encodeHD108_HD()

template<typename InputIterator, typename BrightnessIterator, typename OutputIterator>

void fl::encodeHD108_HD	(	InputIterator	first,
		InputIterator	last,
		BrightnessIterator	brightness_first,
		OutputIterator	out )

Encode pixel data in HD108 format with per-LED brightness.

Template Parameters

InputIterator	Iterator yielding fl::array<u8, 3> (3 bytes in wire order)
BrightnessIterator	Iterator yielding uint8_t brightness values
OutputIterator	Output iterator accepting uint8_t

Parameters

first	Iterator to first pixel
last	Iterator past last pixel
brightness_first	Accepted and never read. hd108BrightnessHeader() pins all gains at 31 regardless, so per-LED brightness has no effect on the output (FastLED#4042) – it used to be dereferenced once per LED and thrown away, which cost a call each time (FastLED#4402). Kept for source compatibility.
out	Output iterator for encoded bytes

Note

HD108 uses RGB wire order: pixel[0]=Red, pixel[1]=Green, pixel[2]=Blue

Definition at line 99 of file hd108.h.

                                                                                          {
    FL_UNUSED(brightness_first);
    // Start frame: 64 bits (8 bytes) of 0x00
    for (int i = 0; i < 8; i++) {
        *out++ = 0x00;
    }
 
    // One header for the whole strip. `hd108BrightnessHeader` discards its
    // argument -- `(void)brightness_8bit`, then three `constexpr` gains of 31
    // -- so it returns 0xFF 0xFF for every input. There was a per-LED cache
    // here guarding against recomputing a value that cannot change, and a
    // `*brightness_first` read feeding it; both computed the same two bytes
    // every time. Reading the brightness per LED was the more expensive half:
    // `ScaledPixelIteratorBrightness::load()` is out of line, so it was a call
    // per LED to fetch a number that reached nothing. See FastLED#4402.
    u8 f0, f1;
    hd108BrightnessHeader(0, &f0, &f1);
 
    // LED data: 2-byte header + 6-byte RGB16 (count as we go)
    size_t num_leds = 0;
    while (first != last) {
        const fl::array<u8, BYTES_PER_PIXEL_RGB>& pixel = *first;
 
        // Apply gamma correction (2.8) to 16-bit (RGB order: 0=R, 1=G, 2=B)
        u16 r16 = hd108GammaCorrect(pixel[0]);  // Red
        u16 g16 = hd108GammaCorrect(pixel[1]);  // Green
        u16 b16 = hd108GammaCorrect(pixel[2]);  // Blue
 
        // Transmit: 2 header + 6 color bytes (16-bit RGB, big-endian)
        *out++ = f0;
        *out++ = f1;
        *out++ = static_cast<u8>(r16 >> 8);
        *out++ = static_cast<u8>(r16 & 0xFF);
        *out++ = static_cast<u8>(g16 >> 8);
        *out++ = static_cast<u8>(g16 & 0xFF);
        *out++ = static_cast<u8>(b16 >> 8);
        *out++ = static_cast<u8>(b16 & 0xFF);
 
        ++first;
        ++num_leds;
    }
 
    // End frame: (num_leds / 2) + 4 bytes of 0xFF
    const size_t latch = num_leds / 2 + 4;
    for (size_t i = 0; i < latch; i++) {
        *out++ = 0xFF;
    }
}

References FL_NO_EXCEPT, FL_UNUSED, hd108BrightnessHeader(), and hd108GammaCorrect().

Here is the call graph for this function: