RCA 21CT55 B8802897 is Alive!

[zenithfan1]

The exact chassis number that uses I and Q is the CTC-121, mine's from '84 and has a beautiful picture.

[old_tv_nut]

I just ran some pictures of roses against dark green foliage through the simulation to show what the extra I bandwidth does in this case:


Original full bandwidth RGB:



Wideband I (and narrow Q, of course):


Narrowband I and Q (equivalent to R-Y/B-Y or X, Z or any other narrow equiband demodulation):


the wideband I:


the narrow band I:


the Q:

[Pete Deksnis]

On my little Eee netbook, the loss of detail in the roses is subtle but real; neat simulation.

[old_tv_nut]

What you should do is save the images so you can use your favorite viewer, then zoom in until the image is the same size as on a 25 inch diagonal tube. The differences will be much more apparent, just as they would in real viewing.

[old_tv_nut]

Here's a try at posting an animated gif of the differences:



Edit: It worked! Sometimes I amaze myself!

[Tomcomm]

Great simulations Wayne! I know nothing regarding Photoshop Pro, but their filter capabilities must be awesome. My little Photo Impact’s custom filters only effect a 5 X 5 pixel zone, resulting in an overall softening or other distortions. No way can it filter color content of red, blue or green components. Obviously, Photoshop is capable of simulating the color separation of a dicroic filter into their primary colors. Developing I and Q components from the 32bit chroma of each pixel and combining them in quadrature and then bandpass filtering them to produce perfect NTSC baseband bit streams. Is this a canned program within Photoshop or did you develop this NTSC simulation on your own? Whatever, you certainly are an expert in TV system simulations.

When you conducted comparisons of the 1984 RCA solid state I, Q, WB receiver with the current Zenith R-Y, B-Y, NB receivers you noticed improvement in the RCA’s red detail, but also the presence of quad crossover distortion artifacts along color transition edges, The 1954 IRE Color TV issue had a 1953 paper by Hazeltine on the very subject of “Quadrature Cross Talk in NTSC Color TV”. Seems this problem was solved in the fully compliant NTSC WB, IQ implementation of RCA’s first production models., the CT 100 and 21CT55, using the CTC2/B chassis. That same IRE issue described a true laboratory standard color receiver built by Sylvania in 1953 which was fully NTSC compliant and very similar to the RCA CTC2/B WB IQ decoder. It had an additional NB R-Y, B-y decoder similar to the RCA CTC4 thru CTC38 at least. These decoders could be switched to the same 15in CRT display for picture quality comparisons. Regarding the NB R-Y, B-Y , their comment was "Although this method gives the desired result, much detail is lost due to narrow-band operation”. I have to assume if any cross-talk artifacts were present in the lab standatd WB, IQ receiver it would have been well reported.

[old_tv_nut]

The built-in filters in Photoshop are not up to this job. I had to write my own using an authoring program called "Filtermeister." The filters are phase-linear FIR filters designed for the proper I, Q, and chroma trap bandwidth. There is also a sharp cutoff lowpass to simulate the IF amplifier.

The modulation and demodulation is done by assuming that the pixels fall exactly on 57 degrees (I), 147 degrees (Q), 237 degrees (-I), and 327 degrees (-Q). This results in a 720x540 4x3 square-pixel image, close enought to the real NTSC sampled values for illustration.

[Dave A]

The irony of this thread is that all things digital in 2010 are being used to recreate the missing information of the color analog world in 1954. And probably with more information than they had at hand.

Keep it up because we all become more informed and more educated...digital or analog.

Dave A

[jeyurkon]

Originally Posted by old_tv_nut

Here's a try at posting an animated gif of the differences:

...

Edit: It worked! Sometimes I amaze myself!

And us!

