brianbwlim
April 7th, 2006, 10:27 PM
Sorry to bother you again.
Unfortunately, I have found another performance shortage of HQV(DPX-1300)
for "Mixed 3:2 with Titles" processing.
Details are here:
http://www.avsforum.com/avs-vb/showthread.php?t=664769
Althohgh, HQV claims that
-----------
HQV algorithms make a film vs. video decision on a per-pixel basis. This is useful, for example, for deinterlacing film that has a video-based text 'crawl' superimposed on top of it (film credits, storm advisories on TV, etc). The film portion is de-interlaced with cadence detection, and the video portion is deinterlaced through motion interpolation.
-----------
My test results tells me something different.
My eyes tells me that they just "bob" the whole frame when titles are running.
For reference, you can see the picture below.
none.jpg - The shape of guitar cords - this is correct image from 3:2 pull down except for running titles.
bob.jpg - This is the same as DPX-1300 image.
http://rapidshare.de/files/17409995/mixed.zip.html
Jason,
Could clarify that VXP do different processing from "bob".
And if so, could explain how it is done.
Best Regards,
Unfortunately, I have found another performance shortage of HQV(DPX-1300)
for "Mixed 3:2 with Titles" processing.
Details are here:
http://www.avsforum.com/avs-vb/showthread.php?t=664769
Althohgh, HQV claims that
-----------
HQV algorithms make a film vs. video decision on a per-pixel basis. This is useful, for example, for deinterlacing film that has a video-based text 'crawl' superimposed on top of it (film credits, storm advisories on TV, etc). The film portion is de-interlaced with cadence detection, and the video portion is deinterlaced through motion interpolation.
-----------
My test results tells me something different.
My eyes tells me that they just "bob" the whole frame when titles are running.
For reference, you can see the picture below.
none.jpg - The shape of guitar cords - this is correct image from 3:2 pull down except for running titles.
bob.jpg - This is the same as DPX-1300 image.
http://rapidshare.de/files/17409995/mixed.zip.html
Jason,
Could clarify that VXP do different processing from "bob".
And if so, could explain how it is done.
Best Regards,