Fix Art-Net DMX channel limitation: support full 512 channels per universe

[Copilot]

Problem

Art-Net device limited to 108 DMX channels (36 RGB LEDs) instead of 512 channels (170 RGB LEDs). Root cause: write() function iterated over _ledRGBCount (byte count = LEDs × 3) while treating loop index as LED count, terminating after 108 bytes.

Buggy code:

for (unsigned int ledIdx = 0; ledIdx < _ledRGBCount; ledIdx++)  // _ledRGBCount = 108 for 36 LEDs
{
    artnet_packet.Data[dmxIdx++] = rawdata[ledIdx];
    if ((ledIdx % 3 == 2) && (ledIdx > 0))
        dmxIdx += (_artnet_channelsPerFixture - 3);
    // Loop ends at ledIdx=107, leaving channels 109-512 uninitialized
}

Changes

Core fix

• Iterate over LED count, not byte count: for (ledIdx = 0; ledIdx < ledValues.size(); ledIdx++)
• Explicit RGB channel access: rawdata[ledIdx * 3 + 0/1/2] instead of sequential byte iteration

Buffer safety

• Check packet capacity before writing: if (dmxIdx + _artnet_channelsPerFixture > DMX_MAX)
• Send final packet after loop completion (was conditional on loop iteration)

Error handling

• Fixed return value: if (writeBytes(...) < 0) retVal = -1
• Previously: retVal &= writeBytes(...) with retVal = 0 always returned 0

Validation

• Runtime check: _artnet_channelsPerFixture >= 3
• Schema constraint: "minimum": 3 for channelsPerFixture

Fixed code:

for (unsigned int ledIdx = 0; ledIdx < static_cast<unsigned int>(ledValues.size()); ledIdx++)
{
    if (dmxIdx + _artnet_channelsPerFixture > DMX_MAX)
    {
        // Send packet before overflow
        prepare(thisUniverse, _artnet_seq, dmxIdx);
        if (writeBytes(18 + dmxIdx, artnet_packet.raw) < 0)
            retVal = -1;
        memset(artnet_packet.raw, 0, sizeof(artnet_packet.raw));
        thisUniverse++;
        dmxIdx = 0;
    }
    
    artnet_packet.Data[dmxIdx++] = rawdata[ledIdx * 3 + 0]; // Red
    artnet_packet.Data[dmxIdx++] = rawdata[ledIdx * 3 + 1]; // Green
    artnet_packet.Data[dmxIdx++] = rawdata[ledIdx * 3 + 2]; // Blue
    dmxIdx += (_artnet_channelsPerFixture - 3);
}

Impact

Enables full 512-channel DMX transmission per universe, supporting ~170 RGB LEDs vs. previous 36 LED limitation.

Original prompt

Bug Description

There is a critical bug in the Art-Net LED device implementation that limits transmission to only 108 DMX channels (36 RGB LEDs) instead of the full 512 channels (170 RGB LEDs) per universe.

Symptoms

1. Channel limitation: Only channels 1-108 are transmitted correctly
2. Unstable values: Channels 109-512 show random/constantly changing DMX values when using Art-Net
3. Expected behavior: Should support up to 512 channels (approximately 170 RGB devices) per universe

Root Cause

The bug is in libsrc/leddevice/dev_net/LedDeviceUdpArtNet.cpp in the write() function (lines 103-122).

The loop incorrectly iterates over _ledRGBCount (total number of RGB bytes: LEDs × 3) while treating ledIdx as if it counts individual LEDs. This causes the loop to terminate after processing exactly 108 bytes (36 LEDs × 3 colors).

Current buggy code:

for (unsigned int ledIdx = 0; ledIdx < _ledRGBCount; ledIdx++)
{
    artnet_packet.Data[dmxIdx++] = rawdata[ledIdx];
    if ( (ledIdx % 3 == 2) && (ledIdx > 0) )
    {
        dmxIdx += (_artnet_channelsPerFixture-3);
    }

    if ( (ledIdx == _ledRGBCount-1) || (dmxIdx >= DMX_MAX) )
    {
        prepare(thisUniverse, _artnet_seq, dmxIdx);
        retVal &= writeBytes(18 + qMin(dmxIdx, DMX_MAX), artnet_packet.raw);

        memset(artnet_packet.raw, 0, sizeof(artnet_packet.raw));
        thisUniverse ++;
        dmxIdx = 0;
    }
}

Example: With 36 LEDs:

• _ledRGBCount = 108 (36 LEDs × 3 bytes)
• Loop runs: ledIdx = 0 to 107
• At ledIdx = 107, the condition (107 % 3 == 2) is true
• After processing byte 107 and adding channel spacing, the loop ends because it reaches _ledRGBCount
• Channels 109-512 are never written, containing garbage/uninitialized data

