[mythtv-users] Correctly behaving 576i (or 480i) from vdpau-capable chip?

Paul Gardiner lists at glidos.net
Tue May 25 10:26:10 UTC 2010


Christopher Kerr wrote:
> On Tue, May 25, 2010 at 5:38 PM, Paul Gardiner <lists at glidos.net> wrote:
>> I can't say with certainty yet because I don't yet have my ION board,
>> but I'm pretty sure that it can because I have several sources and
>> the 600Hz feature makes a striking difference on all of them. The
>> motion becomes fluid in a way that I've never seen before. My girl
>> friend raved about it more than I did: "How is it doing that? It's
>> like being there", and then when I turned it off in an attempt to
>> demonstrate that it was quite smooth even with out 600Hz: "TURN IT
>> BACK ON!!". :-)
>>
>> It's not without a price. It definitely creates artefacts with
>> complex objects moving across a complex background, the object
>> can drag bits of background. The worst is a full screen pan
>> with a semitransparent logo on the screen. The logo can become
>> deformed. Most of the artefacts aren't very noticable unless you
>> are looking for them, and I find them worth putting up with for
>> the sake of the striking smoothness. I imagine others wouldn't
>> though.
> 
> Indeed. Frankly I think it looks gross - much like the current crop of
> 3D movies, it makes the picture more striking at the cost of looking
> wrong. IMHO, it's more of an "effect" than an "improvement".

Hmmm, maybe. Certainly a matter of opinion. Other than the artefacts,
to me it gives the impression of seeing a real scene rather than
a recording of a scene. People move like people rather than blurring
and juddering.

> While the TV may notionally manage a 600hz response time, in reality,
> the DECAY time of images on both LCD and plasma displays is often
> upwards of 40ms. Fortunately, it turns out that we can't actually see
> the resulting motion blur, so it doesn't matter, but it renders the
> issue of high refresh rate screens moot.

That doesn't fit with what I see. The 600Hz mode seems to make moving
objects hold their sharpness and markedly reduces motion blur. And
motion blur is in itself an effect: objects moving at lowish speeds
don't blur in reality; your eyes track them and they stay perfectly
sharp.

> Taking the above into account, then, any differences you CAN see can
> be attributed to the motion processing algorithms applied by the
> display. Since they're interpolating data, it's unlikely that it
> actually looks any better than the original, but if you happen to like
> the effect, you might as well leave it turned on.

Because of the artefacts, I turn it off every now and then. but then
hate the motion blur.

Of course my comparison is with just a few other TVs. It may be that,
rather than the Panasonic being good with 600Hz enabled, it's actually
poorer than most with it disabled. Perhaps other TVs do a better job
at 60Hz. Also I'm watching mostly SD, which although supposedly 50i,
actually has frames who's fields are not separated in time, so I'm
really watching 25p content. Perhaps the 600Hz is advantageous in
that case but wouldn't be for 50p or true 50i content. In the UK
we still have some SD programs that are truly 50i and I could very
much see the difference in smoothness when watching on my CRT.

I remember seeing a whole wall of TVs dealing with a slow pan, and
just one handling it well, that being a Phillips model. Maybe they
do a better job without using anything like this 600Hz mode. I don't
know. I've never had a chance to do a comparison.

> The only thing a 600hz display can do that a 100hz display can't is
> properly display NTSC (30fps) PAL (25fps) and cinema (24fps) content
> without mucking about with quality degrading techniques like telecine
> or nasty hacks like PAL speedup. For that reason, and that alone,
> "600hz" displays will probably become a worldwide standard within a
> few years.

Before I owned the set, that's what I thought it was all about. I did
the 25 x 24 = 600 calculation and thought that must be the point, but
there seems to be more going on than that.

Paul.



More information about the mythtv-users mailing list