Drop timeBase from convertAVFrameToFrameOutput API

dvrogozh · dvrogozh · commit 8f899e16d815 · 2025-09-12T17:30:51.000Z
Signed-off-by: Dmitry Rogozhkin &lt;dmitry.v.rogozhkin@intel.com&gt;
diff --git a/src/torchcodec/_core/CpuDeviceInterface.cpp b/src/torchcodec/_core/CpuDeviceInterface.cpp
@@ -117,7 +117,6 @@ std::unique_ptr<FiltersContext> CpuDeviceInterface::initializeFiltersContext(
 // `dimension_order` parameter. It's up to callers to re-shape it if needed.
 void CpuDeviceInterface::convertAVFrameToFrameOutput(
     const VideoStreamOptions& videoStreamOptions,
-    [[maybe_unused]] const AVRational& timeBase,
     UniqueAVFrame& avFrame,
     FrameOutput& frameOutput,
     std::optional<torch::Tensor> preAllocatedOutputTensor) {
diff --git a/src/torchcodec/_core/CpuDeviceInterface.h b/src/torchcodec/_core/CpuDeviceInterface.h
@@ -33,7 +33,6 @@ class CpuDeviceInterface : public DeviceInterface {
 
   void convertAVFrameToFrameOutput(
       const VideoStreamOptions& videoStreamOptions,
-      const AVRational& timeBase,
       UniqueAVFrame& avFrame,
       FrameOutput& frameOutput,
       std::optional<torch::Tensor> preAllocatedOutputTensor =
diff --git a/src/torchcodec/_core/CudaDeviceInterface.cpp b/src/torchcodec/_core/CudaDeviceInterface.cpp
@@ -263,7 +263,6 @@ std::unique_ptr<FiltersContext> CudaDeviceInterface::initializeFiltersContext(
 
 void CudaDeviceInterface::convertAVFrameToFrameOutput(
     const VideoStreamOptions& videoStreamOptions,
-    [[maybe_unused]] const AVRational& timeBase,
     UniqueAVFrame& avFrame,
     FrameOutput& frameOutput,
     std::optional<torch::Tensor> preAllocatedOutputTensor) {
@@ -281,11 +280,7 @@ void CudaDeviceInterface::convertAVFrameToFrameOutput(
 
     FrameOutput cpuFrameOutput;
     cpuInterface->convertAVFrameToFrameOutput(
-        videoStreamOptions,
-        timeBase,
-        avFrame,
-        cpuFrameOutput,
-        preAllocatedOutputTensor);
+        videoStreamOptions, avFrame, cpuFrameOutput, preAllocatedOutputTensor);
 
     frameOutput.data = cpuFrameOutput.data.to(device_);
     return;
diff --git a/src/torchcodec/_core/CudaDeviceInterface.h b/src/torchcodec/_core/CudaDeviceInterface.h
@@ -28,7 +28,6 @@ class CudaDeviceInterface : public DeviceInterface {
 
   void convertAVFrameToFrameOutput(
       const VideoStreamOptions& videoStreamOptions,
-      const AVRational& timeBase,
       UniqueAVFrame& avFrame,
       FrameOutput& frameOutput,
       std::optional<torch::Tensor> preAllocatedOutputTensor =
diff --git a/src/torchcodec/_core/DeviceInterface.h b/src/torchcodec/_core/DeviceInterface.h
@@ -48,7 +48,6 @@ class DeviceInterface {
 
   virtual void convertAVFrameToFrameOutput(
       const VideoStreamOptions& videoStreamOptions,
-      const AVRational& timeBase,
       UniqueAVFrame& avFrame,
       FrameOutput& frameOutput,
       std::optional<torch::Tensor> preAllocatedOutputTensor = std::nullopt) = 0;
diff --git a/src/torchcodec/_core/SingleStreamDecoder.cpp b/src/torchcodec/_core/SingleStreamDecoder.cpp
@@ -1289,7 +1289,6 @@ FrameOutput SingleStreamDecoder::convertAVFrameToFrameOutput(
     }
     deviceInterface_->convertAVFrameToFrameOutput(
         streamInfo.videoStreamOptions,
-        streamInfo.timeBase,
         avFrame,
         frameOutput,
         preAllocatedOutputTensor);

Original file line number	Diff line number	Diff line change
`@@ -1289,7 +1289,6 @@ FrameOutput SingleStreamDecoder::convertAVFrameToFrameOutput(`
`1289`	`1289`	`}`
`1290`	`1290`	`deviceInterface_->convertAVFrameToFrameOutput(`
`1291`	`1291`	`streamInfo.videoStreamOptions,`
`1292`		`- streamInfo.timeBase,`
`1293`	`1292`	`avFrame,`
`1294`	`1293`	`frameOutput,`
`1295`	`1294`	`preAllocatedOutputTensor);`