21CT55 Screen-Shot Test
 

Attempting to post full monitor screen size on VK Forum postings.

http://i677.photobucket.com/albums/v...1-2123_IMG.jpg[/URL]

http://i677.photobucket.com/albums/v...hots/Lama1.jpg[/URL]


http://i677.photobucket.com/albums/v...ps7e748ef5.jpg[/URL]


[http://i677.photobucket.com/albums/v...licker2010.jpg[/URL]

http://i677.photobucket.com/albums/v...pse3715511.jpg[/URL]

http://i677.photobucket.com/albums/v...psda96737b.jpg[/UR]

http://i677.photobucket.com/albums/v...ps7f8a985b.jpg[/URL]

Hey Tom, that CT-55 has one hell of a picture on it.:thmbsp:

You know, I thought that as color television matured, the colors etc. would have improved exponentially. Obviously, I was wrong. I remember early color but didn't know about the differences.

What CRT has the "correct" colors, red especially and when did it change please? For instance, was the 15GP22 the only one with correct colors, or does a 21CPY22 have the same colors. I realize I skipped ahead by several steps.

Just wondering...


Thanks!

"Glorious Lollipop Color"....(grin)

Nice job, very good picture.... good convergence!

me likey.....

can you believe the saturation and the fleshtones?awesome pics.this is the pinnacle of real color tv.loving this!

Amazing picture, & it was done w/o all those hokey "Color Correction" circuits they had later..They HAD virtual HDTV in '54, they just didn't realize it..

Very nice screen shots!

The color "correction" circuits were really color "poor adjustment compensation" circuits. Ruin a good thing for the sake of making it "good enough for average, sloppy users". Way too much of the world is made that way, it seems.

What's wrong with this picture?
I remember that the CTC2B chassis sat flat in its cabinet and that the CTC4 chassis was the first to be mounted vertically as shown.

Wow that third screenshot, looks so sharp and defined! I thought this was S-Video modified for a second.

Loved the llama shot.

Pete

Beautiful pics as usual, Tom. Thanks for sharing!:yes:
Kevin

See post #12 in this thread:
http://www.videokarma.org/showthread...color+phosphor

The only correct NTSC tubes were the 15GP22 and the early 21AXP22's.

The main difference that affected both greens and reds was the change from P1 green to sulfide green . The sulfide red was somewhat more orangy than other reds, but the rare-earth reds usually are quite close to NTSC red.

NTSC red, by the way, is a bit orange - it can reproduce car tail lights but not traffic signal red.

Also, when the NTSC blue, which is somewhat towards cyan, was replaced by more-violet sulfide blue, this affected flesh tones by moving yellows towards green.

This has been a mishmosh in NTSC for all tubes after the 15GP22 and 21AXP22, and only PAL and now HDTV have got it right by changing the cameras to match the modern phosphors. You can see "chery red" as distinct from other reds with the modern phosphors if the camera is designed for them. Otherwise, the approximate corrections built into NTSC sets result in overdriving the bright reds, making them all overly bright and the same color as the red phosphor.

Having a 15GP22 no longer guarantees you correct color on NTSC sources, since at the same time receiver manufacturers were putting in compensation, camera manufacturers were fudging to get acceptable pictures on monitors with the new phosphors and monitor manufacturers were including switchable matrices that could either demodulate straight (but with the modern phosphors) or with an electrical correction similar to home receivers.

This was like the factory time keeper blowing the whistle based on the train departure while the train conductor set his watch by the factory whistle.

