I'm not sure about this example but keep in mind the traditional resolution measurements were actually based on horizontal resolution and not vertical. You would find the horizontal resolution of a monitor for example and take 3/4th of that to come up with a number. IIRC this was to take into account the 4x3 nature of the image so taking 3/4th of the horizontal resolution would give you the total number of visible black to white transitions, resolution, that would occur in a length that was equivalent to the height of the picture. This is why there were cameras and monitors in the SD days that had resolutions far greater than the 486 lines possible vertically. The various Pal and NTSC scan rates required non square pixel math to get things to mesh. IIRC NTSC is taller than wide by a factor of .9 and Pal is Wider than it is tall by the same .9 factor. It is usually considered cleaner to stretch things horizontally than vertically. That's why in the Avid Graphics class they suggested that it was cleaner to build things in Photoshop at 648x486 rather the 720x540. I've always believed this is because the traditional CRT display technology had a fixed vertical resolution dictated by the scan rate while the horizontal resolution could increase as technology advanced. I figure it's just all the math to fit a square peg into a non square hole.
--- In Avid-L2@yahoogroups.com, Michael Brockington <brocking@...> wrote:
>
> Respectfully, Terry, I have to disagree.
>
> When sampling non-square pixels for NTSC, it implies you have a higher
> sampling frequency in the horizontal direction than if you were sampling
> square pixels.
>
> Imagine you have a sensor that is 1 inch wide. You sample it one line
> at a time from top to bottom, to get your 480 lines -- that part doesn't
> change between square and non-square pixels. For each individual line,
> if you take 720 samples, each sample will be 1/720th of an inch apart
> horizontally. If you take 640 samples, they will be more widely spaced
> -- each 1/640th of an inch apart -- you will be sampling that line of
> the sensor at a lower resolution horizontally.
>
> So in this example, which I have made up to make the math easy, the
> following is true:
>
> In the case of square pixels, the distance between horizontal
> samples is identical to the distance between vertical lines - 1/640th of
> an inch in both directions.
>
> In the case of non-square pixels, the distance between samples will
> be 1/720th horizontally, but 1/640th between vertical lines.
>
> Regards,
> --Michael
>
> On 12-01-25 9:01 PM, Terence Curren wrote:
> >
> > You folks don't seem to grok square pixels vs. non square pixels.
> >
> > http://en.wikipedia.org/wiki/Pixel_aspect_ratio
> >
> > And here you can see the ATSC standards for SD, both square and non
> > square:
> >
> > http://en.wikipedia.org/wiki/ATSC_standards
> >
> > --- In Avid-L2@yahoogroups.com <mailto:Avid-L2%40yahoogroups.com>,
> > Michael Brockington <brocking@> wrote:
> > >
> > > I'm with Steve on this one.
> > >
> > > 720x486 = 349,920 pixels
> > > 640x480 = 307,200 pixels
> > >
> > > They might both be SD technically, but the first one has about 14% more
> > > resolution.
> > >
> > > Cheers,
> > > --Michael
> > >
> > > On 12-01-25 3:14 PM, Mark Spano wrote:
> > > >
> > > > Steve,
> > > >
> > > > 480 lines of resolution is SD. The math there is that .9 ratio
> > times 720
> > > > equals 648 but that is representative over 486 lines. 640 is the
> > best math
> > > > for 4x3 to get 480. (640 *3 / 4 = 480). And we're talking about square
> > > > pixels, so that math works out perfectly.
> > > >
> > > > The artful look of the square is the best reason I've ever heard why
> > > > anyone
> > > > would shoot this way. But possibly the fear of downconversion
> > softness is
> > > > why they shot SD for an SD spot.
> > > >
> > > > good luck...
> > > > -Mark
> > > >
> > > > On Wed, Jan 25, 2012 at 6:08 PM, Steve Hullfish
> > > > <steve4lists@ <mailto:steve4lists%40veralith.com>>wrote:
> > > >
> > > > > I don't buy that. I mean I know it's CLOSE to real SD, but...
> > > > >
> > > > > The difference between square pixels and 601 pixels is .9 right?
> > The
> > > > math
> > > > > (720x.9) works out to slighly more than 640. And 480 is a little
> > shy
> > > > of 486.
> > > > >
> > > > > Avid used to claim 640x 480 was broadcast SD, but it's really
> > 720x486.
> > > > > With 720x480 being DV and DVD.
> > > > >
> > > > > So it's CLOSE to SD, but I say it's NOT SD. The bigger question
> > is WHY
> > > > > shoot 4:3 SD when you have a 16:9 camera? For the artful look of the
> > > > > square? :-)
> > > > >
> > > > > Steve Hullfish
> > > > > contributor: www.provideocoalition.com
> > > > > author: "The Art and Technique of Digital Color Correction"
> > > > >
> > > > > On Jan 25, 2012, at 4:38 PM, Terence Curren wrote:
> > > > >
> > > > > >
> > > > > >
> > > > > > --- In Avid-L2@yahoogroups.com
> > <mailto:Avid-L2%40yahoogroups.com> <mailto:Avid-L2%40yahoogroups.com>,
> > > > Steve Hullfish <steve4lists@> wrote:
> > > > > >
> > > > > > <<it has a 640x480 setting. So that's actually LESS than SD.>>
> > > > > >
> > > > > > Not actually. If it's square pixels that is SD.
> > > > > >
> > > > > >
> > > > >
> > > > >
> > > > >
> > > > > [Non-text portions of this message have been removed]
> > > > >
> > > > >
> > > > >
> > > > > ------------------------------------
> > > > >
> > > > > Search the official Complete Avid-L archives at:
> > > > > http://archives.bengrosser.com/avid/
> > > > > Yahoo! Groups Links
> > > > >
> > > > >
> > > > >
> > > > >
> > > >
> > > > [Non-text portions of this message have been removed]
> > > >
> > > >
> > >
> > >
> > > [Non-text portions of this message have been removed]
> > >
> >
> >
>
>
> [Non-text portions of this message have been removed]
>
No comments:
Post a Comment