Hi List,
I'm currently trying to convert raw audio data from 64bit float (ieee
double precision) Big Endian to other formats (i.e. WAV, OGG, etc.) for
output using some linux audio player. The problem I'm running into is
that when importing this data into various programs (see below), the
audio clips outside of +/- magnitude=1. I don't understand why this is
and, more importantly, how to avoid/workaround this limitation.
So far I've tried and been able to successfully import the data in SoX,
Audacity, Rezound. For each of these programs, I've simply had to
specify RAW format, 64bit float, sample rate, # channels, Big Endian.
Using sox, I've successfully resampled, written other RAW files, written
WAV files....no problem. When the original raw audio file contains data
that is constrained within +/- 1 magnitude scale (i.e. 0.5 peak
magnitude sine wave w/no DC offset), there is no problem.
Questions:
1) is there something fundamentally incompatible with my original data
and "standard" audio data wrt range of values?
2) is the reason I'm seeing all of these programs clip my data at +/- 1
amplitude because they all use libraries with this limitation?
What I really want is a simple way to convert data from my original
format to other "playable" formats. SoX seems perfect if not for the
clipping issue. Though, I'd appreciate advice on other suitable
applications.
Any/All help greatly appreciated!
Rick
