Happy Birthday 1953NTSC Color :)

[earlyfilm]

Originally Posted by old_tv_nut

Much too late to be a patent on basic combination of luminance and color subcarrier. . . . . . . similar to the way NTSC chroma and luma are interleaved.

Philips claims priority, with application in the Netherlands on May 1, 1952 and US patent applied for April 28, 1953 and this patent took 5 years to be approved in the US.

I suspect that there originally more claims and this potential interference patent could be the reason that RCA dropped their obviously superior I & Q system like a hot potato!

Jas.

[old_tv_nut]

Interesting that this might be construed to have priority - it put the color difference signals on different subcarriers instead of a common subcarrier. However, it did use unequal sidebands, and that might be the interference.

[Pete Deksnis]

Originally Posted by earlyfilm

I suspect that there originally more claims and this potential interference patent could be the reason that RCA dropped their obviously superior I & Q system like a hot potato!

Jas.

More speculation...

It's true that altering the CT-100's CTC2 chassis I/Q demodulation to R-Y/Q demodulation in the CTC2B chassis "saved one resistor in the matrix," as a retired RCA exec retorted on the internet some years ago.

It was then soon speculated that patent infringement had a hand in the switch.

Certainly the latter is the finer argument, as the quadrature transformer in the CTC2 chassis was replaced with another part number in the CTC2B. The reason: I/Q demodulation is in quadrature and R-Y/Q demodulation is not. Hence the need to go to the added expense of replacing a true quadrature transformer with one that isn't.

The really interesting question becomes: is there really any significant cross-color resuling from demodulation in the 21-CT-55 versus the CTC2?

I suspect very little.

Pete

[old_tv_nut]

Originally Posted by Pete Deksnis

...
The really interesting question becomes: is there really any significant cross-color resuling from demodulation in the 21-CT-55 versus the CTC2?

I suspect very little.

Pete

Actually, moving the I axis to R-Y would produce no crosstalk. The Q signal is narrow band, and the I signal has both low frequency sidebands (double) and high frequency (lower sideband only).

Since the wideband R-Y demodulator is properly matrixed to form the R signal when combined with the Y signal, it gets the correct proportions of I and Q low frequencies (in the double sideband region below +/- 500 kHz).

What does the R-Y demodulator get in the high-frequency I region? Because it is not exactly in phase with the I signal, it gets mostly the correct I high frequencies; but it also gets a small proportion of the Hilbert transform of the I high frequencies (that is, I high frequencies with a 90 degree phase shift). The result is that the R-Y shows some distortion of high frequency transients that would be perfect in a true I demodulator.

sin (33) = .545 (pretty large) - multiplier for the undesired Hilbert component.
cos (33) = .839 - multiplier for the desired I signal, which also is the correct multiplier for the I component of R-Y.

Wikipedia has an illustration of a square wave and its Hilbert transform:

http://en.wikipedia.org/wiki/Hilbert_transform

If the original I signal is a square wave of frequency greater than 500 kHz , what comes out of the R-Y demodulator is a sum of the square wave and a smaller proportion of the shown Hilbert transform. So, rising edges get pushed towards cyan, and falling edges get pushed towards red.

Whether this is noticeable or not will depend on the subject matter and also on whether the viewer knows what to expect. In most cases, these kinds of distortions do not become strongly apparent unless you have a correct picture to compare to. There might be exceptions for things like colored title letters, which the viewer may assume to originally have uniform color on the left and right edges.

[old_tv_nut]

Here is an article by James O'Neal that appeared in TVTechnology, and now appears to be posted in the clear, with a nice shot of RCA field engineers working on a prototype:

http://www.tvtechnology.com/article/...versary/222948

[Penthode]

The question is really whether the shifting of one demodulation axis was actually noticeable in the reproduced picture. I suspect that in the lab, and with the pressure to reduce the cost of the receiver, trimming the demodulation circuitry was an important option.

When the CTC2B chassis was offered, it was already nearing obsolescence with the simplified CTC4 already underway. The other distortions in the production and delivery chain in 1955 surely must have offset any noticeable distortion in the displayed picture color transitions.

The original I/Q quadrature modulation scheme was only a means to try and maintain some sort of a minimum chrominance bandwidth which at the end of the day I imagine in practice was not really noticed when it was shaved off.

Originally Posted by old_tv_nut

...Whether this is noticeable or not will depend on the subject matter and also on whether the viewer knows what to expect. In most cases, these kinds of distortions do not become strongly apparent unless you have a correct picture to compare to. There might be exceptions for things like colored title letters, which the viewer may assume to originally have uniform color on the left and right edges.