PAL/NTSC - chroma phase vs. colour documentation

[Silver Dream !]

Hi group! I am looking for documentation that would clearly say what
colour is expected at which phase difference between the reference
phase (colour burst) and the actual signal subcarrier phase in given
time. Sure, I can connect generator to a vectorscope and measure it,
but I am looking for the official source/documentation that engineers
who needed to build colour decoders back then had to use. Any clues?

--
SD!

 

[Jeff Urban]

From what I've read PAL is very similar to NTSC. It has a scheme with a comb filter that makes a phase (tint) control unnecessary, 50Hz frame rate rather than 60 and a higher horizontal rate resulting in more lines in the raster. (your spelling says you are across the pond)

The burst phase is considered -I, where I is 180Âş out that is blue or actually B-Y, which is blue without the Y component which is what the black and white signal is called here. For red there is Q for quadrature which is 90Âş out of phase with the I and therefore the two do not interfere with each other. our vector algebra will tell you that a mixed 90Âş signal with the original phase does not affect the amplitude. Thus you get two discrete signals from one carrier. In the US it is 3.579545MHz which inherently causes interlace, I imagine the PAL carrier, which I think is 4.43MHz accomplishes the same in PAL.

A vectorscope would give you the data. It is a bit easier to see when you have a phase control. You DO but it is not user adjustable because of the system making undesirable phase shifts cancel each other out.

These color levels and phases are selected due to several factors having to do with visual perception. They also match the contour - for lack of a better word - to the makeup of the Y signal. It is not 33%, 33%, 33%, it is actually mostly green. Thus the -Y for the green is the least precedent in comparison.

The I and Q do not exactly match red and blue for whatever reason. They are a hair off, so many manufacturers went with a different demodulation angle than 90Âş, usually wider like 105Âş or whatever.

In the US the I signal is transmitted with a wider bandwidth and there are rare receivers that can use that. It requires an extra delay line. Also with that the only choice is true I and Q so whatever matrixing that has to be done falls on the engineer. Very very few units had that. I do not know if anything similar to that exists in the PAL system.

The bottom line is that I is blue and Q is red. Green was derived from those usually here, not demodulated itself.

I have worked (in the states) on some sets that had PAL capability but it was mostly all in chips so further detail would have to come from a manufacturer training courses or the standards committee of the government.

So you may gloat that your analog color TV system was better than ours, however it is based on the same principles. Also your PAL, phase alternate by line system's one advantage depends on the comb filter, which I BELIEVE was invented or at least made practical by Sony for the color under home video recorder format.

But we had it first. And from what I heard the Russian system was the worst..

Now the French developed system called SECAM, I really have little to no idea how that works. However the US was first and the government insisted on compatibility with the kagillions of black and white sets out there in use. I think PAL also has that capability but I doubt it with SECAM.

Still, we invented the dot matrix color CRT without which all this would be impractical. All color TVs would be projection of some type and that means three CRTs. We did use that scheme quite a bit for projection TVs but it took much better technology to accomplish it. I was one of the few who really knew the convergence (registration) circuitry in those things and that was my job for some time.

Form what I recollect, which means was told, the first color broadcasts were in the CBS system which was not compatible. Separate carriers for each of the three primary colors. Used on the moon. Funny now that we are back to that. What do you think is on them HDMI cables or comes out of the switcher ? Three channels, one for each color.

Anyway, the compatible NTSC system owes something to FM stereo which uses a very similar approach but no Q carrier.

In parts of Europe they used to call the NTSC system Never The Same Color but that was more due to engineers taking liberties with the accuracy of the design. For one the wider angle >90Âş actually had the benefit of auto flesh tone. When the phase was a bit of the red and green wound up demodulated differently and the phase shift affected the picture less. And their attitude was "Who cares if a wall is aqua or blue ?". They actually concerned themselves much less with accuracy than with producing a picture pleasing to the eye. Sony was one major exception. Their attitude was to go for the accuracy and if the customer has to adjust the tint (phase) control so be it, but you KNOW if the wall is blue or aqua. It also got pleasing real quick like and other sets with all these liberties taken in the design started having their own "look". That was mostly brown and blue. When people saw the Sony after watching all this faux color they liked the fact that it seemed to have a better range of colors. It is like speakers in a way, if everything sounds the same it is not high fidelity. That is the US forya. In fact one manufacturer chose different phosphors for their CRTs to look better under the lights at a supermarket or big box store. Then you got it home and like, ewwww.

