Mailing List Archive

http://www.mythtvtalk.com/forum/ivtv-issues/10841-pvr-150-mythtv-capture-quality-issues-picture-included.html#post44262


>Well the failure mode looks somewhat obvious: the digitizer (like at
>CX25843) is not getting set up right or it's being reverted to the
>autoconfig defaults-.
>All the analog front end gain adjustment and analog to digital
>conversion happens at the digitizer chip. The black with white pixels
>lines at the top are Vertical Blanking Interval test signals and data
>services (like Closed Captioning). Either the digitizer isn't detecting
>vertical sync properly or the digitizer register for the number of lines
>of vertical blanking isn't being filled properly. The digitizer has
>other filters and gain controls that can be applied to an analog signal.
>I2C bus errors/failures could cause registers in the sigitizer not to be
>programmed properly.
>When the machine is realtively quiescent, kill the mythbackend, use
>modprobe -r ivtv cx25840 to unload ivtv and the cx28540 (or whatever
>digitizer you have) modules. Then modprobe ivtv and see if things are
>better.
>Regards,
>Andy

Hmm.. I tried the commands with the CX25840 that you mentioned, and
it seems to have no effect. One thing worth noting is that when I run
the lspci command, I get the PVR-150 listing and it has a number in
parenthesis that looks similar to what you mentioned (CX23416). I
tried substituting your number for my number and modprobe reported
that the CX32416 module is not found. I tried your number and the
command was executed with no errors. Then I executed the "modprobe
ivtv" command and that was also executed successfully. I then open up
MythTV and still see the same lines.

Is the digitizer the problem for both the overscan and the green
shadow? If the problem is only for the overscan, I can ignore it as I
have found a setting in MythTV that allows me to hide the overscan.
The more important problem is the green shadow. I have pinpointed it
down to the tv tuner; I tried several connection configurations that
test where the green shadow is apparent(S-Video, Composite both have
no problems; anything connected through coaxial has problems, like
regular cable and even antenna).

If you need more information, I will be happy to supply it.

_______________________________________________
ivtv-users mailing list
ivtv-users@ivtvdriver.org
http://ivtvdriver.org/mailman/listinfo/ivtv-users

On Sat, 2009-02-21 at 12:09 -0500, WSD33904 wrote:
> http://www.mythtvtalk.com/forum/ivtv-issues/10841-pvr-150-mythtv-capture-quality-issues-picture-included.html#post44262
>
>
> >Well the failure mode looks somewhat obvious: the digitizer (like at
> >CX25843) is not getting set up right or it's being reverted to the
> >autoconfig defaults-.
> >All the analog front end gain adjustment and analog to digital
> >conversion happens at the digitizer chip. The black with white pixels
> >lines at the top are Vertical Blanking Interval test signals and data
> >services (like Closed Captioning). Either the digitizer isn't detecting
> >vertical sync properly or the digitizer register for the number of lines
> >of vertical blanking isn't being filled properly. The digitizer has
> >other filters and gain controls that can be applied to an analog signal.
> >I2C bus errors/failures could cause registers in the sigitizer not to be
> >programmed properly.
> >When the machine is realtively quiescent, kill the mythbackend, use
> >modprobe -r ivtv cx25840 to unload ivtv and the cx28540 (or whatever
> >digitizer you have) modules. Then modprobe ivtv and see if things are
> >better.
> >Regards,
> >Andy
>
> Hmm.. I tried the commands with the CX25840 that you mentioned, and
> it seems to have no effect.

Sorry, I've been busy (I still am), and haven't replied.

Please provide the output of

v4l2-ctl --log-status

so I can see what ancillary chips your card has.


> One thing worth noting is that when I run
> the lspci command, I get the PVR-150 listing and it has a number in
> parenthesis that looks similar to what you mentioned (CX23416). I
> tried substituting your number for my number and modprobe reported
> that the CX32416 module is not found. I tried your number and the
> command was executed with no errors. Then I executed the "modprobe
> ivtv" command and that was also executed successfully. I then open up
> MythTV and still see the same lines.

The PVR-150 uses the CX23416 PCI host/MPEG encoder chip which is
supported by the ivtv, along with the cx2341x helper module. The
digitizer chip is a chip on the card connected to the CX23416 via an
on-card I2C bus between the CX23416 and the digitizer chip - it won't
show up in lspci. The ivtv module is smart enough to load all the I2C
bus drivers and digitizer chip driver all on it's own. Sometimes PCI
bus or I2C bus interactions, or old i2c modules, don't work as expected.

First let's figure out what your card has with the log status output and
go from there.


> Is the digitizer the problem for both the overscan and the green
> shadow?

Yes and most likely. It could also be the analog tuner assembly is
giving you a weird signal into the digitizer, for whatever reason. The
tuner is also controled via I2C bus from the CX23416.


> If the problem is only for the overscan, I can ignore it as I
> have found a setting in MythTV that allows me to hide the overscan.
> The more important problem is the green shadow.


Vertical blanking interval is what you see at the top of the screen.

The digitizer is coming out of vertical blank too soon and digitizing
veritcal blanking data that you then see on the screen. It's indicative
of the digitzer not being set up properly or a bad video signal.


> I have pinpointed it
> down to the tv tuner; I tried several connection configurations that
> test where the green shadow is apparent(S-Video, Composite both have
> no problems; anything connected through coaxial has problems, like
> regular cable and even antenna).

So can you clarify/confirm, you are running the same video standard
(e.g. NTSC-M is the US standard) on both the tuner and the line in? Or
is one of them using something different?


> If you need more information, I will be happy to supply it.

Regards,
Andy


_______________________________________________
ivtv-users mailing list
ivtv-users@ivtvdriver.org
http://ivtvdriver.org/mailman/listinfo/ivtv-users