FreeCalypso > hg > fc-audio-config
changeset 11:46f49aa6ab8c
aud-fcdev3b: new spkr0 config with AEC and mic tuning
author | Mychaela Falconia <falcon@freecalypso.org> |
---|---|
date | Sat, 09 Oct 2021 04:12:55 +0000 |
parents | 7ad65d870b2f |
children | b57c380d78e7 |
files | aud-fcdev3b/Makefile aud-fcdev3b/spkr0.src |
diffstat | 2 files changed, 49 insertions(+), 1 deletions(-) [+] |
line wrap: on
line diff
--- a/aud-fcdev3b/Makefile Mon Aug 02 18:21:55 2021 +0000 +++ b/aud-fcdev3b/Makefile Sat Oct 09 04:12:55 2021 +0000 @@ -1,4 +1,4 @@ -BINFILES= mcsi.cfg mcsi.vol spkrmax.cfg spkrmax.vol +BINFILES= mcsi.cfg mcsi.vol spkr0.cfg spkr0.vol spkrmax.cfg spkrmax.vol INSTDIR= /opt/freecalypso/aud-fcdev3b all: ${BINFILES} @@ -9,6 +9,12 @@ mcsi.vol: Makefile tiaud-mkvol 150 $@ +spkr0.cfg: spkr0.src + tiaud-compile $< $@ + +spkr0.vol: Makefile + tiaud-mkvol 150 $@ + spkrmax.cfg: spkrmax.src tiaud-compile $< $@
--- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/aud-fcdev3b/spkr0.src Sat Oct 09 04:12:55 2021 +0000 @@ -0,0 +1,42 @@ +# The official voice call kit for FCDEV3B consists of a SparkFun COM-09151 +# loudspeaker and a CUI CMC-9745-130T microphone. The present audio mode +# config has been tuned as follows: +# +# 1) AEC is enabled as needed for all not-held-to-ear loudspeaker setups; +# +# 2) The microphone gain has been tuned to provide the most natural +# voice levels to parties on the far end of the call, matching standard +# phones; +# +# 3) The sidetone level has been turned down: the default -5 dB is too loud +# and causes singing effects with the raised microphone gain. + +mic default { + gain 8 # tuned for CMC-9745-130T and FCDEV3B mic input circuit + output-bias 0 # 2.0V appears to be correct for CMC-9745-130T +# Mic has very flat frequency response, thus no FIR needed +} + +# We don't have a FIR filter for our SparkFun COM-09151 loudspeaker. There is +# no datasheet for this speaker on SparkFun website, thus we don't know its +# frequency response curve, and even if we got the needed data, the work to +# create a flattening filter would probably be more than we can justify. + +speaker ear { + gain 0 # left at default for now +} + +# The canon seems to be to disable the sidetone altogether in loudspeaker +# setups, but we set it to the lowest level instead, so we keep exercising +# this chipset feature - our FCDEV3B is unusual in that the loudspeaker +# is the only way to exercise voice calls. + +sidetone -23 + +# The AEC configuration has been copied from Pirelli DP-L10, or more +# specifically, Pirelli's loudspeaker mode AEC config. Unfortunately +# we lack sufficient Calypso DSP documentation to even understand most +# of these tunable settings, let alone actually tune them, hence +# copying Pirelli's config is the best we can do at the moment. + +aec-new 2 1 0x14 0xCCC 0xFFF 0x4000 0x32 0x1000 0x1000 0 4 0x60