The original TK-41 cameras had no matrixing to tweak, but relied on careful selection of trimming filters in each color channel to get the best response for NTSC phosphors. This was a major reason why their optics were so inefficient. Later versions had more efficient wider-passband prism optics, but still no matrixing, and the prism passbands were still optimized for NTSC phosphors. So at this time (early 60s) the futzing was all in the receivers. As soon as Plumbicon cameras appeared, they had to have matrixing because the deep red response was so poor; and the good signal-to-noise ratio of Plumbicons made matrixing possible without excessive noise. The receiver manufacturers had been playing with the color for some years, but now the camera designers were too. This is the point at which PAL standards were developed, so they had the option to make the receivers demodulate without special compensation and do the proper compensation for the new phosphors in the cameras.
NTSC countries were not so blessed, because they already had a population of receivers with various approximate compensation for the new phosphors.

Would the 19VP22 be NTSC correct also? It was my understanding that they were, but I could be mistaken. That set is looking amazing Tom!
-Benny

Thank you very, very much for answering that question in great detail. I have a CTC-9 in queue for restoration. But it, as you must know, does not have the 21AXP22 in it, rather the 21CYP22. And, I hold no hope to ever have access to a 15GP22!

In reading this thread and the thread that was linked, it sure does bring back memories. I remember when color TV was a new thing. I just didn't remember the change in color. But I sure do know what you are talking about. I had a Heathkit GR-2000 (I think that is the number) that had a Zenith CRT in it. I always felt that the colors favored earthtones. I thought that it was similar to the 70s obsession with browns and oranges. No bright blues, greens or reds.

I also remember working for EJ Korvette. I worked in the Audio Dept. which was right next to what they called "Major Appliances". In that dept., they had a room with doors that had all of the color sets in it. It was a very dim room. Hmmm, there's a surprise! I just hope to one day have a set with the early phosphors in it.

Again, thanks for explaining and taking the time to search and post.

Unknown to me. However, a working one could be measured.

--

Here's a laugh: the original patent on the sulfide phosphors touts not only their brightness, but claims that the restricted color gamut is an advantage because the tube cannot make extremely wrong colors! In other words, if you don't correct the reduced saturation (which everyone WOULD do by turning up the color), you can't see variations in transmission so clearly.

Also, there is a story behind the NTSC blue vs. the modern blue. RCA experiments with the triniscope used very pure primary colors, including a blue nearer to present sulfide blue. But sulfide blue is very sensitive to copper impurities, which turn it green. Therefore, RCA chose a different blue phosphor that was slightly toward cyan, but doesn't change due to presence of copper during the processing of the CRT. There is a futher confusion, in that it is not clear that RCA intended the cyanish blue to go into the NTSC specs - the color coordinates that ended up in the FCC rules look suspiciously like a typo.

21CT55 Screen-Shot Test
 

Thanks for your responses. In reviewing this thread I noticed the third screen shot of my Beth, Andy and Baby Jack was not interlaced. So I replaced it with a shot having proper interlace. This correction emphasizes my contention all posted Roundy screen shots should be taken in total darkness, sized full edge to edge and presented on the viewing monitor full size edge to edge. Conveying cabinet detail is a separate issue and should be taken full room lighting. Since I have no cabinet I have no problem with this. I also added two more good shots plus a saturated color bar shot. Check it out. The bar colors appear quite accurate on my 19 inch Acer LCD monitor. The red bar’s hue is virtually indistinguishable from my clock’s red led numerical dial color hue. The CRT is a 21FBP22A rare-earth with a white-paper gray face plate. I always considered this CRT’s red to be more “pure” compared to my Sony Pro monitor which specs SMPTE-C color phosphor. The Sony’s red bar is more “orangish” like all my other CRT TVs.

While my Samsung isn't a roundie, I have found it's reds to be seriously awesome. It's greens seem a little "wrong" in my opinion though. Blue looks ok but doesn't look at all like your blues.

This is the color test my TV performed with a colorimeter.

http://img.photobucket.com/albums/v3...M.png:original

Hontou ni? Ore wa Samusung no taikutsu:D

I have no idea what language that is.

Tom explained at one time, that he modified his CTC-2B by repositioning the chassis vertically (and discarding the cabinet if he ever had it) in the 1960s when he first got that set.

