Recherche avancée

Sur d’autres sites (9255)

• aacenc : remove support for AAC Main profile

  8 février, par Lynne

  aacenc : remove support for AAC Main profile
  The Main profile of AAC is... terrible.
  
  It enables the use of delta coding across coefficients of two frames
  
  to try to increase compression, and it enabled one more pole for TNS
  
  filters.
  What the AAC authors failed to take into account were basic
  
  mathematics, as MDCT leakage (e.g. the spread of each frequency when
  
  represented in a discrete spectrum) is significant in most audio codecs.
  
  This leads to huge variations between each frame, basically rendering
  
  prediction completely pointless.
  In fact, enabling AAC-Main prediction does not, in general, even recoup
  
  the metadata losses from signalling the profile and prediction properties
  
  in the first place. So you lose efficiency by using AAC Main.
  The rumor is that it was put in the AAC spec for patent reasons, though
  
  patent-wise, it has about as much use as a patent for a bicycle designed
  
  for use by snakes.
  The only other thing AAC Main changes is it permits 3-pole TNS filters.
  
  When AAC's bands are absolutely tiny, except for very high frequency bands,
  
  where you're likely to use PNS instead.
  Just get rid of it.
  

  • [DH] doc/encoders.texi
  • [DH] libavcodec/Makefile
  • [DH] libavcodec/aaccoder.c
  • [DH] libavcodec/aacenc.c
  • [DH] libavcodec/aacenc.h
  • [DH] libavcodec/aacenc_pred.c
  • [DH] libavcodec/aacenc_pred.h
  • [DH] libavcodec/aacenc_tns.c
  • [DH] libavcodec/aacenctab.h
  • [DH] tests/fate/aac.mak

• avutil/libm : correct isnan, isinf compat hacks

  15 novembre 2015, par Ganesh Ajjanagadde

  avutil/libm : correct isnan, isinf compat hacks
  isnan and isinf are actually macros as per the standard. In particular,
  
  the existing implementation has incorrect signature. Furthermore, this
  
  results in undefined behavior for e.g double values outside float range
  
  as per the standard.
  This patch corrects the undefined behavior for all usage within FFmpeg.
  Note that long double is not handled as it is not used in FFmpeg.
  
  Furthermore, even if at some point long double gets used, it is likely
  
  not needed to modify the macro in practice for usage in FFmpeg. See
  
  below for analysis.
  Getting long double to work strictly per the spec is significantly harder
  
  since a long double may be an IEEE 128 bit quad (very rare), 80 bit
  
  extended precision value (on GCC/Clang), or simply double (on recent Microsoft).
  
  On the other hand, any potential future usage of long double is likely
  
  for precision (when a platform offers extra precision) and not for range, since
  
  the range anyway varies and is not as portable as IEEE 754 single/double
  
  precision. In such cases, the implicit cast to a double is well defined
  
  and isinf and isnan should work as intended.
  Reviewed-by : Michael Niedermayer <michael@niedermayer.cc>
  
  Signed-off-by : Ganesh Ajjanagadde <gajjanagadde@gmail.com>
  

  • [DH] libavutil/libm.h

• Python : How to decode a mp3 chunk into PCM samples ?

  30 mars 2021, par Bendzko

  I'm trying to catch chunks of an mp3 webstream and decoding them into PCM samples for signal processing. I tried to catch the audio via requests and io.BytesIO to save the data as .wav file.

  



  I have to convert the mp3 data to wav data, but I don't know how. (My goal is not to record a .wav file, i am just doing this to test the algorithm.)

  



  I found the pymedia lib, but it is very old (last commit in 2006), using python 2.7 and for me not installable.

  



  Maybe it is possible with ffmpeg-python, but I have just seen examples using files as input and output.

  



  Here's my code :

  



  import requests
import io
import soundfile as sf
import struct
import wave
import numpy as np


def main():
    stream_url = r'http://dg-wdr-http-dus-dtag-cdn.cast.addradio.de/wdr/1live/diggi/mp3/128/stream.mp3'
    r = requests.get(stream_url, stream=True)
    sample_array = []
    try:
        for block in r.iter_content(1024):
            data, samplerate = sf.read(io.BytesIO(block), format="RAW", channels=2, samplerate=44100, subtype='FLOAT',
                                       dtype='float32')
            sample_array = np.append(sample_array, data)

    except KeyboardInterrupt:
        print("...saving")
        obj = wave.open('sounds/stream1.wav', 'w')
        obj.setnchannels(1)  # mono
        obj.setsampwidth(2)  # bytes
        obj.setframerate(44100)

        data_max = np.nanmax(abs(sample_array))

        # fill WAV with samples from sample_array
        for sample in sample_array:
            if (np.isnan(sample) or np.isnan(32760 * sample / data_max)) is True:
                continue
            try:
                value = int(32760 * sample / data_max)  # normalization INT16
            except ValueError:
                value = 1
            finally:
                data = struct.pack('code>

  



  Do you have an idea how to handle this problem ?