求助：如何通过Camera2 API获取Pixel 8 Pro的50MP原生相机分辨率

1. 定位正确的摄像头ID #

Pixel 8 Pro的50MP主摄对应逻辑摄像头ID，需遍历摄像头特性筛选：

• 优先筛选硬件级别为INFO_SUPPORTED_HARDWARE_LEVEL_FULL的摄像头
• 检查SCALER_STREAM_CONFIGURATION_MAP中是否包含8165×6124分辨率的输出尺寸

代码示例：

CameraManager manager = (CameraManager) getSystemService(Context.CAMERA_SERVICE);
String targetCameraId = null;
for (String cameraId : manager.getCameraIdList()) {
    CameraCharacteristics characteristics = manager.getCameraCharacteristics(cameraId);
    Integer hardwareLevel = characteristics.get(CameraCharacteristics.INFO_SUPPORTED_HARDWARE_LEVEL);
    if (hardwareLevel != null && hardwareLevel == CameraCharacteristics.INFO_SUPPORTED_HARDWARE_LEVEL_FULL) {
        StreamConfigurationMap configMap = characteristics.get(CameraCharacteristics.SCALER_STREAM_CONFIGURATION_MAP);
        if (configMap != null) {
            Size[] jpegSizes = configMap.getOutputSizes(ImageFormat.JPEG);
            for (Size size : jpegSizes) {
                if (size.getWidth() == 8165 && size.getHeight() == 6124) {
                    targetCameraId = cameraId;
                    break;
                }
            }
        }
    }
}

2. 配置高分辨率拍摄请求 #

• 确保ImageReader的尺寸设置为8165×6124，格式对应ImageFormat.JPEG
• 重置传感器裁剪区域为全画幅，避免被默认裁剪：

captureRequestBuilder.set(CaptureRequest.SCALER_CROP_REGION, 
    characteristics.get(CameraCharacteristics.SENSOR_INFO_ACTIVE_ARRAY_SIZE));

• 若开启ZSL（零延迟快门）导致分辨率降级，需关闭CONTROL_ENABLE_ZSL参数

3. 适配高分辨率模式限制 #

• 50MP模式下预览分辨率会自动降级为12MP，仅拍照阶段切换到高分辨率，无需强制预览匹配
• 如需夜景、HDR等优化效果，需添加对应请求参数：

  captureRequestBuilder.set(CaptureRequest.CONTROL_MODE, CameraMetadata.CONTROL_MODE_AUTO);
  captureRequestBuilder.set(CaptureRequest.CONTROL_SCENE_MODE, CameraMetadata.CONTROL_SCENE_MODE_HDR);
  

• 相机会话的输出Surface需包含对应分辨率的ImageReader，避免与低分辨率Surface混用

4. 备选方案：RAW转JPEG #

若直接输出JPEG仍为12MP，可先获取RAW格式（ImageFormat.RAW10/ImageFormat.RAW12）数据，再自行编码为50MP JPEG：

ImageReader rawReader = ImageReader.newInstance(8165, 6124, ImageFormat.RAW10, 1);
rawReader.setOnImageAvailableListener(reader -> {
    Image rawImage = reader.acquireNextImage();
    // 处理RAW数据并编码为JPEG（可使用Android原生Bitmap或第三方编库）
    rawImage.close();
}, null);
// 将RAW reader的Surface添加到相机会话配置
List<Surface> surfaces = new ArrayList<>();
surfaces.add(rawReader.getSurface());
cameraDevice.createCaptureSession(surfaces, sessionCallback, null);

内容的提问来源于stack exchange，提问作者MayurDev