This is a write-up I received that I find most intriguing. A couple of years ago I built up a system very similar to the one done by J. L. Baird in 1928. There are some pictures and a few words about it, elsewhere in this website. Well this, (as they say) is a horse of another color, or is it colour?
NBTV colour? by Sam Hunt
About 6 months ago,
I handed in for my Electronics course work a system which I had devised which
gave colour to 405 line TV (I got an A). This system was very simple, and allowed
you to get full colour on a connection with no bandwidth limitation. When the
bandwidth started to become limited, it dropped first down to a 2 colour system,
then to a mono system. It worked very well, and gave very good colours at 3 colour,
and quite good colours on the 2 colour system. The mono signal was just the same
as standard 405 line TV. If the colour signal was being played back on a standard
405 line TV (The school, amazingly, still had one of these!), then I could not
notice any extra noise, or reduction of picture quality.
I went through my
proposal for this system, which outlined all of the key points, then changed all
the frequencies to more sensible frequencies for NBTV, then changed some of the
content.
I have put this explanation of the system onto the end of this message.
The only difference between a colour and a mono mechanical receiver would be that
it has three light emitters, not one (One for red, one for green and one for blue),
and it has a colour decoding circuit.
Anyway, here is the explanation. I thought you may be interested.
Explanation of frequency multiplexing colour drop down system
This
system is that at 0-10KHZ there is a channel containing all three colours. At
10-20KHZ there is a channel containing both the green and the blue colours. At
20-30KHZ there is a channel containing just the blue colour.
At the top of
each frequency band, there is a pilot tone.
How
it works is like this.
If all frequencies are being received by the reliever:
The reliever tests for a 30KHZ pilot tone, and detects one. It then decodes the
green + blue channel (10-20KHZ), and subtracts this from the mixed channel (0-10KHZ).
This leaves the red channel, and the blue + green channel. Next the blue channel
is decoded (20-30KHZ), and this is subtracted from the green + blue channel. This
leaves the green and blue channel. You now have all three colours to feed into
all three emitters.
If frequencies
past 20KHZ are being cut off.
The reliever tests for a 30KHZ pilot tone, and
fails. It then tests for a 20KHZ pilot tone, and succeeds. It then decodes the
green + blue channel (10-20KHZ), and subtracts it from the mixed channel (0-10KHZ).
This leaves the red and the green + blue channel. This is a fairly acceptable
two colour system (Red + Cyan as the primaries).
Using this system, flesh
tones are very accurate. Sky tones are quite accurate, although you can tell something
isn't quite right. Sky still looks pleasing to the eye. Grass and trees are quite
inaccurate, but they are still acceptable. Any shade of gray is 100% accurate.
If
frequencies past 10KHZ are being cut off (Or even less than 10KHZ)
If this
is happening, the reliever has no choice but to revert to the mono signal. It
tests at 30KHZ and 20KHZ, and fails to detect a tone. It then feeds the signal
directly to all three emitters. This results in a mono image.
A mono signal
would display in black and white, since there are no pilot tones.
Backwards compatibility.
Colour relievers
can display mono signals because they would not detect any pilot tones, so could
revert to the mono signal (0-10KHZ)
Mono relievers can display colour signals
because they cannot display frequencies above about 10KHZ anyway. Newer sets may
be "Colour enhanced" - They have a low - pass filter, which ensures
that there is no chance of the colour signal interfering with the mono display.
The
advantages over field sequential colour?
This system gives much less flicker
than field sequential colour, and it has backwards compatibility. It does not
need any extra synchronization, and uses up the "spare" bandwidth which
is not used for anything else at the moment.
This system could also have different bandwidth for different components - the mono (0-10KHZ) channel could have 10KHZ bandwidth, then the green + blue channel could have a lower bandwidth (About 7.5KHZ), since the eye is less sensitive to resolution in green and blue than it is to resolution in red. Finally, the blue channel could be about 5KHZ. This is because the eye is very insensitive to resolution in blue. This also would give a more realistic bandwidth of 22.5KHZ, which can be stored on CD, compared to 30KHZ, in which part of the blue channel would be ignored.
Your comments on this subject would be appreciated.
Thank you for taking time to read it.
Peter Yanczer
HOME