Slashdot Mirror

One thing worth pointing out: something vaguely like this was studied for post h.264 compression. It sounds like the way to go, right? Include time in the compression scheme, don't just look at each frame in isolation. Turns out temporal compression artifacts are quite jarring, and the whole thing had to be abandoned.

So what you're suggesting sounds like something useful for games and other video generated locally and never compressed, but I don't think it would work for recordings.

The relationship between CCD elements and pixels is a whole nother rant. Even in 2d, pixels in an image format never should have been "little squares". It's a really suboptimal way to represent image data (especially since we don't have really displays with little squares that can display any color).

It's important to distinguish between these legacy modes. Composite merges chroma and luma information in a single signal. Component video (s-video) that keeps them separate and is a big quality improvement. If you want little rectangles as pixels, well, it's not clear that was ever the design intent of these games. A pixel is not a little square.

Are you sure that *this* matters? After all, a pixel is not a little square, and by the time you take the perceptual issues into account, your "proper LCD display" turns out to be not such a good idea anyway.

Csuri's work is certainly 3D computer graphics that predates Catmull's film, but it's the particular techniques shown here that make Catmull's film remarkable. Csuri's work from this period (that I've seen) is only rendered points and wireframes without any hidden-surface removal. Catmull shows fully shaded polygons with correct depth ordering (it's likely he used techniques other than Z-buffering to achieve this).

The founders of Pixar are Ed Catmull and Alvy Ray Smith. See: http://alvyray.com/pixar/default.htm . (I would not be surprised if the list of founding employees includes some student's of Csuri's.)

It is probably worth referencing Alvy Ray Smith's (co-founder of Pixar) enlightening memo from 1995 entitled: A Pixel Is Not A Little Square, A Pixel Is Not A Little Square, A Pixel Is Not A Little Square! (And a Voxel is Not a Little Cube)"

Non square pixels are not a new idea, see for example sensors of cameras.

Alvy Ray Smith, computer graphics researcher and co-founder of Pixar, will tell you that A Pixel Is Not A Little Square, A Pixel Is Not A Little Square, A Pixel Is Not A Little Square! (And a Voxel is Not a Little Cube). I don't necessarily agree with everything in the article (e.g. coordinate systems) but it is worth a read.

Non square pixels are not a new idea, see for example sensors of cameras.

Alvy Ray Smith, computer graphics researcher and co-founder of Pixar, will tell you that A Pixel Is Not A Little Square, A Pixel Is Not A Little Square, A Pixel Is Not A Little Square! (And a Voxel is Not a Little Cube). I don't necessarily agree with everything in the article (e.g. coordinate systems) but it is worth a read.

A Pixel Is Not A Little Square, A Pixel Is Not A Little Square, A Pixel Is Not A Little Square! (And a Voxel is Not a Little Cube)

Obligatory paper on the subject of pixels: A Pixel Is Not A Little Square, A Pixel Is Not A Little Square, A Pixel Is Not A Little Square! (And a Voxel is Not a Little Cube).

Repeat after me: A pixel is not a little square. CRT monitors reconstruct an image using something closer to a gaussian distribution, rather than a crisp rectangular one as you'd get if you simply doubled pixels. The graphics of games made when CRTs were common were made on CRTs and thus take advantage of this. The video game console emulation crowd has faced a similar issue, only there it's more than just a CRT; there's also the distortions introduced by the various composite video encoding schemes (color bleed, fringing, artifacts). You might think that removing these distortions would improve the image, but you have to realize that the artists viewed things on the same systems, and thus tailored the art to look good in those circumstances. It's sort of like a web page designer getting a page to look just right in a buggy browser, even though it looks all wrong in one with proper rendering; here you want the buggy browser, at least if you want to see the page as it was intended.

The thing that gets me is that a high-resolution LCD could horizontally display exactly what a Trinitron CRT did, as the vertical stripe phosphor pattern matches that on most LCDs. The scaling algorithm would need to simulate the blurred-edge electron beam and mixing between pixels. There would be some sub-pixel action too, as on a CRT.

Not to mention, it's not like Anti-Aliasing in ray tracing is impossible.

ftp://ftp.alvyray.com/Acrobat/6_Pixel.pdf (from 1995 no less)

You make several good point, especially about the compression.

Basically, 1080i is the broadcast format and 1080p will be reserved for HD DVD / Blueray (and possibly content generated on the fly, like games). The reason for this is the darn compression ratios. There is simply not enough bandwidth (on a digital TV channel) to transfer the 1080p information without compressing the hell out of it, which defeats the point of higher resolutions.

You can read about this at Alvy Ray's website - he's one of the guys that founded Pixar. He covers naming conventions, bandwidth and compression ratios pretty well: http://alvyray.com/DigitalTV/ A very good read, even if you know your HD stuff.

You're right that resolutions are fixed by channel. (E.g., see here). The networks all either chose 720p or 1080i. However, in most areas of the country. ABC is 720p through most of the country, except some areas of Texas. See here

Regardless of the HDTV format being broadcast, all new HDTV receivers can receive both formats. New HDTV televisions will convert any received signal to a format that is compatible with your new display. The 720p format uses progressive scanning, which is just like your computer monitor. Progressive scan offers crystal clear images that virtually eliminates those scanning lines that are visible on most large screen televisions. ABC broadcasts all of its programming using the 720p format except in Dallas, where the ABC station broadcasts in 1080i. Many new flat panel displays use progressive scan

It looks like it may vary per network and not per show?

http://alvyray.com/DigitalTV/Naming_Proposal.htm

I wouldn't be suprised if Lucasfilm bought back Pixar from Jobs since Lucasfilm sold Pixar in 1986. I just hope reclaimed Pixar would work on other project besides Star Wars films.

Well, they did have the inventor of the alpha channel working for them. MS bought out his company Altimira, whose Composer product became the basis of MS's photo apps.
I think they didn't go after the pro market partly because they didn't want to, and mainly because the majority of the pro market is Mac based.

Pixels are not square, or even rectangular, any more than a number in a sound file is an interval of time. A pixel is a point sample, which can be resampled into square areas if needed.

This is all explained in Alvy Ray Smith's technical memo. Read it and consider yourself enlightened.

It's essentially a POINT - it has no dimensions. When you see those little squares you actually see a poor (and fast) representation of pixels - pixels themselves are not square or non-square. Pixels won't come in various sizes, they'll still be regular 0-sized points.

Here's a good paper on why it's important to keep in mind the true nature of pixels (by Alvy Ray Smith):

A Pixel Is Not A Little Square, A Pixel Is Not A Little Square, A Pixel Is Not A Little Square! (And a Voxel is Not a Little Cube)

The definitive reference on the history of alpha is Alvy Ray Smith's technical memo from 1995. It seems pretty clear that he co-invented the technique with Ed Catmull as a solution to a problem that Catmull was having with his sub-pixel hidden surface algorithm.

The earliest dated documentation on his alpha channel code is January 13, 1978, although it was probably written the previous December. See footnote 4, page 6 of the memo for details.

Pixels and voxels are zero-dimensional samples of some 2D image or 3D volume. Thinking about them as squares, gaussian splats, or something other than samples is the path to the Dark Side.

For more info, read Alvy Ray Smith's Tech Memo, "A Pixel is Not a Little Square, a Pixel is Not a Little Square, a Pixel is Not a Little Square! (And a Voxel is Not a Little Cube)" available here.