If I remember right doesn't this set have video inputs installed?
I remember a thread about it a while ago. If so, can I get a link to the thread?

It must be modified for standard composite video input, I see no RF distortion AT ALL.

If it is RF, I am highly surprised.

21CT55 # 2897 Modifications
 

I bought this TV in 1964 for $50 not working. Once I got it operational I decided to keep it in the garage TV shop area. I modified the consul cabinet into a table model by rotating the chassis from horizontal to vertical to take up less room in the cramped garage, also make chassis modification and repair much easier. Yes, in 2007 it became direct composite video input only. Reactivating the RF/IF portions seemed unnecessary since I had good DVD players and the 21CT55 was never intended as a restoration item.

More BIG Screen Shots
 

Got back to the 21CT55 after 16 weeks “sanitizing” my “bought new” 1980 BMW 320i sports coupe. The CRT set came right up with nominal voltages, same as last operating. At 115vac in, B+ is 420v, Boost is 750v, FBI is 190ma, Ultor is 28kv and Focus is 5.1kv.

Took a few screen shots with my Olympus SP550UZ on full auto from 64in distance at 120 iso, 1/6 sec, 3.9 F.

http://i677.photobucket.com/albums/v...ps7719cd16.jpg

http://i677.photobucket.com/albums/v...ps0791802a.jpg

http://i677.photobucket.com/albums/v...ps492e2827.jpg

http://i677.photobucket.com/albums/v...ps43dedfd4.jpg

http://i677.photobucket.com/albums/v...ps80d0c085.jpg

http://i677.photobucket.com/albums/v...psaec6d637.jpg

http://i677.photobucket.com/albums/v...ps0f1cc6a5.jpg

That is a really good picture, and very good convergence..... Nice job....
And really good pictures....

Inx64;

Your color "gamut ICE " graph is wrong, in the green by a lot..... Someone here did those measurements and posted them on one of those 15gp's and another tube, and while people maintain the 15g's had better "full"gamut, the graph they posted showed a smaller triangle, than the other tube. I would sure be interested in seeing the response of your tube again, once you figure out what is going on with the green.... Also, try it and be sure that when the screens are turned down, the tube goes into cutoff. I have no idea how the test is run, but I find this testing interesting...

Neither did Google translate.

Only thing I can say for sure that it's not English. As I'm a single language white trash American. :D

This gamut shows colors better:
http://videokarma.org/attachment.php...3&d=1269294026

Is that line pairing on the test pattern, or just a fast shutter catching one field?

??
Not sure I follow what you said.

Inx64 did not explain his colorimeter software, but I believe what the CIE gamut display is showing is:
The light colored triangle and square boxes are the aim points for REC709 (HDTV) or sRGB monitors. The circles are the measured points for his set.


edit: bottom line, his green is not wrong, but is fairly typical for all tubes with sulfide green, which is all tubes after the 15GP22 and 21AXP22 (don't recall what green was in the 21CYP22 at the moment).


You can see that the triangle made by the RGB points for his set is slightly smaller than the REC709 spec because the green is slightly yellower.

The magenta is far off from the aim point towards red, and this has to be because of the demodulator gains and angles in his set, since moving the green primary towards yellow causes the magenta to move towards blue in order to keep the white point within spec (which it is). This error in magenta is not necessarily a bad thing, since it helps to prevent purple faces when there is an error in burst phase, and there are few recognizable magenta/purple objects (except maybe Barney the dinosaur). 2nd edit: it also compensates for the movement towards blue of magenta caused by the yellowish sulfide green, making the magenta a closer approximation to NTSC magenta.

A 15GP22 or 21AXP22 CRT has a green that is much less yellow, so the triangle is larger on the cyan side; but the blue in those CRTs is less violet and more cyan, so the triangle is smaller in the blue/purple area.

nihongo ga sukoshi dake dekimasu...

:lmao:

Absolutely beautiful screen shots! Such wonderful color rendition, and sharpness! Nah, I'd rather watch a flat screen :-) Haha.....

