> Date: Tue, 28 Feb 2006 22:27:43 -0600
> From: Michael Halcrow <[EMAIL PROTECTED]>
> On Wed, Mar 01, 2006 at 05:03:08AM +0100, J.O. Aho wrote:
> > On Tue, 28 Feb 2006, Michael Halcrow wrote:
> >=20
> > >On Thu, Feb 23, 2006 at 11:06:29PM -0600, Michael Halcrow wrote:
> > >>Every so often, I will get a blip that looks something like this
> > >>(image from a digital camera, so it looks pretty bad; only pay
> > >>attention to the artifact in the bottom quarter of the screen):
> > >>
> > >>http://halcrow.us/blip.jpg
> > >
> > >This is happening when I cat /dev/video0 > file.mpg too.
> > >
> > >Anyone have any suggestions for approaching this problem? Should I
> > >mess with PCI bios settings? Try another card? Try another driver? Has
> > >*anyone* here seen this before? These blips really are driving me
> > >nuts.
> > I had something that looked similar when using
> > gentoo-sources-2.6.15-r1 (seemed to be quite bad kernel as it
> > crashed when using rmmod), but on the top of the screen.
> I am running a custom built vanilla 2.6.15.4 kernel...
> > I did manage to get this to work okey if I did select the channel
> > twice with ivtvctl, not sure if this will help for you.
> I'm using an S-Video input. The problem also manifests w/ a Composite
> input from the same video source.
> I'm starting to suspect that this is the result of a glitchy signal
> coming from my Dish 311 satellite receiver; maybe it's throwing off
> the PVR-250's ability to synchronize the signal every once in a while?
Does it sound like it might be related to this problem? No one
replied when I reported this a couple weeks ago, but it's looking
like the 250 (at least) is having problems recovering from a single
sync problem for the rest of the recording. (My glitch doesn't look
like yours---it recurs regularly, it hits the top 1/4 of the screen,
and it looks more like a whole bunch of regularly-spaces squares,
rather a simple displacement like yours, but I wonder if it's really
the same phenomenon...
- - - Begin forwarded message - - -
Date: Wed, 8 Feb 2006 04:47:08 -0500 (EST)
From: [EMAIL PROTECTED]
To: [email protected], [email protected]
Subject: Momentary sync loss trashes recording forever---peculiarly
I've just observed some extremely bizarre behavior. Does anyone have
a clue or (even better) a workaround? This is the second time I've
seen something like this in as many days.
I'm running ivtv 0.4.1-r1 in a Myth backend that's got 5 PVR-250's in
it. I was recording the same RF signal twice---once as four 30-min
episodes, and once as one 2-hour "episode", from a PBS affiliate that
sometimes has a marginal signal. Thus, the first two tuners of the 5
were recording the same thing. The other 3 were idle. I was -also-
recording the same signal on a TiVo.
The PBS feed glitched in a way that looks like the transmitter lost
video sync momentarily; it faded back over the course of a couple of
seconds, with distorted color for the next few seconds.
-One- of the PVR-250's signals (the one recording a 2-hour segment)
was destroyed from then until the end of the recording (about 30
minutes). The upper 1/4 of the video had a periodic (about 1 Hz)
diagonal pattern of dots that flashed on for a frame or two. The dots
were fairly large and blocky---superpixels. Maybe VBI data that was
being "smeared" digitally into the signal? It also had an audible
pop in the audio.
The other PVR-250 did -not- have this problem---it recovered from the
sync loss and continued to give good video. (Unfortunately, that's
the one that was capturing individual episodes---but the PBS affiliate
has crappy scheduling late at night and most of the episodes were
offset a few minutes from when the Myth started & stopped each one.)
[I don't recall right now if it popped, but I could find out if anyone
thinks it matters.]
The TiVo also observed the glitch; its audio continued uninterrupted,
but its video froze for 1-2 seconds and then jumped discontinuously
(fairly standard behavior when it rides out a loss of video sync).
My recording from this PBS affiliate was similarly destroyed last
night, but last night, I wasn't recording on multiple Myth tuners,
nor did I have TiVo coverage, so I thought the transmitting station
had some sort of dying digital VTR. Now I know better---it's an ivtv
issue.
Do later versions of ivtv have fixes that should prevent this from
happening? Has anyone seen this kind of behavior before? I'm
reluctant to upgrade unless someone thinks this has been actually
fixed, since (except for this) I have a "working" and stable system.
And if I do upgrade, what's the recommended version? 0.4.2? 0.5.x?
Thanks!
P.S. Note that I -am- capturing closed-captioning data on all tuners,
in case it matters. (US NTSC cable feed.) I'm doing this by running
"ivtvctl -b wss,cc -x 1 -d /dev/video0", etc, on each tuner just
before Myth's backend starts.
- - - End forwarded message - - -
_______________________________________________
ivtv-users mailing list
[email protected]
http://ivtvdriver.org/mailman/listinfo/ivtv-users
