Warn if decoding YUV images with different plane size (#3201)

Summary:
StreamReader behaves differently when dealing with YUV formats.
It implicitly converts the image format to YUV444P because
otherwise image planes do not have the same shape and it is not
possible to express it as a regular PyTorch Tensor with dedicated
dimension for each color channel.

This is commit adds warnings to such conversions.

Pull Request resolved: https://github.com/pytorch/audio/pull/3201

Reviewed By: nateanl

Differential Revision: D44311017

Pulled By: mthrok

fbshipit-source-id: 73a02a19c013c0263f349e1f3a3603e3d3eddb6a

torchaudio/csrc/ffmpeg/stream_reader/conversion.cpp

@@ -207,7 +207,12 @@ torch::Tensor PlanarImageConverter::convert(const AVFrame* src) {
 ////////////////////////////////////////////////////////////////////////////////
 YUV420PConverter::YUV420PConverter(int h, int w)
     : ImageConverterBase(h, w, 3),
-      tmp_uv(get_image_buffer({1, 2, height / 2, width / 2})) {}
+      tmp_uv(get_image_buffer({1, 2, height / 2, width / 2})) {
+  TORCH_WARN_ONCE(
+      "The output format YUV420P is selected. "
+      "This will be implicitly converted to YUV444P, "
+      "in which all the color components Y, U, V have the same dimension.");
+}
 
 void YUV420PConverter::convert(const AVFrame* src, torch::Tensor& dst) {
   TORCH_INTERNAL_ASSERT_DEBUG_ONLY(src);
@@ -269,7 +274,12 @@ torch::Tensor YUV420PConverter::convert(const AVFrame* src) {
 ////////////////////////////////////////////////////////////////////////////////
 NV12Converter::NV12Converter(int h, int w)
     : ImageConverterBase(h, w, 3),
-      tmp_uv(get_image_buffer({1, height / 2, width / 2, 2})) {}
+      tmp_uv(get_image_buffer({1, height / 2, width / 2, 2})) {
+  TORCH_WARN_ONCE(
+      "The output format NV12 is selected. "
+      "This will be implicitly converted to YUV444P, "
+      "in which all the color components Y, U, V have the same dimension.");
+}
 
 void NV12Converter::convert(const AVFrame* src, torch::Tensor& dst) {
   TORCH_INTERNAL_ASSERT_DEBUG_ONLY(src);
@@ -331,7 +341,12 @@ NV12CudaConverter::NV12CudaConverter(int h, int w, const torch::Device& device)
       tmp_uv(get_image_buffer(
           {1, height / 2, width / 2, 2},
           device,
-          torch::kUInt8)) {}
+          torch::kUInt8)) {
+  TORCH_WARN_ONCE(
+      "The output format NV12 is selected. "
+      "This will be implicitly converted to YUV444P, "
+      "in which all the color components Y, U, V have the same dimension.");
+}
 
 void NV12CudaConverter::convert(const AVFrame* src, torch::Tensor& dst) {
   TORCH_INTERNAL_ASSERT_DEBUG_ONLY(src);
@@ -403,7 +418,12 @@ P010CudaConverter::P010CudaConverter(int h, int w, const torch::Device& device)
       tmp_uv(get_image_buffer(
           {1, height / 2, width / 2, 2},
           device,
-          torch::kInt16)) {}
+          torch::kInt16)) {
+  TORCH_WARN_ONCE(
+      "The output format P010 is selected. "
+      "This will be implicitly converted to YUV444P, "
+      "in which all the color components Y, U, V have the same dimension.");
+}
 
 void P010CudaConverter::convert(const AVFrame* src, torch::Tensor& dst) {
   TORCH_INTERNAL_ASSERT_DEBUG_ONLY(src);