Additional Issues Found

1. Return value bug (line 116): retVal &= writeBytes(...) with retVal initialized to 0 will always return 0 (failure) even on success
2. Incorrect size calculation (line 116): Using qMin(dmxIdx, DMX_MAX) is unnecessary and incorrect since the condition already prevents dmxIdx from exceeding DMX_MAX

Required Fix

Replace the write() function in libsrc/leddevice/dev_net/LedDeviceUdpArtNet.cpp with the corrected version that:

1. Iterates over LEDs (not RGB bytes): Loop over ledValues.size() instead of _ledRGBCount
2. Explicitly writes R, G, B for each LED: Access rawdata[ledIdx * 3 + 0/1/2]
3. Fixes return value handling: Initialize retVal properly and accumulate errors correctly
4. Removes unnecessary qMin(): Use dmxIdx directly in the size calculation

Corrected implementation:

int LedDeviceUdpArtNet::write(const QVector<ColorRgb> &ledValues)
{
	int retVal = 0;
	int thisUniverse = _artnet_universe;
	auto rawdata = reinterpret_cast<const uint8_t *>(ledValues.data());

	/*
	This field is incremented in the range 0x01 to 0xff to allow the receiving node to resequence packets.
	The Sequence field is set to 0x00 to disable this feature.
	*/
	if (_artnet_seq++ == 0)
	{
		_artnet_seq = 1;
	}

	int dmxIdx = 0; // offset into the current dmx packet

	memset(artnet_packet.raw, 0, sizeof(artnet_packet.raw));
	
	// Iterate over LEDs, not RGB bytes
	for (unsigned int ledIdx = 0; ledIdx < static_cast<unsigned int>(ledValues.size()); ledIdx++)
	{
		// Write R, G, B for this LED
		artnet_packet.Data[dmxIdx++] = rawdata[ledIdx * 3 + 0]; // Red
		artnet_packet.Data[dmxIdx++] = rawdata[ledIdx * 3 + 1]; // Green
		artnet_packet.Data[dmxIdx++] = rawdata[ledIdx * 3 + 2]; // Blue
		
		// Skip extra channels if fixture needs more than RGB
		dmxIdx += (_artnet_channelsPerFixture - 3);

		// is this the last LED of last packet || last LED that fits in current packet
		if ( (ledIdx == static_cast<unsigned int>(ledValues.size()) - 1) || (dmxIdx >= DMX_MAX) )
		{
			prepare(thisUniverse, _artnet_seq, dmxIdx);
			
			if (writeBytes(18 + dmxIdx, artnet_packet.raw) < 0)
			{
				retVal = -1;
			}

			memset(artnet_packet.raw, 0, sizeof(artnet_packet.raw));
			thisUniverse++;
			dmxIdx = 0;
		}
	}

	return retVal;
}

Testing

This fix has been validated against the reported issue where a Swisson XMT-500 device (supporting both Art-Net and sACN) only received valid data on channels 1-108, with channels 109-512 showing unstable/random values.

After the fix, all 512 channels per universe should work correctly, supporting up to ~170 RGB LEDs per universe.

Files to Modify

• libsrc/leddevice/dev_net/LedDeviceUdpArtNet.cpp - Replace the write() function (lines 85-125)

This pull request was created from Copilot chat.

---

💬 We'd love your input! Share your thoughts on Copilot coding agent in our 2 minute survey.

[hyperion-project]

Hello @copilot 👋

I'm the Hyperion Project Bot and I want to thank you for
contributing to Hyperion with your pull requests!

To help you and other users test your pull requests faster,
I'll create a link for you to your workflow artifacts.

🔗 https://github.com/hyperion-project/hyperion.ng/actions/runs/22155635218

Of course, if you make changes to your PR, I will create a new link.

Best regards,
Hyperion Project

[hyperion-project]

Hey @copilot I created a new link to your workflow artifacts:
🔗 https://github.com/hyperion-project/hyperion.ng/actions/runs/22155710172

[hyperion-project]

Hey @copilot I created a new link to your workflow artifacts:
🔗 https://github.com/hyperion-project/hyperion.ng/actions/runs/22155776739

[hyperion-project]

Hey @copilot I created a new link to your workflow artifacts:
🔗 https://github.com/hyperion-project/hyperion.ng/actions/runs/22155796478

[hyperion-project]

Hey @copilot I created a new link to your workflow artifacts:
🔗 https://github.com/hyperion-project/hyperion.ng/actions/runs/22239665784

[hyperion-project]

Hey @Copilot I created a new link to your workflow artifacts:
🔗 https://github.com/hyperion-project/hyperion.ng/actions/runs/22240455249

[hyperion-project]

Hey @copilot I created a new link to your workflow artifacts:
🔗 https://github.com/hyperion-project/hyperion.ng/actions/runs/22773721041