The BBC bought up the rights to a bunch of Star Trek TNGs. I wonder if they got the original series. The test of a TV color system ? Is Spock green and everyone else normal color ? I shit you not. they only put a little bit of green tint on him so many TVs displayed the same color. Sonys, certain RCAs, Zeniths, they were the main ones to give at least somewhat accurate color rendition.

We had an RCA, one of the good ones - a CTC25 chassis.

Enough. This is probably more information than you wanted. (I am an expert on the NTSC system) But you can prove it all with the vectorscope. All you need to do is to find a way to vary the phase of the subcarrier a bit to see what happens. It will come clear.

 

[Fred Smith]

On 2020-03-17, Jeff Urban <jurb6006@gmail.com> wrote:
> In parts of Europe they used to call the NTSC system Never The Same Color

NTSC: Never Twice Same Color

SECAM: System Essentially Contrary to AMericans

PAL: Perfect AT Last.

 

[Jeff Urban]

Fuck you ! LOL

No offense intended, this time...

 

[Silver Dream !]

On 2020-03-17 02:48:00 +0000, Jeff Urban said:

> From what I've read PAL is very similar to NTSC.

Colur encoding concept is the same.

It has a scheme with a comb filter that makes a phase (tint) control
unnecessary,

It inverts phase every other line so that phase errors cancel each
other out when summed.

50Hz frame rate rather than 60 and a higher horizontal rate resulting
in more lines in the raster.
(your spelling says you are across the pond)

The burst phase is considered -I, where I is 180ş out that is blue or
actually B-Y, which is blue without the Y component which is what the
black and white signal is called here. For red there is Q for
quadrature which is 90ş out of phase with the I and therefore the two
do not interfere with each other. our vector algebra will tell you that
a mixed 90ş signal with the original phase does not affect the
amplitude. Thus you get two discrete signals from one carrier. In the
US it is 3.579545MHz which inherently causes interlace, I imagine the
PAL carrier, which I think is 4.43MHz accomplishes the same in PAL.

Yes, the PM variant of QAM modulation is used in both. The differences
are in frequencies and PAL error cancellation.

A vectorscope would give you the data. It is a bit easier to see when
you have a phase control. You DO but it is not user adjustable because
of the system making undesirable phase shifts cancel each other out.
These color levels and phases are selected due to several factors
having to do with visual perception. They also match the contour - for
lack of a better word - to the makeup of the Y signal. It is not 33%,
33%, 33%, it is actually mostly green. Thus the -Y for the green is the
least precedent in comparison.
The I and Q do not exactly match red and blue for whatever reason. They
are a hair off, so many manufacturers went with a different
demodulation angle than 90ş, usually wider like 105ş or whatever.
In the US the I signal is transmitted with a wider bandwidth and there
are rare receivers that can use that. It requires an extra delay line.
Also with that the only choice is true I and Q so whatever matrixing
that has to be done falls on the engineer. Very very few units had
that. I do not know if anything similar to that exists in the PAL
system.
The bottom line is that I is blue and Q is red. Green was derived from
those usually here, not demodulated itself.
I have worked (in the states) on some sets that had PAL capability but
it was mostly all in chips so further detail would have to come from a
manufacturer training courses or the standards committee of the
government.

So you may gloat that your analog color TV system was better than ours,

It was better in terms of colour fidelity and resolution. It was worse
in terms of flickering, due to lower refresh rate and some el cheapo
decoders being employed too often.

however it is based on the same principles. Also your PAL, phase
alternate by line system's one advantage depends on the comb filter,
which I BELIEVE was invented or at least made practical by Sony for the
color under home video recorder format.
But we had it first. And from what I heard the Russian system was the worst.
Now the French developed system called SECAM, I really have little to
no idea how that works.

