[FFmpeg-user] ffprobe -show_frames and coded_picture_number

thljcl jiachielee at live.com
Thu Aug 15 02:37:19 CEST 2013

It’s technically impossible to change the playback speed of audio without
affecting pitch, length, or other issues due to round-off
errors/quantization. As you said, though, it’s counting on less noticeable
perceptional difference. The waveforms of the audio would change
accordingly, albeit possibly very minor if done correctly, when you speed up
the audio. The original sound (analog) information is being encoded so that
it could be represented by combinations of digits. With enough sampling
rate, floating points, etc., the accuracy could be achieved that our
speakers could reproduce the audio signal that we still perceive pretty
close to the original signal. I do believe that digital audio is the right
solution compared with analog storage medium for audio; but the debate for
analog vs. digital does still persist in these days, especially among
audiophiles. I am personally in favor of digital audio because our
perception does have limitation; we cannot listen to sound wave of all
frequencies; we cannot see photons of every frequency either even though
technically the frequency of photon is frame-dependent.
As a matter of fact, when the audio is still in production stage, it’s very
common to have twice the sampling rate of the eventual sampling rate for
distribution. Some of the high-cost or high-value production of audios
producers may still keep “master copy” of such sampling rate. At the
production stage, they may re-mix the audios, speed up, changing speed,
pitch, etc. The famous song “Billie Jeans” has the mix of Michael Jackson’s
voice of various pitches. Of course, in the early days, the audio is being
produced without the assistance of computer; but Michael Jackson and many
other producers later do use computers to assist in audio production. Of
course, the audios for distribution would have its sampling rate to be
down-sampled. In fact, recently the rise of MP3 online store makes the audio
quality to be even worse for commercially-released songs due to lossy
compression. I bet the teenagers couldn’t care less with their headset.
Reproduction of frame-by-frame is not a perfect solution either, so to
speak. Only when the producers work hard, smart, and careful, then only we
can achieve high-quality reproduction. It needs substantial testing, of
course. But I do admit such reproduction is much laborious compared with
changing the playback speed of audio.

View this message in context: http://ffmpeg-users.933282.n4.nabble.com/ffprobe-show-frames-and-coded-picture-number-tp4660602p4660725.html
Sent from the FFmpeg-users mailing list archive at Nabble.com.

More information about the ffmpeg-user mailing list