CBS Field-Sequential Color from PC - First Pictures! (sort-of...)

[Kat Manton]

Hi,

As regular readers of my random ramblings will know, I'm busy working out more modelines for the FotH TV System.

CBS Field Sequential is one that's fascinated me; so armed with the timing information, I worked out a modeline.

With this modeline, the computer is producing 405-line interlaced RGB video with 29.160kHz line rate and 144Hz field rate. (Or thereabouts, the closest I can get is 29.159kHz.)

To get a true CBS sequential colour video signal, a sequential switch is required, which, every field-sync pulse, switches the video output to the outputs from the PC in the sequence Red, Blue, Green. Additionally, seperate gain controls need to be provided for each colour so that white balance can be adjusted. For completeness, equalisation pulses and the red sync pulse need adding.

While looking for more information on this standard I'd found this web page describing how experimenters managed to get 4 fields of the CBS broadcasts displayed on 525/60 sets with a tweak of horizontal and vertical hold. Of course, I had to try it. Below is what happened after setting the Pye monitor on 525 and tweaking it. The trusty nixie-tube frequency counter is hung on the line sync signal from the PC, proving the line frequency matches my calculations.

Now all I have to do is build the electronics described above... oh... and build a CBS field-sequential TV set...

Regards, Kat

[Sean Williams]

Ok, This is too clever for me!

Looks good, even if I havent a clue !

Cheers
Sean

[Kat Manton]

Hi,

Originally Posted by Sean Williams

Ok, This is too clever for me!

Looks good, even if I havent a clue !

In the screenshot, the monitor is running at half the line frequency, and half the field frequency; locked to alternate line and field sync pulses. Result, four pictures

I suspect the chance of me ever acquiring either a Gray Research monitor or an actual receiver are, for all practical purposes, zero. But, I'd like to see how well the system worked. I've arguably got a source of video at the right line and field rate, at least.

So, what I (or anyone else who's mad... sorry... enthusiastic enough) needs to do, I guess, is pounce on some poor unsuspecting little monochrome monitor or TV set, persuade the 'victim' to run at 29.160kHz line and 144Hz field rates then build a colour wheel to go in front of it. The colour wheel needs some method of being synchronised; the electronics to select the RGB outputs from the computer in the correct sequence need building.

The technical challenges sound fun... anyone fancy having a go?

Regards, Kat

[ppppenguin]

Originally Posted by Kat Manton

So, what I (or anyone else who's mad... sorry... enthusiastic enough) needs to do, I guess, is pounce on some poor unsuspecting little monochrome monitor or TV set, persuade the 'victim' to run at 29.160kHz line and 144Hz field rates then build a colour wheel to go in front of it.

An old VGA monitor will do that H scan rate and may not be hard to adapt for V scan. Just drive RGB together to give a monochrome image. I know this seems a bit strange but it's not going to be easy to make a TV monitor scan at almost double its normal H rate.

For trial purposes you could sequence the signal to the RGB of the monitor and not use a colour wheel. Even more strange!

[Kat Manton]

Hi Jeffrey,

Thanks for prodding my memory... lurking among my piles of obsolete computer junk are:

• White phosphor MDA monitor (18.43kHz/50Hz)
• Amstrad Monochrome VGA monitor (31.5kHz/60Hz)
• NEC Multisync mono monitor - MDA/CGA/EGA/VGA I think...
• Apricot white-phosphor mono monitor - probably MDA (18.43kHz/50Hz) and small, 9" or 10"

I think, of these, the Apricot and NEC ones should be cowering in fear, maybe the Amstrad too... I'm glad I didn't throw these away, now.

Regards, Kat

[mickjjo]

The system looks to work quite well according to this site:-

http://home.att.net/~pldexnis/potpou...color2004.html

Regards, Mick.

[ppppenguin]

Originally Posted by Kat Manton

White phosphor MDA monitor (18.43kHz/50Hz)
Amstrad Monochrome VGA monitor (31.5kHz/60Hz)
NEC Multisync mono monitor - MDA/CGA/EGA/VGA I think...
Apricot white-phosphor mono monitor - probably MDA (18.43kHz/50Hz) and small, 9" or 10"

I would go for the NEC because it's multisync and shouldn't have any problems with 29kHz. A fixed 18kHz monitor might not be easy to modify.

I'm sure that Darryl (tubesrule) will have some worthwhile comments. He has made a version of his Aurora converter that generates CBS sequential and used it on a real set.

[Kat Manton]

Hi,

I think I'm leaning in favour of the NEC; with a bit of luck it'll only need modification to the field scan circuitry. I'll dig it out and see if it vaguely considers locking to the output of the PC; I'd expect it to give two squashed pictures one above the other with a bit of a tweak.

I've also thought about reducing the scale of colour wheel construction (The NEC monitor is 12" if memory serves), just to try things out... Once I've got a monitor locked and the colours sequenced correctly, I could build... CBS field-sequential glasses...

Regards, Kat

[tubesrule]

Kat,
I think your on the right track going for a multi-sync monitor. I used an NEC multi-sync during development as it was the only monitor I could find that handled the frequencies with no problem.
Keep in mind that the extra red field sync pulse was not used by the CBS sets, so you really don't have to create it to be correct. The sets had a push button the user just kept fiddling with until the wheel phase was correct.
Also, instead of an external RBG sequencer, would it be possible for you to just use say the red output, and do the sequencing in software? This way your current sync combiner could be used without modification.

Darryl

[Kat Manton]

Hi Darryl,

Originally Posted by tubesrule

I think your on the right track going for a multi-sync monitor.

This one's a little ancient, I'm not sure it'll be up to the task of running at 144Hz but modifying the field scan circuitry's easier than modifying line-scan circuitry.

