C++ SDK

Seamless integration of Arctan SDK to C++ projects using CMake FetchContent, this will automatically download the required source archive and link them to your project.

Requirements

• CMake 3.15 or newer.
• A C++11 compiler.
• OpenSSL and libcurl available to CMake.
• An Arctan SDK license key.
• Network access

Integration

Add this to your CMake

include(FetchContent)

FetchContent_Declare(
    arctan_vi_sdk
    URL      "https://mediagateway.getarctan.com:8080/sdk/release?language=cpp&dist=41c09d5c-9be2-4753-ab01-20a28fbf7727&version=0.2.0"
    URL_HASH SHA256=d86e415a53c5103e4952e8f9d3ebf85f8c4a2fd2b17be2cf6771c2a19d45589f
)
FetchContent_MakeAvailable(arctan_vi_sdk)

target_link_libraries(my_audio_app PRIVATE arctan-vi)

and build your application normally

cmake -B build -S .
cmake --build build -j

License key

export ARCTAN_SDK_KEY="arc_sk_live_..."

You can also pass the key directly to arctan::Processor::create(...).

Process audio

#include "arctan.hpp"

#include <cstdlib>
#include <vector>

int main() {
  const char* license_key = std::getenv("ARCTAN_SDK_KEY");

  arctan::Model model = arctan::Model::create_default().take();
  uint32_t sample_rate = model.get_optimal_sample_rate();
  size_t num_frames = model.get_optimal_num_frames(sample_rate);

  arctan::ProcessorConfig config(sample_rate, num_frames, 2);
  arctan::Processor processor =
      arctan::Processor::create(model, license_key).take();

  processor.initialize(
      config.sample_rate,
      config.num_channels,
      config.num_frames,
      config.allow_variable_frames);

  std::vector<float> audio(config.num_channels * config.num_frames, 0.0f);
  processor.process_interleaved(
      audio.data(),
      config.num_channels,
      config.num_frames);
  processor.close();
}

Troubleshooting

1. If CMake cannot find OpenSSL or libcurl, install the development package for your platform and rerun CMake.

2. If authorization fails, confirm the SDK license key is valid and the application process has outbound HTTPS access.

3. If processing returns a frame-size error, use model.get_optimal_num_frames(sample_rate) or pass a frame count that contains a whole number of native processing blocks.