fix dual output + enhanced broadcast (Twitch set vertical)

Repro steps:

In dual output, set horizontal to any platform and Twitch to vertical platform
Enable enhanced broadcasting
Go Live
Check feed on Twitch. You'll see stream with vertical video settings and content from horizontal display

Author: sandboxcoder

obs-studio-server/source/nodeobs_service.cpp

@@ -1401,7 +1401,8 @@ bool OBS_service::startMultiTrackStreaming(StreamServiceId serviceId, bool dualS
 	std::vector<OBSEncoderAutoRelease> audio_encoders;
 	std::shared_ptr<obs_encoder_group_t> video_encoder_group;
 	auto output =
-		osn::SetupOBSOutput("Enhanced Broadcasting", go_live_config.value(), audio_encoders, video_encoder_group, audio_encoder_id, 0, vod_track_mixer);
+		osn::SetupOBSOutput("Enhanced Broadcasting", go_live_config.value(), audio_encoders, video_encoder_group, audio_encoder_id, 0, vod_track_mixer,
+				    canvases);
 	if (!output) {
 		throw std::runtime_error("startStreaming - failed to create multitrack output");
 	}

obs-studio-server/source/osn-enhanced-broadcasting.hpp

@@ -81,7 +81,7 @@ template<typename BaseStreaming> class EnhancedBroadcasting : public BaseStreami
 		std::vector<OBSEncoderAutoRelease> audio_encoders;
 		std::shared_ptr<obs_encoder_group_t> video_encoder_group;
 		auto output = osn::SetupOBSOutput("Enhanced Broadcasting", go_live_config.value(), audio_encoders, video_encoder_group, audio_encoder_id, 0,
-						  vod_track_mixer);
+						  vod_track_mixer, canvases);
 		if (!output) {
 			throw std::runtime_error("startStreaming - failed to create multitrack output");
 		}

obs-studio-server/source/osn-multitrack-video-output.cpp

@@ -172,7 +172,8 @@ static OBSEncoderAutoRelease create_video_encoder(DStr &name_buffer, std::size_t
 	return video_encoder;
 }
 
-static bool create_video_encoders(const Config &go_live_config, std::shared_ptr<obs_encoder_group_t> &video_encoder_group, obs_output_t *output)
+static bool create_video_encoders(const Config &go_live_config, std::shared_ptr<obs_encoder_group_t> &video_encoder_group, obs_output_t *output,
+				   const std::vector<obs_video_info *> &canvases)
 {
 	DStr video_encoder_name_buffer;
 	if (go_live_config.encoder_configurations.empty()) {
@@ -186,18 +187,17 @@ static bool create_video_encoders(const Config &go_live_config, std::shared_ptr<
 		return false;
 	}
 
-	const auto max_canvas_idx = obs_get_video_info_count() - 1;
-
 	for (size_t i = 0; i < go_live_config.encoder_configurations.size(); i++) {
 		auto &config = go_live_config.encoder_configurations[i];
-		if (config.canvas_index > max_canvas_idx) {
-			blog(LOG_ERROR, "create_video_encoders - got unexpected large canvas index - i: %d", config.canvas_index);
+		if (config.canvas_index >= canvases.size()) {
+			blog(LOG_ERROR, "create_video_encoders - canvas_index %d out of range (canvases.size()=%zu) - encoder: %zu", config.canvas_index,
+			     canvases.size(), i);
 			return false;
 		}
 
-		obs_video_info *ovi = obs_get_video_info_by_index2(config.canvas_index);
+		obs_video_info *ovi = canvases[config.canvas_index];
 		if (!ovi) {
-			blog(LOG_ERROR, "create_video_encoders - failed to get canvas video info by index - i: %d", config.canvas_index);
+			blog(LOG_ERROR, "create_video_encoders - null canvas at index %d - encoder: %zu", config.canvas_index, i);
 			return false;
 		}
 
@@ -321,12 +321,12 @@ static OBSOutputAutoRelease create_output()
 
 OBSOutputAutoRelease SetupOBSOutput(const std::string &multitrack_video_name, const Config &go_live_config, std::vector<OBSEncoderAutoRelease> &audio_encoders,
 				    std::shared_ptr<obs_encoder_group_t> &video_encoder_group, const char *audio_encoder_id, size_t main_audio_mixer,
-				    std::optional<size_t> vod_track_mixer)
+				    std::optional<size_t> vod_track_mixer, const std::vector<obs_video_info *> &canvases)
 {
 
 	auto output = create_output();
 
-	if (!create_video_encoders(go_live_config, video_encoder_group, output))
+	if (!create_video_encoders(go_live_config, video_encoder_group, output, canvases))
 		return nullptr;
 
 	std::vector<speaker_layout> requested_speaker_layouts;

obs-studio-server/source/osn-multitrack-video-output.hpp

@@ -13,7 +13,7 @@ namespace osn {
 
 OBSOutputAutoRelease SetupOBSOutput(const std::string &multitrack_video_name, const Config &go_live_config, std::vector<OBSEncoderAutoRelease> &audio_encoders,
 				    std::shared_ptr<obs_encoder_group_t> &video_encoder_group, const char *audio_encoder_id, size_t main_audio_mixer,
-				    std::optional<size_t> vod_track_mixer);
+				    std::optional<size_t> vod_track_mixer, const std::vector<obs_video_info *> &canvases);
 
 OBSServiceAutoRelease create_service(const Config &go_live_config, const std::optional<std::string> &rtmp_url, const std::string &in_stream_key);