问题背景:
在windows下配置了anaconda的深度学习的tensorflow环境来跑语音数据,在load数据时发现竟然报错Raise NoBackendError(),简直疯了,what?竟然没有后台是神马情况,只能看源码。
debug进入源码:
with audioread.audio_open(os.path.realpath(path)) as input_file:
这里报错,继续深究,原来它使用的是这个源码:
def audio_open(path):
"""Open an audio file using a library that is available on this
system.
"""
# Standard-library WAV and AIFF readers.
from . import rawread
try:
return rawread.RawAudioFile(path)
except DecodeError:
pass
# Core Audio.
if _ca_available():
from . import macca
try:
return macca.ExtAudioFile(path)
except DecodeError:
pass
# GStreamer.
if _gst_available():
from . import gstdec
try:
re