Slashdot Mirror
The First High-Definition TV, Circa 1958
An anonymous reader sends us to Gizmag for a look at a recent auction of a large collection of antique TVs. The star of the show was the Teleavia type P111, one of the earliest examples of high-definition TV. This rare 1958 console-stand television was designed by Flaminio Bertroni, who was also responsible for the iconic Citroen DS. The TV featured dual resolution capability, with the higher setting offering better resolution than 720p — 819 lines. This early attempt at a high-def standard, originating in France in 1949, didn't catch on in the marketplace.
Just look closely at the fine kerning!
. .
The TV featured dual resolution capability, with the higher setting offering better resolution than 720p â" 819 lines.
Nice try, but "by today's standards, it could be called 737i with a maximum theoretical resolution of 816x737 pixels with a 4:3 aspect ratio (10Mhz * 40.8 / 1000 *2 = 816)" Now compare this to the 720p standard which is 1280x720 pixels and a much higher resolution.
Ask Slashdot: Where bad ideas meet poor googling skills.
Computer displays are the same way. Twelve years ago I had a vertical resolution of 1200px in a 21" monitor. Today on a 24" monitor, that's still the best sold in any store. It's sickening.
It's not that hard to do high-definition monochrome TV. You just need to crank up the horizontal sweep rate and use higher-bandwidth amplifiers. Color, though, requires more holes in the shadow mask or stripes on the screen, and the alignment tolerances are tighter.
France had 819-line monochrome broadcast TV in the 1950s. But with the transition to color around 1960, Europe went to a uniform 625 lines. Kind of sad, but making special color TV tubes for France just wasn't worth the trouble.
The screens in the black and white tubes didn't limit resolution, but the spots size (focus) of the beam could. In practice that's mainly a problem with very small screens and high brightness levels, as seen with c.r.t.s in projection sets. Those sure could look awful...
In practice the resolution from left to right is limited by the video bandwidth. On a high end analog computer monitor that may exceed 100 MHz. That essentially limits the minimum width of vertical lines.
But unlike the case with analog computer monitors where stored digital pixel information has a corresponding fixed position on a line, a true analog signal can have intensity changes occur anywhere along the line. To approximate that digitally would take a minimum of two pixels being averaged. (It's the same theory that dictates using at least 40 KHz sampling to sample 20 KHz audio). Trying to use too few of digital pixels (sub-sampling) is what causes aliasing (the jaggies). Analog tv does have that problem, but only in the vertical direction due to the fixed line count/position.
In an analog television, the bandwidth is limited not by the video amplifier section, but by the "i.f." intermediate frequency strip of filters/amplification. By mixing the incoming signals with an adjustable internal oscillator, the tv tuner shifts the desired channel down to the intermediate frequency, there the i.f. filters pass the desired signal while attenuating that of the adjacent channels. That design approach avoids the need to retune a whole group of filters just to change channels. (When first done with A.M. radios, the breakthrough was called SuperHetrodyne) To get higher horizontal detail requires wider filters, and tv channels spaced more widely (greater spectrum bandwidth). The use of too much spectrum was the main limiting factor in preventing opting for higher quality analog. Also, a wider channel means more noise bandwidth (more is captured), so higher resolution would require increased transmitter power to get the desired signal to noise ratio (not notice snow).
The U.S. system used A.M. transmission, but with only part of the lower sideband transmitted in order to save bandwidth. Normal A.M. sidebands are mirror images of each other. With that redundant carrying of information, one sideband could actually be eliminated (you've heard of S.S.B. or single-sideband), but that was too big of a feat to be viable when tv standards were set. The compromise of vestigal sideband gave U.S. black and white tv slightly less than 4.5 MHz of bandwidth out of a 6 Mhz channel. The sound signal (F.M.) was placed 4.5 MHz up from the visual carrier frequency, so the usable video spectrum could extend quite that far. As with single-sideband, putting the same sideband transmission power as A.M. into a narrow channel reduces noise, so coverage is improved.
N.T.S.C. color stuffs additional information into the spectrum used by black and white. Because of the horizontal (line) scan rate being a samping rate of sorts, the video bands exist in clusters spaced that rate (15.750 Khz for B&W, changed to 15.734 Khz for color) occupying spectrum like the teeth of a comb. The added color information centers on a frequency 3.579545 MHz above the video carrier, a choice which causes the sidebands created by the color information to have a comb=like spectrum with the peaks falling right between those of the black and white. If you every had someone trying to sell you a tv that used comb filtering, maybe now you can almost understand why that was a good thing. It allowed recovering as much as possible of the detail present in both the black and white and color parts of the signal while minimizing interferrence effects between them. On old black and white tvs with pretty good signal bandwidth one could actually see a pattern in the parts of the picture where there was bright color content. It looked sort of like regularly spaced lighter/darker dots from left to right on each line. But the choice of frequencies/spacing was such that al