Instead of a variant of QAM modulation to encode U and V (I/Q) it sends
the two in sequence, one per each line.

However the US was first and the government insisted on compatibility
with the kagillions of black and white sets out there in use. I think
PAL also has that capability but I doubt it with SECAM.

All three have. The requirement was the same everywhere.

[...]

Enough. This is probably more information than you wanted. (I am an
expert on the NTSC system) But you can prove it all with the
vectorscope. All you need to do is to find a way to vary the phase of
the subcarrier a bit to see what happens. It will come clear.

Yeah, thanks - the thing is though: I can measure this with somewhat
reasonable accuracy by sending pure R/G/B on to vectorscope and this
will tell me how _does_ the generator encode the signal. What I am
looking for though is the official/norm/standard specification telling
how it _should_ encode it.

If chroma phase in reference to burst phase is the hue, what is the
colour of 0 degrees, what angle should pure red, green, blue lie at in
reference to the burst. You say "it's a hair off". Why so? Etc.
Something like a real standard specification.

--
SD!

 

On Tuesday, 17 March 2020 02:48:03 UTC, Jeff Urban wrote:

From what I've read PAL is very similar to NTSC. It has a scheme with a comb filter that makes a phase (tint) control unnecessary, 50Hz frame rate rather than 60 and a higher horizontal rate resulting in more lines in the raster. (your spelling says you are across the pond)

The burst phase is considered -I, where I is 180Âş out that is blue or actually B-Y, which is blue without the Y component which is what the black and white signal is called here. For red there is Q for quadrature which is 90Âş out of phase with the I and therefore the two do not interfere with each other. our vector algebra will tell you that a mixed 90Âş signal with the original phase does not affect the amplitude. Thus you get two discrete signals from one carrier. In the US it is 3.579545MHz which inherently causes interlace, I imagine the PAL carrier, which I think is 4.43MHz accomplishes the same in PAL.

A vectorscope would give you the data. It is a bit easier to see when you have a phase control. You DO but it is not user adjustable because of the system making undesirable phase shifts cancel each other out.

These color levels and phases are selected due to several factors having to do with visual perception. They also match the contour - for lack of a better word - to the makeup of the Y signal. It is not 33%, 33%, 33%, it is actually mostly green. Thus the -Y for the green is the least precedent in comparison.

The I and Q do not exactly match red and blue for whatever reason. They are a hair off, so many manufacturers went with a different demodulation angle than 90Âş, usually wider like 105Âş or whatever.

In the US the I signal is transmitted with a wider bandwidth and there are rare receivers that can use that. It requires an extra delay line. Also with that the only choice is true I and Q so whatever matrixing that has to be done falls on the engineer. Very very few units had that. I do not know if anything similar to that exists in the PAL system.

The bottom line is that I is blue and Q is red. Green was derived from those usually here, not demodulated itself.

I have worked (in the states) on some sets that had PAL capability but it was mostly all in chips so further detail would have to come from a manufacturer training courses or the standards committee of the government.

So you may gloat that your analog color TV system was better than ours, however it is based on the same principles. Also your PAL, phase alternate by line system's one advantage depends on the comb filter, which I BELIEVE was invented or at least made practical by Sony for the color under home video recorder format.

But we had it first. And from what I heard the Russian system was the worst.

Now the French developed system called SECAM, I really have little to no idea how that works. However the US was first and the government insisted on compatibility with the kagillions of black and white sets out there in use. I think PAL also has that capability but I doubt it with SECAM.

Still, we invented the dot matrix color CRT without which all this would be impractical. All color TVs would be projection of some type and that means three CRTs. We did use that scheme quite a bit for projection TVs but it took much better technology to accomplish it. I was one of the few who really knew the convergence (registration) circuitry in those things and that was my job for some time.

Form what I recollect, which means was told, the first color broadcasts were in the CBS system which was not compatible. Separate carriers for each of the three primary colors. Used on the moon. Funny now that we are back to that. What do you think is on them HDMI cables or comes out of the switcher ? Three channels, one for each color.

Anyway, the compatible NTSC system owes something to FM stereo which uses a very similar approach but no Q carrier.

