Ignore return_complex when returning real-valued tensor in spectrogram. (#1551)

Author: mthrok

torchaudio/functional/functional.py

@@ -73,11 +73,11 @@ def spectrogram(
         onesided (bool, optional): controls whether to return half of results to
             avoid redundancy. Default: ``True``
         return_complex (bool, optional):
-            ``return_complex = True``, this function returns the resulting Tensor in
-            complex dtype, otherwise it returns the resulting Tensor in real dtype with extra
-            dimension for real and imaginary parts. (see ``torch.view_as_real``).
-            When ``power`` is provided, the value must be False, as the resulting
-            Tensor represents real-valued power.
+            Indicates whether the resulting complex-valued Tensor should be represented with
+            native complex dtype, such as `torch.cfloat` and `torch.cdouble`, or real dtype
+            mimicking complex value with an extra dimension for real and imaginary parts.
+            This argument is only effective when ``power=None``.
+            See also ``torch.view_as_real``.
 
     Returns:
         Tensor: Dimension (..., freq, time), freq is
@@ -92,11 +92,6 @@ def spectrogram(
             "for more details about torchaudio's plan to migrate to native complex type."
         )
 
-    if power is not None and return_complex:
-        raise ValueError(
-            'When `power` is provided, the return value is real-valued. '
-            'Therefore, `return_complex` must be False.')
-
     if pad > 0:
         # TODO add "with torch.no_grad():" back when JIT supports it
         waveform = torch.nn.functional.pad(waveform, (pad, pad), "constant")

torchaudio/transforms.py

@@ -60,11 +60,11 @@ class Spectrogram(torch.nn.Module):
         onesided (bool, optional): controls whether to return half of results to
             avoid redundancy Default: ``True``
         return_complex (bool, optional):
-            ``return_complex = True``, this function returns the resulting Tensor in
-            complex dtype, otherwise it returns the resulting Tensor in real dtype with extra
-            dimension for real and imaginary parts. (see ``torch.view_as_real``).
-            When ``power`` is provided, the value must be False, as the resulting
-            Tensor represents real-valued power.
+            Indicates whether the resulting complex-valued Tensor should be represented with
+            native complex dtype, such as `torch.cfloat` and `torch.cdouble`, or real dtype
+            mimicking complex value with an extra dimension for real and imaginary parts.
+            This argument is only effective when ``power=None``.
+            See also ``torch.view_as_real``.
     """
     __constants__ = ['n_fft', 'win_length', 'hop_length', 'pad', 'power', 'normalized']