Why is video inverted for transmission?

[Joerg]

Joerg wrote:

(snip)



I believe DVD recorders with ATSC tuners are now available for around
$200 or less. They at least have composite video output, probably
RF out, too.

Yes, I believe Costco has one right now. Probably only online though.

Otherwise the government is supposed to subsidize the converters.

It's a bit late in the game now. If serious issues for example with
multipath are discovered that won't leave much time to deal with that.
Since the standard is cast in concrete the only option would be
repeaters. The permit procedure alone would eat up much of the remaining
time.

 

[Jerry Avins]

There is a difference between bit error rate and multipath distortion.
Big difference.

With digital, and of course, the FEC that gets included, it is either
ALL there or not displayable. There is NO distortion ever.

Get on board. You're locked into distortion of the signal delivered *by*
the receiver, while the chaps who know the real world are discussing
distortion of the signal delivered *to* the receiver. If you took the
time to bone up on how things are done, you might not be bored any more.

Jerry

 

[Jerry Avins]

At first I wanted to agree with you - it makes sense in one way.

But a sync pulse is really nothing more than the horizontal scan rate,
right? You can build a PLL to pull this way out of the noise by making
a very narrow-band loop filter. The video information is a different
matter since you need a wide bandwidth to utilize it.

So I'd say the sync pulse power can be lower than the video power, by the
ratio of the bandwidths, and still be appropriate for a usable video signal.

Randy, back when TV first went commercial, a frequency divider was a
multivibrator with a little of the pilot signal introduced somewhere. It
wasn't possible to make large divisions stably, so cascades of smaller
divisions were used. RCA's original system used 441 lines -- I assume
you know about odd line count and interleave -- and synced horizontal to
vertical with two divisions by 7 and two by 3. The FCC, guessing that
CRTs larger than ten inches would eventually come, insisted on 525. That
factors to 5x5x7x3. Don't confuse what is easy today with what was
practical using tubes and discrete components. The original TV sync
circuits were designed when the Eccles-Jordan bistable was a novelty.

Jerry

 

[Jerry Avins]

Oops, my error, I was forgetting timing. For UK PAL, the sync area
occupies 12uSec, and the video 52uSec. So the sum should be (.7
*52)/(.3*12) or roughly 10:1 video to sync power ratio.

That's for horizontal sync. More area is used for vertical, so your
original figure may be close to correct overall. (I haven't worked it out.)

The idea was that sync should start dropping at about the same time as
the picture started becoming unwatchable because of noise. This ratio
achieved that balance. There would be no point in having a picture
that was clear but rolling, or one that was locked solid and looking
like boiling rice.

Strictly, "rolling" is loss of vertical sync. (If we don't watch out,
we'll be discussing whether the equalizer notches during vertical sync
are still needed.)

Jerry

 

[Don Pearce]

That's for horizontal sync. More area is used for vertical, so your
original figure may be close to correct overall. (I haven't worked it out.)

Certainly lower, but probably not that close.

Strictly, "rolling" is loss of vertical sync. (If we don't watch out,
we'll be discussing whether the equalizer notches during vertical sync
are still needed.)

Jerry

Yebbut, vertical sync is just made of a load of horizontal syncs with
variable mark-space ratios in the pulses.

d

 

[Jerry Avins]

Randy Yates wrote:
...


Yeah BUT
What techniques and hardware were commonly used/available almost a
century ago. Cost would be important.

The TV of a century ago -- Nipkow, 1884 patent -- has absolutely no
influence on today's standard. There was one on-air broadcaster[1] and
one maker of bare chassis and kit receivers[2] (build your own cabinet)
in 1939.

Jerry

[1] RCA, experimental, 441 lines.
[2] Andrea Radio, successor to the depression-bankrupt FADA Radio &
Electric Co. FADA stood for Frank A. De Andrea. Frank was an old man
when I worked for him in 1956, maybe even older than I am now.

 

[Jerry Avins]

Randy Yates wrote:

...

What does "prognostigation" mean? I can't find it in my dictionary.

Pro: before. Gnosis: knowing. Literally, "knowing before". Practically,
prediction. You need a better dictionary, but the context in which the
word showed up here would have given me no hint either.

Jerry

 

[Jerry Avins]

Yebbut, vertical sync is just made of a load of horizontal syncs with
variable mark-space ratios in the pulses.

