[mythtv-users] playback not smooth regardless of source

[Seth Daniel]

On Mon, Feb 26, 2007 at 09:37:29AM +0000, Mark Kendall wrote:
> On 2/26/07, Seth Daniel <mythtv.org at sethdaniel.org> wrote:
> > Hello,
> > I have searched this mailing list up and down looking for a workable
> > solution to my playback problems.  No matter what I playback I get
> > jitters or stutters (not sure what the difference is).
> 
> Can you provide some logs  from:-
> 
> mythtv -v playback

Certainly.  Here is what it looks like:

[...]

2007-02-26 22:10:18.257 NVP: 2000 interlaced frames seen.
'video_output' mean = '33358.17', std. dev. = '1532.92', fps = '29.98'
'video_output' mean = '33360.87', std. dev. = '1493.42', fps = '29.98'
'video_output' mean = '33387.94', std. dev. = '1571.79', fps = '29.95'
'video_output' mean = '33355.04', std. dev. = '1429.20', fps = '29.98'
2007-02-26 22:10:31.605 NVP: 2400 interlaced frames seen.
'video_output' mean = '33357.72', std. dev. = '1495.36', fps = '29.98'
'video_output' mean = '33357.09', std. dev. = '1432.55', fps = '29.98'
'video_output' mean = '33359.59', std. dev. = '1443.98', fps = '29.98'
'video_output' mean = '33360.24', std. dev. = '1470.51', fps = '29.98'
2007-02-26 22:10:44.950 NVP: 2800 interlaced frames seen.
'video_output' mean = '33359.68', std. dev. = '1447.85', fps = '29.98'
'video_output' mean = '33389.30', std. dev. = '1493.57', fps = '29.95'
'video_output' mean = '33354.28', std. dev. = '1404.68', fps = '29.98'
'video_output' mean = '33356.44', std. dev. = '1454.28', fps = '29.98'
2007-02-26 22:10:58.298 NVP: 3200 interlaced frames seen.
'video_output' mean = '33358.40', std. dev. = '1482.65', fps = '29.98'
'video_output' mean = '33359.01', std. dev. = '1716.96', fps = '29.98'
'video_output' mean = '33359.29', std. dev. = '1498.63', fps = '29.98'
'video_output' mean = '33360.92', std. dev. = '1430.63', fps = '29.98'
2007-02-26 22:11:11.642 NVP: 3600 interlaced frames seen.
'video_output' mean = '33386.85', std. dev. = '1472.15', fps = '29.95'
'video_output' mean = '33356.66', std. dev. = '1444.67', fps = '29.98'
'video_output' mean = '33356.54', std. dev. = '1474.86', fps = '29.98'
'video_output' mean = '33357.27', std. dev. = '1502.69', fps = '29.98'
2007-02-26 22:11:24.990 NVP: 4000 interlaced frames seen.
'video_output' mean = '33359.21', std. dev. = '1653.13', fps = '29.98'
'video_output' mean = '33359.82', std. dev. = '1453.20', fps = '29.98'
'video_output' mean = '33360.66', std. dev. = '1497.56', fps = '29.98'
'video_output' mean = '33389.63', std. dev. = '1695.92', fps = '29.95'
2007-02-26 22:11:38.335 NVP: 4400 interlaced frames seen.
'video_output' mean = '33352.73', std. dev. = '1466.74', fps = '29.98'
'video_output' mean = '33358.24', std. dev. = '1410.70', fps = '29.98'
'video_output' mean = '33358.90', std. dev. = '1456.11', fps = '29.98'
'video_output' mean = '33357.29', std. dev. = '1465.41', fps = '29.98'
2007-02-26 22:11:51.682 NVP: 4800 interlaced frames seen.
'video_output' mean = '33359.73', std. dev. = '1622.67', fps = '29.98'
'video_output' mean = '33360.81', std. dev. = '1489.99', fps = '29.98'
'video_output' mean = '33386.09', std. dev. = '1484.03', fps = '29.95'
'video_output' mean = '33356.85', std. dev. = '1554.74', fps = '29.98'
2007-02-26 22:12:05.030 NVP: 5200 interlaced frames seen.
'video_output' mean = '33357.82', std. dev. = '1454.10', fps = '29.98'
'video_output' mean = '33357.81', std. dev. = '1727.32', fps = '29.98'
'video_output' mean = '33358.65', std. dev. = '1618.32', fps = '29.98'
'video_output' mean = '33359.34', std. dev. = '1504.26', fps = '29.98'
2007-02-26 22:12:18.375 NVP: 5600 interlaced frames seen.
'video_output' mean = '33361.24', std. dev. = '1607.24', fps = '29.97'
'video_output' mean = '33389.36', std. dev. = '1520.77', fps = '29.95'
'video_output' mean = '33353.55', std. dev. = '1472.60', fps = '29.98'
'video_output' mean = '33036.85', std. dev. = '2794.12', fps = '30.27'

[...]


It continues on like this.  During this time there was subtle, yet
obvious, stuttering.  This time I was using the Kernel deinterlacer.

I finally hooked up my HDHomeRun.  Using the kernel deinterlacer it
looks pretty good.  It even seems more smooth than SD when you consider
the stuttering (although the HD seemed to stutter at times too, but
seemed to be less frequently).  If I hadn't mentioned earlier I am using
a PVR-150 to record SD.  My ivtv drivers are v0.9.0.

When playing HD with Bob, oh man.  That was not pretty.  The CPU cores
weren't any higher than when using Kernel deint (1 is between 50-60% and
the other hovers around 35%), but the video was extremely jumpy,
jittery, stuttery.  You name it.  Kernel deint was much better (but
there was still some jitter).

-- 
seth / @sethdaniel.org
Immortality...a fate worse than death.