Hans,

More feedback on the beta cx18 driver....

Problem statement: auto detection of broadcast audio mode doesn't appear
to work, mute doesn't work, and stereo audio is not available.

$ v4l2-dbg -d /dev/video1 --log-status 

always reports "mono" and "muted"

   cx18-0: Detected audio mode:       mono
   cx18-0: Detected audio standard:   BTSC
   cx18-0: Audio muted:               yes
   cx18-0: Audio microcontroller:     running
   cx18-0: Configured audio standard: automatic detection
   cx18-0: Configured audio system:   BTSC
   cx18-0: Specified audio input:     Tuner (In8)
   cx18-0: Preferred audio mode:      stereo


My findings:

1. cx18 driver initially starts up using Tuner AF line audio via the
CS5345 multiplexer/digitizer and not the Tuner SIF to the
CX23418/CX25843 demodulator/digitizer.  Since the AF from the tuner is
only 1 wire, it is monoaural and never stereo.

Here's how to prove that the AUDIO8(CVBS8/SIF5) input on the CX23418 is
*not* the source for the audio in the default MPEG encoding stream, by
toggling the CS5345 Mute:

$ mplayer /dev/video1
(in second window)
# /usr/local/bin/v4l2-dbg -d /dev/video1 -R type=i2cdrv,chip=cs5345
ioctl: VIDIOC_DBG_G_REGISTER

          00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F
00000000:    e2 00 08 01 10 40 00 00 01 00 00 c0 00 00 00 
00000010: 00 
# /usr/local/bin/v4l2-dbg -d /dev/video1  \
        -r type=i2cdrv,chip=cs5345,reg=0x04,val=0x05
register 0x4 set to 0x5
# /usr/local/bin/v4l2-dbg -d /dev/video1  \
        -r type=i2cdrv,chip=cs5345,reg=0x04,val=0x01
register 0x4 set to 0x1


2. I cannot mute the audio by setting all the mutes in the cx25843 core
in registers 0x02c408d3 and 0x02c408e3.  The registers appear to have
only the mute bits documented in the cx25843 data sheet and no new ones,
as writes to higher order bits in those registers don't work (the higher
bits stay 0).

The only way to mute appears to be via the CS5345 mute or by commanding
the encoder to mute.


3. In trying to investigate the audio routing through the system, I
found a block diagram of the CX23418 at 
http://www.saint-tec.co.jp/conexant.html
showing that the CX23418 has 2 serial audio inputs.

Since the CX23416 (not a typo) had 1 serial audio input, IIRC, I am
*guessing* one serial audio input on the CX23418 corresponds to the
CX25843 core and the other serial audio input goes to the encoder.  *If*
this is the case (can anyone confirm?), then I wouldn't know how to get
the encoder to listen on the serial audio output of the CX25843 core
anyway.  This would leave me stuck with the Tuner AF mono output from
the CS5345.  

Does anyone have a real CX23418 data sheet I can look at?  Will Conexant
want me to sign an NDA to do so, if I go to them?


4. The cx18 driver doesn't setup the Analog Front End (AFE) of the
CX25843 core to properly to process Sound IF (SIF) on the AUDIO8 input.
Specifically, I think VGA_SEL_CH3, DROOP_COMP_CH3, CLAMP_EN_CH3 are
wrong compared to what the CX25843 datasheet says in section 3.20 on
page 3-83.


5. The cx18 driver allows CX25843 Chip autoconfig based on the setting
in register 0x02c40102.  This is okay, except that writes to the
INPUT_MODE bits register 0x02c40401 have a side effect of changing some
of the settings required for SIF on AUDIO8: VGA_SEL_CH3, DROOP_COMP_CH3,
CLAMP_EN_CH3, etc, according to Section 3.4.2.1 pages 3-17,3-18 of the
CX25843 datasheet.  From what I can see, the driver makes no attempt to
adjust auto-configured registers after writing to registers
0x02c40400-0x02c40403, except in the function cx18_av_initialize().  So
I don't think that will be sufficient to preserve proper demodulation of
SIF on AUDIO8, if changing from S-Video back to Tuner.


6.  v4l2-ctl --log-status reports an autodetected standard of "BTSC"
when that status register at 0x02c40805 is reporting "0 - not defined"
as opposed to "0xF -BTSC" and the microcontroller hasn't bothered to
unmute any of the audio paths.  Reporting BTSC in this situation seems
misleading, as it doesn't appear that the microcontroller has determined
the standard yet.


-Andy


_______________________________________________
ivtv-devel mailing list
[email protected]
http://ivtvdriver.org/mailman/listinfo/ivtv-devel

Reply via email to