[mythtv-users] Upscaling standalone DVD vs. MythTV

[Michael T. Dean]

On 02/16/2007 03:41 PM, Craig Partin wrote:
>>>> Yes, but this is only fairly primitive scaling. To compete with hardware
>>>> scalers in high-end DVD players, TVs or receivers one has to employ
>>>> fancy software scaling algorithms which are much more CPU intensive.
>>> Are any of these algorithms in myth currently?
>>>       
>> No.  And, TTBOMK, they're not in Windows Media Player or whatever else
>> was used in the previously-mentioned quality test comparing set-top
>> upscaling DVD players to computer output.
>>     
> So is it better to have myth do the scaling?  Or send the video
> untouched to the TV and let it do the scaling?  I suppose that depends
> on the quality of the TV, or is there more to it?

I think Myth/Xv does a better job of scaling than the ATI Xilleon 220 in 
my TV.  Most HDTV's in use today have one of only a couple of video 
processing chips (and, last I checked, the ATI Xilleon was the most 
common--at least in the US).

I also find that the Xilleon has trouble keeping up with the real-time 
requirements of video playback--even though it's a dedicated video 
processing SOC that's only processing the video I'm playing back--giving 
the occasional (i.e. once per hour or so) blockiness and missed frames 
(even at 480p--as on a DVD).  Basically, it looks like it's trying to do 
too much "fancy" stuff (like the TV's stupid DNIe--"Digital Natural 
Image engine"--color "correction").  My Myth box does much better...  Of 
course, if you can't send your Myth output to your TV such that the TV's 
video processor doesn't touch it (which depends on the TV's inputs and 
wiring), the point is moot.

Of course, for such a subjective comparison (quality of output), others 
may have different opinions.

Mike