Horizontal sync is gotten by differentiating. Vertical sync is gotten by
integrating. During vertical retrace, we need to offset horizontal sync
by half a line. Equalizer notches do that, but I won't discuss it.

Jerry

 

[Don Pearce]

Horizontal sync is gotten by differentiating. Vertical sync is gotten by
integrating. During vertical retrace, we need to offset horizontal sync
by half a line. Equalizer notches do that, but I won't discuss it.

Yup, that's what I was talking about.

d

 

[Bob Myers]

You are talking about the situation post sync-separator. In the
composite video signal, sync tip is 0V. Black level is 0.3V and peak
white is 1V.

Nope; check the standards. The reference is always
the blanking level; if the sync happens to be at "0V"
with respect to the local reference at some point in the
circuit, that's simply a coincidence. You never, ever
reference the rest of the video signal to the sync tips
if you're following the standards, so to speak of sync
as being at "0V" is meaningless.

Bob M.

 

[Gary Tait]

[email protected] (Ray Fischer) wrote in [email protected]:

It is indeed. Too bad it's going to be moot in 17 months when all the
broadcasters go over to digital only.

It will still be very relavent even when broadcasters and cable turn off
their analog. Stet-top boxes and displays will still output analog video
for some time to come.

 

[Don Pearce]

Nope; check the standards. The reference is always
the blanking level; if the sync happens to be at "0V"
with respect to the local reference at some point in the
circuit, that's simply a coincidence. You never, ever
reference the rest of the video signal to the sync tips
if you're following the standards, so to speak of sync
as being at "0V" is meaningless.

Bob M.

Sure the standards use the blanking level as the reference, but you
must apply a sync = 0 signal to a modulator.

d

 

[Bob Myers]

The sync pulse is defined in the standard as occupying 30%. The video
occupies 70%, So the video has more than 5 times the power of the sync
(70/30)^2.

I just realized that we're talking similar but conflicting standards
here. In the UK and other "PAL" countries, sync is indeed
30% of the peak-to-peak signal, and the black-to-white range
the other 70% - with no distinction made between the black
and blank levels (i.e., "zero setup"). That is, however, NOT the
case in the U.S. or other countries still adhering to the original
NTSC and EIA video signal definitions. Those consistently
described the peak-to-peak video signal in terms of "IRE
units," with the blanking level as the reference and the blank-
to-white range pegged at "100 IRE." Under that definition, the
sync tips are 40 IRE "down" from blank, and black is 7.5 IRE
"up". What you wind up with in terms of absolute voltages
depends on the standard p-p swing of the signal. The older
RS-170 standard established a 1.4 Vp-p video signal, so the
sync tips were -0.400V with respect to blank, while white was
+ 1.000 V. Later, RS-343 revised this to a 1.000 Vp-p
(including sync), which resulted in the currently-typical values
of ROUGHLY 0.7V for white - but there is still a distinction
between North American and European practice. Simply
rescaling the 140 IRE range to a 1.000 Vp-p swing results
in some odd numbers - sync is -0.286V with respect to blank,
black is +0.054V, and white is +0.714V. When the European
standards were established (Europe lagged a bit in producing
TV standards, of course, due to the effects of WW II), a more
sensible approach was taken, as the blank-to-black "setup" or
"pedestal" was eliminated and the 1 V swing divided in the
30/70 ratio you mentioned, between sync and video. We've
also both been right about the reference - apparently the signal
is "built up from the sync tips" in standard European practice,
while we in North America always have considered the blanking
level to be the "0V" reference.

Bob M.

 

[Bob Myers]

You're describing the ideal spec. In practice, the clamping diode is
referenced to ground, but the narrow sync tips don't have as much effect
as the wider blanking porch (except during vertical sync. You can see this
on some sets as a slight black-level shift at the top of the screen.
Overscan often masks that anyway.)

Actually, see my later post on the same topic; Don and I
were actually both right, as we were describing differing
(European vs. North American) standards.

Yes, the clamping diode is referenced to ground, but I was
talking about what's used as the "internal" reference point for
the video signal itself.

Bob M.