[Tomcomm]

Still pondering the relative picture quality of the NTSC spec’ed WB IQ chroma demodulation used in the RCA CTC2/B chassis and the simplified NB R-Y B-Y chroma demodulation used in all other RCA CTVs thru 1980 at least. When did any CTV return to WB IQ if ever! My two premium Diamond-view Mitsubishis of 1988 and 1996 both have comb filters but I can’t identify WB IQ in either. Admittedly I have very few detailed schematics of my Sonys but both 1982 and 1992 jungle chips identify R-Y, B-Y demodulators, no mention of IQ anything. So does anyone know for sure any modern CTVs that for sure use WB IQ processing? My 1988 13in Sony RVM 1344 Pro Monitor must surely use WB IQ demodulators along with it’s comb filter, the direct screen is awesome. We’re there ever fully digital implemented CTV, AD rf inputs thru all digital demods to LCDs before the end of NTSC? How about the high-end ATSC to NTSC converter boxes that outputted component analog video or the best DVD players, did they encode MPEG to NTSC in WB IQ? It seems that NTSC spec’ed WB IQ started and ended with the RCA CTC2/B chassis, or am I wrong……..Tom

[old_tv_nut]

AS stated above, the RCA CTC-121 was the only consumer set that revived IQ demodulation. DVD players, DTV converters, etc, all use equiband modulation, but it may be wider band than the standard Q channel. If you input this to your TV as baseband composite, or especially S-video, you may get some chroma bandwidth improvement. However, the best chroma bandwidth by far comes from using the component Y Pr Pb output of such devices if they have them, and skip the chroma mod/demod process altogether.

No idea if your 1988 13in Sony RVM 1344 Pro Monitor used IQ demods. Tektronix monitors did not. In fact, the smaller screen Tek monitors had a strictly limited luma bandwidth also to prevent moire patterns due to chroma dots beating with the relatively coarse screen stripes. TV station engineers loved them - the pictures looked CLEAN because they also hid any garbage in the upper end of the luma signal.

By the way, I know someone who worked on a super high-res sharp focus NTSC monitor for Tek. The project was dropped when they saw that it made NTSC pictures look awful when you had full-contrast resolution of all the luma artifacts and the scanning lines with black spaces in between.

[stromberg67]

I don't have the service notes at hand at the moment, but I believe my CTC-136B has WB IQ demod, using a chipset by NEC. I'll dig out the folder and take a look when I can.
Thanks for the comparison.
Kevin

[Tomcomm]

So far only the CTC 121 and CTC136B were suspected to have used WB IQ chroma demodulators. Wayne states he evaluated the performance of a RCA CTC13X that had WB IQ demodulators but gave no details of its implementation; did it use a NEC IC? A whole bunch of guys say they have or had either CTC 136Bs or the CTC121s which also is claimed to have WB IQ processing:
CTC136B stromberg67 Kevin
CTC121 zenithfan1 Mark, waltchan, zenith2123, wa2ise, nrx37, blast
Surely at least one of these has a schematic or other documentation that definitely proves the CTC136B or CTC121 used WB IQ and how it was implemented. Since I have a CTC2B open on the bench I would appreciate any information why RCA tried WB IQ on only a couple of CTVs and quickly dropped it. Thanks, Tom

[old_tv_nut]

I don't know which chassis number I looked at. I only ever knew of one with IQ chroma, and I don't know who actually manufactured the IC, but I do recall that some very clever analog design was done by RCA to minimize the pin count.

[vintagecollect]

I did my own test to see improvement in picture using the 3 seperate color outputs from a DVD fed into a digital widescreen projector. Compared single composite video and then three seperate colors, the three colrs image had slightly more depth in color, shading. The detail differance was minimal. High end Proxima PJ w/ full resolution capibilities of a DVD slightly displaying their differences displaying a 15 to 18 ft screen.

NTSC really shows it's full potential to be in the area somewhat near wideband color in quality. The only real way is for the Museum, OR one of the old color TV gurus who can compare a wide band color to narrow band sets side by side from the same vintage.