Re: 1080p...and why did I?


"Robin" <me@xxxxxxxx> wrote in message
news:4d9ab31jbqhmb4nh2cv4bffr8g8m505e8r@xxxxxxxxxx
On Fri, 03 Aug 2007 20:00:05 -0400, Some Other Guy
<bgates@xxxxxxxxxxxxx> wrote:

Ward Abbott wrote:
I am setting here wondering why I bought a 1080p. It is
the Sharp
52" and the picture is breathtaking....stunning. But
now I find
that NOTHING is broadcast.....nothing available in 1080p.
Why did I
buy that?

This might make you feel better:

http://www.hometheaterhifi.com/volume_14_1/feature-article-1080p-3-2007-part-1.html


If I read correctly this article claims that when scaling up
low
resolution sources, the higher the target resolution the
better the
result.
In other words when scaling up 480p source, it will be better
if
scaled up to 1080p rather than to 720p.
And in support of this it gives some line drawings.

I find this claim rather incredible.
I don't pretend to know the math or the algorithm behind the
upscaling.

One point of the article is that there is more than one
algorithm - not mentioned is the relative cost of licensing any
of them other than by implication, stating that one is better
than others.

Math is neither the authors' nor the editor's forte or none of
them would have suggested that 1 part of a 10x10 array is
1/10th.

And the somewhat snipe-ish style of the authors reduces the
"authority" of their article to that of a newsgroup exchange of
"facts." ;-0)

But from common sense, if you scale up 480p source
to 1080p, then you have to conjure up a lot more "missing"
data
to fill than you would with 720p target.
No matter how sophisticated the scaling algorithm is, doesn't
this
introduce more artifacts for 1080p target?


"Common sense" and math are not always obviously related.
Interpolation of fractional differences may be better or worse
as the distance between points to be interpolated increases and
the resolution of the interpolated point coincides with a
display point. The more display points available, the better
the interpolation can be represented.

The article states that MOST people wouldn't know good video
from bad. I can only assume that the aim of the authors, by
avoiding algorithm specifics and using overly simple line
drawings, was to somewhat rectify that. Whether it failed
totally, or merely miserably, could be judged by now sampling
the audience of know-nothings who read it.

Even in conclusion the authors waffle; muddling fact, opinion,
and concession in two brief but befuddling paragraphs. There
must be politics in their future!


.

Relevant Pages

  • Re: New guy trying to avoid the mistakes of others
    ... math does make a lot of sense for compression algorithms. ... Quantizer design comes to my mind - one often simplifies the design by considering infinitely extended probability densities, - and even *densities* - that is, continuous signals. ... I use the sum as a tool to come to an algorithm, ...
    (comp.compression)
  • Re: New guy trying to avoid the mistakes of others
    ... math does make a lot of sense for compression algorithms. ... one doesn't really need to invoke math to know this though, because common-sense reasoning should be sufficient to show that magic compression doesn't work. ... The algorithm may then make use of such results. ...
    (comp.compression)
  • Re: New guy trying to avoid the mistakes of others
    ... math does make a lot of sense for compression algorithms. ... The algorithm may then make use of such results. ... into Mathematica, for example, I see a lot of things represented you ... but that you "need math" to design them and design good algorithms. ...
    (comp.compression)
  • Re: Confusion about splitting classes to allow sharing of resources
    ... Then the Spline would ... actual interpolation to the algorithm de jour one set of points at a time. ... regeneration behavior since it is the only one that uses the data. ...
    (comp.object)
  • Re: Sapient Fridge
    ... particular target number e.g. ... that it showed that evolution itself "locked". ... and should be reasonably amenable to a genetic algorithm, ... At that point Spinny declares that my program doesn't work because ...
    (talk.origins)