 

[Don Pearce]

I just realized that we're talking similar but conflicting standards
here. In the UK and other "PAL" countries, sync is indeed
30% of the peak-to-peak signal, and the black-to-white range
the other 70% - with no distinction made between the black
and blank levels (i.e., "zero setup"). That is, however, NOT the
case in the U.S. or other countries still adhering to the original
NTSC and EIA video signal definitions. Those consistently
described the peak-to-peak video signal in terms of "IRE
units," with the blanking level as the reference and the blank-
to-white range pegged at "100 IRE." Under that definition, the
sync tips are 40 IRE "down" from blank, and black is 7.5 IRE
"up". What you wind up with in terms of absolute voltages
depends on the standard p-p swing of the signal. The older
RS-170 standard established a 1.4 Vp-p video signal, so the
sync tips were -0.400V with respect to blank, while white was
+ 1.000 V. Later, RS-343 revised this to a 1.000 Vp-p
(including sync), which resulted in the currently-typical values
of ROUGHLY 0.7V for white - but there is still a distinction
between North American and European practice. Simply
rescaling the 140 IRE range to a 1.000 Vp-p swing results
in some odd numbers - sync is -0.286V with respect to blank,
black is +0.054V, and white is +0.714V. When the European
standards were established (Europe lagged a bit in producing
TV standards, of course, due to the effects of WW II), a more
sensible approach was taken, as the blank-to-black "setup" or
"pedestal" was eliminated and the 1 V swing divided in the
30/70 ratio you mentioned, between sync and video. We've
also both been right about the reference - apparently the signal
is "built up from the sync tips" in standard European practice,
while we in North America always have considered the blanking
level to be the "0V" reference.

Bob M.

So we have! I've always lived with the PAL-I standard (the highest
resolution and best colour in the world, one might say) and I'd
forgotten that others do things differently.

Unfortunately it is all moot because in a couple years we switch off
our nice PAL in favour of digital which is, frankly, shit. The colours
aren't right, there is no resolution to speak of and the sound is so
far out of sync that we are lucky if it accompanies the right
programme. So much for progress.

d

 

[Martin Griffith]

On Mon, 24 Sep 2007 16:31:22 GMT, in sci.electronics.design

So we have! I've always lived with the PAL-I standard (the highest
resolution and best colour in the world, one might say) and I'd
forgotten that others do things differently.

Unfortunately it is all moot because in a couple years we switch off
our nice PAL in favour of digital which is, frankly, shit. The colours
aren't right, there is no resolution to speak of and the sound is so
far out of sync that we are lucky if it accompanies the right
programme. So much for progress.

d

And you didn't even mention DAB Vs FM radio <duck>


Martin

 

[Don Pearce]

And you didn't even mention DAB Vs FM radio <duck>

Arrrrrrggggghhhhhhh!

Ooh! Just for interest, my spell checker thought I was trying to say
arteriosclerosis. That kind of sums up DAB.

d

 

[Richard Fry]

There seems to be an idea floating around that if the sync tips are at
the 100% modulation end rather than the 0% modulation end (ie
inverted, not non-inverted) they are represent a somehow stronger
signal. In AM that just isn't so.

_______

It is when video is transmitted in inverse polarity, using AM. The
power radiated from the transmit antenna is maximum at the tip of
sync, which means that the received field strength also is maximum at
the tip of sync.

If video were transmitted in normal polarity the output power of the
transmitter would be considerably less at the tip of sync, and TV sets
in weak signal locations would not lock up as reliably.

RF

 

[Bob Myers]

Unfortunately it is all moot because in a couple years we switch off
our nice PAL in favour of digital which is, frankly, shit. The colours
aren't right, there is no resolution to speak of and the sound is so
far out of sync that we are lucky if it accompanies the right
programme. So much for progress.

I have to admit that I also haven't been keeping very
current on what's going on re broadcast DTV in Europe,
other than still being upset that we got stuck with 8-VSB and
you didn't. What are you expecting now in terms of
the standard resolutions that will be carried, etc.?

Bob M.

 

[Don Pearce]

I have to admit that I also haven't been keeping very
current on what's going on re broadcast DTV in Europe,
other than still being upset that we got stuck with 8-VSB and
you didn't. What are you expecting now in terms of
the standard resolutions that will be carried, etc.?

Bob M.

I just don't know. The ambition of the multiplex operators here
appears to be simply to cram as many programmes on as the mux will
carry. There are no sharp programmes on our digital service.

d