Witam!!!

Na stronie:
http://www.avrfreaks.net/index.php?module=Freaks%20A
cademy&func=viewItem&item_id=502&item_type=project

Jest takie zdanie, którego nie mogę "przetrawić"...

[...]
To decode FSK, you need to delay the incoming signal by half a
bit-period (the easy bit) and then low-pass filtering the result (the
hard bit). At the end of the process the resulting signal is passed
through a bit-slicer to determine the logic level of the original bit.
In this case, the input signal is a unipolar 8-bit sample but this is
promoted to a signed 16-bit result; otherwise, signal-to-noise issues
swamp the filter. Internally, the filter uses 16-bit signed scaled
values; temporary accumulators and multiplication results are 32
bits.

[...]


O co chodzi z tym "easy bit" i "hard bit" ?


Pozdrawiam