[theora] Theora player for Nokia Series 60
nlf482n at med.usyd.edu.au
Sun Apr 18 00:12:47 PDT 2010
Quoting "Stuart Fisher" <stuart.fisher at fig7.com>:
> If you're willing, maybe I could get you to run a few more tests?
Yes sure. It's the least I could do :)
> I'm intrigued as to why you're getting so many dropped frames (2)
> and not a single dropped frame (1). The results also look much worse
> than I was would have expected. That's the kind of result I get on
> my Nokia 5800 but in that case the phone is also having to scale the
> image to the size of the screen. The only thing I can think of is
> that your phone is scaling the image for some reason. That shouldn't
> be necessary though as the screen on the E71 is the same size as the
> video. It doesn't make much sense really.
Nokia specs for the phone:
say that the screen dimensions are 320x240 pixels. So doesn't that
mean that my phone is having to rescale the 320x180 clips and black
bands at the top and bottom of the screen to keep proportions correct
is being "inserted" by OgvPlay on the fly? If so, can you please
re-encode for 320x240 for me to try those?
I still stand by my claim that the 320x180 videos are perfectly
watchable even with this drop rate. Maybe I've got low expectations -
the only other videos I've really watched on the phone are ones shot
with the in-built camera, which are pretty attrocious, and a few
videos from youtube, which are equally attrocious. Certainly though,
the two sample (h264?) clips that came with the phone aren't obviously
better than BBB or ED theora clips - granted they don't have any slow,
smooth, large-scale movement to compare. Maybe I need some BBB & ED
clips in optimised h264 for a comparison.
On a related point - I've read some criticism somewhere that BBB isn't
particularly challenging material to use for testing the quality of
video codecs. If that's true, should we be using some different
material, or a range of material?
More information about the theora