Just kinda says quietly & eloquently that we HAVEN'T really made all THAT much progress in nearly 60 years-Tweaking things a little one way or another, yeah, better reliability-Maybe- but better rendition of color ? Not so much...Leastways, that's the way I see it...

Observed line-pairing
 

Wayne...........Screen-shot # 5 displays only one field not a full frame time. In checking the parameters of that jpg shot I noted it's shutter opening time was 1/13 second or quite sufficient for a complete 1/30 second interlaced frame. It is apparent that line-pairing occurred during screen-shot #5 as you observed. Since all the other shots taken within a few minutes had good interlacing, an anomaly had occurred. Early-on I observed that vertical interlace quality is greatly influenced by the front panel vertical hold manual setting, even when the hold is well removed from causing vertical roll. I considered this imperfection to be of secondary importance and accommodated it by manually tweaking the vertical hold to achieve interlace immediately before any quality screen shot. Apparently, I neglected to tweak the vertical hold before shooting #5. I intend to get into the vertical deinterleaver some day to find out why the waveform differential circuitry is not operating correctly…..Tom

Interesting about the interlacing. The earlier black and white sets especially from the more expensive RCAs (630TS through 9TC270) had more complicated sync amplification and two sync clippers. Along with the good, more comprehensive vertical sync integration and the standard blocking oscillator, the interlace is generally excellent over a wider range of the vertical hold control. The cleaness of the vertical sync applied to the vertical oscilator is demonstrated when you intentionally roll the picture with the vertcal hold ids a "smooth" roll rather than a "jerky" roll which implies fully stripped sync free from any remaining video.

The mid 50's RCA sets had considerably cheapened sync circuits and disposal of the blocking oscillator transformer with feedback from the output stage expands the problem of oscilator drift from one to two stages. (I however thought the 'CT55 was the last color set to maintain the transformer). Maintaining good interlace would be affected.

It suggests to me that the designers in the mid '50's were on a tight budget to remain competitive. I'd be curious of others thoughts and if others have had the same problem with mid 50's RCA vertcal circuits. (I am currently tracing a vertical sync drift in my CTC5 as the set warms up over 1/2 hour).

Line pairing generally comes from horizontal energy getting into the vertical sync - but it doesn't have to be a sync signal separation circuit problem necessarily. It can also occur due to ground currents from the H sweep getting into the vertical, and this can be very difficult to find and fix, as it depends on the physical arrangement of ground points of the various circuits of the chassis. If the sync circuit itself is clean, then experimenting with extra ground leads and lead dress, or actually moving existing ground points may show some effect.

I have had my bouts with interlace issues on RCA sets (some same models some fine some more critical on vert hold setting), I have read that the .001 cap often found from the primary lead of the vert out trans to ground (.001) is there to ground horz noise that works its way back from the yoke thru the transformer.

I know I had a BW maggie that had a real bad interlace issue, the sync pulse looked fine, the vert hold could not be set to get rid of it (like most sets). in desperation I switched the yoke, voila interlace issue gone. I cant recall the details that lead me to suspect the yoke (I was scoping stuff and some how noticed hash that seemed to be coming from the vert out trans).

Test big 21CT55 screen shots
 

http://i677.photobucket.com/albums/v...ps4c78518f.jpg

http://i677.photobucket.com/albums/v...psa82bcb96.jpg

http://i677.photobucket.com/albums/v...pseb242eb2.jpg

http://i677.photobucket.com/albums/v...psc8b86c8a.jpg

http://i677.photobucket.com/albums/v...ps54286ef5.jpg

http://i677.photobucket.com/albums/v...ps0a3ed219.jpg

http://i677.photobucket.com/albums/v...psf15525ff.jpg

http://i677.photobucket.com/albums/v...psff15ed5b.jpg

Ferris Bueller, you're my hero.

Cameron

disgustingly good picture..... !