1.0 compatibility values, and themed icons for
speaker/microphone/webcam.
Name LED zone locations by device kind: the wire location codes are
physical-zone handles whose meaning depends on the device kind — the
existing table (Primary/Logo/Left Side/Right Side) matches mice and
keyboards. On speakers, codes 1-4 denote the left front, right
front, left rear and right rear lamps. Setting names derive from the
numeric code and are unchanged, so existing configurations keep
working.
Add the Audio Visualizer LED effect (0x07), the G560's factory
default: pulses lamp brightness with the audio signal. Wire layout
(hardware-derived): control byte at offset 0 (0 = pulse a fixed
color, 1 = self-cycle hue), RGB at 1-3, BE16 period in ms at 5-6.
Control values >= 2 select unmapped non-pulsing sub-modes and are not
exposed. The effect has no working intensity parameter; overall
brightness is governed by the separate BRIGHTNESS CONTROL (0x8040)
feature. Without a param table the effect displayed as Unknown(0x07)
and saved settings mis-split the period across the Cycle layout.
Strip the request echo from LED zone effect reads: GetEffect (0x8070
fn 0xE0) replies with the zone index echoed ahead of the effect slot
and parameters, but the reply was parsed as if it began with the
slot, shifting every readback one byte (only accidentally right when
the zone index happened to match the slot). HeteroValidator gains a
read_skip_byte_count option.
Repaint saved zone effects when LED Control is re-enabled: the
firmware starts a SW-control session blank, so without this the claim
looks like it did nothing (mirrors RGBControl on 0x8071).
All changes verified against a live G560 (046D:0A78); device scan
included.
38b5665e51