Hello,
recently I stumbled upon the (probably well-known) FM-specific filtering called
"Pre-Emphasis" (before broadcasting) and "De-Emphasis" (after receiving) which
is used for FM broadcasts worldwide (with slightly different filtering
parameters e.g. in US and Europe).
Here is the quick summary of this concept:
https://en.wikipedia.org/wiki/FM_broadcasting#Pre-emphasis_and_de-emphasis
Here at the community radio where I am doing a bit of technical stuff (mostly
IT) we use a single sound processing server for the final signal preparation
(for FM broadcasting).
Until now we use the very same signal for a web stream.
This is probably not a brilliant idea due to the above Pre-Emphasis (boosting
high frequencies).
Sadly I do not see a reasonable way to grab the processed signal before the
Pre-Emphasis filtering is applied. Thus my only option seems to be to apply a
De-Emphasis filter afterwards (similar to the one that is built into all FM
receivers).
I have to admit that I have a very limited amount of understanding with regards
to signal processing. Thus the technical description of the "De-Emphasis" filter
does not help me a lot:
"a simple RC filter circuit with a 50 µs time constant" (for Europe)
I found a forum thread [1] where the audacity community discusses such a
De-Emphasis filter, but I failed to understand how I could apply a similar
filter with liquidsoap's processing functions ([2]).
Additionally I found another source of information for a "De-Emphasis" filter
implementation in a 5-page design sheet [3]. It contained the following
four components for De-Emphasis:
VCO: Output Amplitude is 2 V; Quiescent Frequency is 20 kHz; Input Sensitivity
is 5 kHz/V; Initial Phase is 0.
PLL Low-Pass Filter: Butterworth; Filter Order is 1; Passband Edge Frequency
is 2π × 10500 rad/s.
Deemphasis Filter (First-Order Low-Pass Filter):
Time Constant is 1/(2π × 2100) s.
Low-Pass Filter: Butterworth; Filter Order is 4;
Passband Edge Frequency is 2π × 4000 rad/s.
Sadly again my understand of signal processing failed to turn this description
into a set of liquidsoap filters.
I would highly appreciate any partial hints pushing me into the right direction
in order to solve this puzzle.
Thank you for your time!
Cheers,
Lars
[1] http://forum.audacityteam.org/viewtopic.php?f=13&t=9141&start=10
[2] http://liquidsoap.fm/doc-svn/reference.html#Source___Sound_Processing
[3] http://www.ece.iastate.edu/~zluo/321f.pdf
------------------------------------------------------------------------------
_______________________________________________
Savonet-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/savonet-users