Keep in mind that the extra red field sync pulse was not used by the CBS sets, so you really don't have to create it to be correct. The sets had a push button the user just kept fiddling with until the wheel phase was correct.

That's useful to know; in that case having a button to unlock the wheel sync and let it drift round a bit it probably more authentic than trying to get the wheel to sync automatically...

Also, instead of an external RBG sequencer, would it be possible for you to just use say the red output, and do the sequencing in software? This way your current sync combiner could be used without modification.

That is a possibility; there's provision in MythTV and at least one of the DVD player applications for introducing 'plug-in' filters on the video signal; this could be hi-jacked and a filter written which masks out two colours with the mask changing on every field. None of the existing software needs modifying - which is how I'd prefer to keep things.

(...and a friend just suggested a bizzare take on the 'glasses' idea; colour LCD shutter glasses... a rather perverse way to resurrect an obsolete TV standard... )

Regards, Kat

[tubesrule]

Hi Kat,
The NEC monitor I used was a prety old one, I think a Multisync 3fx or 3fg. It was a 14" one at any rate, so you may be pleasantly surprised with yours!
As far as we know, only the Gray Research monitor utilized the red field pulse to synchronize the wheel. It actually works pretty well on Steve's set. The CBS sets work just like you describe; you push a button, let the wheel phase slip a bit, and try it again. I'm not sure how well this would have been received by the public back then, and they would have had to make sure this didn't change when the station swapped video feeds like for commercials. I just can't see people having to constatly keep pushing the button ever time the station changed feeds!

Darryl

[Dave Moll]

Originally Posted by Kat Manton

(...and a friend just suggested a bizzare take on the 'glasses' idea; colour LCD shutter glasses... a rather perverse way to resurrect an obsolete TV standard... )

Regards, Kat

But it might be more comfortable than having spinning discs attached to your face

[chipp1968]

Maybe just take some L S D no only joking , all above my head here but keep up the good work

[Steve_McVoy]

This approach is close to having spinning disks attached to your face:

http://www.earlytelevision.org/cbs_c...al_viewer.html

It actually works, though is very hard to keep in sync.

[oldeurope]

http://audiokarma.org/forums/showthread.php?t=76990

look at this thrad, mexican colour

Darius

[Kat Manton]

Hi,

Darius, the Mexican colour system sounds interesting, thanks for bringing it to my attention; I find various field-sequential methods of acheiving colour television (and film) to be particularly fascinating.

At first glance this appeared to be similar to Claude Friese-Green's 'Biocolour' system but further reading revealed that Biocolour is somewhat more complicated than that. In Biocolour, monochrome film is exposed through three successive filters of red, orange and clear, then the film is stained alternately red and green/blue. So you get six successive frames of different colours before it repeats, as follows (I think I've got this right..!)

1. red -> red
2. orange -> green/blue
3. clear -> red
4. red -> green/blue
5. orange -> red
6. clear -> green/blue

I've also realised that with the software filters as mentioned above, along with cranking the field-rate of a colour monitor as high as it will go I can simulate pretty well any field/frame-sequential colour system. I think I'll have to have a play around with this - it doesn't involve building anything, just writing some software.

Checking the specs of one of my monitors (a massive 20" Silicon Graphics workstation monitor) this has a horizontal frequency range of 30-96 kHz and vertical range of 48-160 Hz. So, that'll handle vertical rates fast enough to reduce flicker on field-sequential systems and covers the correct original rate of several of them.

It's interesting to think about this - what if CBS sequential had been adopted as a standard which persisted to this day? As colour CRTs appeared the colour wheel would have been dispensed with and the system would operate exactly as I can emulate with a colour monitor.

I feel some further experiments coming on... but I still feel very inclined to construct a monitor using a monochrome CRT and colour wheel just because... well... it sounds like fun.

Regards, Kat

[oldeurope]

I think about making a colour wheel (mexican system) for a home made set using a MW 17-69. At the moment I have no good idea for the colour wheel.
Generating mexican colour is very easy, my set top box makes RGB at its scart output and the mexican signal can be handled like a black and white signal. It can be converted to 405 or recorded on a monochrom vcr.
The problem with the wheel is it must have a red and a cyan and a big black section. This black section is the problem, it makes losses in brightness.
There must be a way to change from red to cyan from frame to frame (frame= 1/2 picture).

Darius

[Kat Manton]

Hi,

Originally Posted by oldeurope

There must be a way to change from red to cyan from frame to frame (frame= 1/2 picture).

(Terminology clarification - what I call a field is what you call a frame and what I call a frame is what you call a picture. Odd field + even field = frame. Unless you're using triple interlace... Hope that clears up any confusion. )

Well, since cyan = green+blue you can either:

1. With a software filter (as per the system I'm using), mask off red, then mask off green and blue on alternate fields. So you'd get, say, the odd field as red and the even field as cyan.
2. Use hardware to do the same thing; add red, green and blue as usual to monochrome but insert switches in the RGB feeds, then trigger a flip-flop on the field-sync pulse then use that to turn off first red, then green and blue together.

...or am I missing something subtle here and it doesn't work like that..?

Regards, Kat

[Andy Green]

For what it's worth, I remember that a Wireless World from the early 70's had an article on how to build a colour wheel set, and AFAIR it included a circuit to convert a PAL signal to field sequential compatable one for feeding to a monochrome set. I had a copy of this article, maybe ten years ago, it might still be in the loft in my daughters house (which was at on time my house I will check soon - I have to install a WifI router for her, so will rooot through the attic then) - anyone else remember this?

[Panrock]

For those hankering after frame sequential colour (as we all do at some time) but who can't get their hands on a color wheel right now, click HERE for some quick relief. Make sure your movie viewer is set to 'repeat' mode first though...

Steve