Floatbeat format assumes code output is −1.0 to 1.0, while usual
bytebeat expects 0 to 255. Basically, floatbeat code can be
converted and played as bytebeat, and vice versa:
floatbeat = (bytebeat - 127.5) / 127.5
bytebeat = floatbeat * 127.5 + 127.5
But there will be differences in sound! The differences start in
signal processing. The bytebeat code is converted to the range from
0 to 255 like this:
output = value & 255
Example:
output = 0 & 255 = 0
output = 0.5 & 255 = 0
output = 100 & 255 = 100
output = 100.2 & 255 = 100
output = 255 & 255 = 255
output = 256 & 255 = 0
output = 300 & 255 = 44
output = 300.99 & 255 = 44
So, bytebeat signal has a limited range of only 256 values. I.e. 8
bit, or 1 byte. That's why it called BYTEbeat.
On the other side, floatbeat is of very high quality because it is
limited only by the double precision floating point data type, and
in the range from -1 to 1 it has an almost infinite number of
floating point values:
output = -0.1
output = 0.004
output = 5.5e-7
That is, if you convert floatbeat as
bytebeat = floatbeat *
127.5 + 127.5 then when playing in bytebeat mode there will
be a loss of sound quality and you can notice the unwanted noise.
It can be very noticeable on some quiet songs. Listen for