In parts of Europe they used to call the NTSC system Never The Same Color but that was more due to engineers taking liberties with the accuracy of the design. For one the wider angle >90Âş actually had the benefit of auto flesh tone. When the phase was a bit of the red and green wound up demodulated differently and the phase shift affected the picture less. And their attitude was "Who cares if a wall is aqua or blue ?". They actually concerned themselves much less with accuracy than with producing a picture pleasing to the eye. Sony was one major exception. Their attitude was to go for the accuracy and if the customer has to adjust the tint (phase) control so be it, but you KNOW if the wall is blue or aqua. It also got pleasing real quick like and other sets with all these liberties taken in the design started having their own "look". That was mostly brown and blue. When people saw the Sony after watching all this faux color they liked the fact that it seemed to have a better range of colors. It is like speakers in a way, if everything sounds the same it is not high fidelity. That is the US forya. In fact one manufacturer chose different phosphors for their CRTs to look better under the lights at a supermarket or big box store. Then you got it home and like, ewwww.

The BBC bought up the rights to a bunch of Star Trek TNGs. I wonder if they got the original series. The test of a TV color system ? Is Spock green and everyone else normal color ? I shit you not. they only put a little bit of green tint on him so many TVs displayed the same color. Sonys, certain RCAs, Zeniths, they were the main ones to give at least somewhat accurate color rendition.

We had an RCA, one of the good ones - a CTC25 chassis.

Enough. This is probably more information than you wanted. (I am an expert on the NTSC system) But you can prove it all with the vectorscope. All you need to do is to find a way to vary the phase of the subcarrier a bit to see what happens. It will come clear.

We got the UK PAL system 10 years later than the US, it was effectively the mark 2 version of NTSC.

Re B&W compatibility, it was, but PAL colour was totally not compatible with the B&W standard it replaced. The older standard (adopted in the 1930s) was 405 lines, positive modulation, VHF transmission and AM sound on a separate broadcast frequency. 625 line PAL was -ve video modulation, UHF carrier, embedded FM sound. For a while there were dual standard sets with a huge sliding switch that ran the length of the main PCB.

Re dot matrix tubes, there were previous non-projection systems that were never popular. There was the colour wheel, a 3 colour filter wheel than spun on front of a B&W CRT giving field sequential colour. And there was the spinning mirror type, effectively an upgraded nipkov-esque system. That approach is still used in laser printers.

The first colour TV was nipkov type in the 1920s. Not a widley rolled out standard though.

'Never The Same Colour twice' was what I always heard. PAL certainly wasn't perfect but it was a significant upgrade.

Re colour setup, most PAL TVs were much the same but some sets were quite off. I never found out whether that was due to decoder or differing phosphors. I remember a minority of sets giving a rather yellowy green in lieu of green - the overall result was a quite pleasant warmer picture but it wasn't accurate.

I designed an analogue NR system that was effective at addressing noise in dark picture areas, PAL's biggest shortcoming, but that was just before digital took over so went nowhere.

Delta tube sets typically didn't converge accurately. That marred resolution as well as giving miscoloured fringes. It did not look good. The Sony tube really solved that. Also its higher output enabled more tinting giving better contrast as well as brighter picture. And less curvature meant less geometric distortion. Plus the Sonys were generally significantly lower video noise sets. OTOH early Sony colour sets used a throwback tuning knob, no channel preselectors. Sonys were also the only PAL sets I know of to have a tint control.


NT

 

[Stephen Wolstenholme]

On Tue, 17 Mar 2020 04:38:02 -0700 (PDT), tabbypurr@gmail.com wrote:

>Re colour setup, most PAL TVs were much the same but some sets were quite off. I never found out whether that was due to decoder or differing phosphors. I remember a minority of sets giving a rather yellowy green in lieu of green - the overall result was a quite pleasant warmer picture but it wasn't accurate.

I worked on TV when PAL was introduced in the UK. I set up a NTSC set
and a PAL set so they could be compared. PAL was clearly the better
system.

Steve

--
http://www.npsnn.com

 

[N_Cook]

On 17/03/2020 12:06, Stephen Wolstenholme wrote:

On Tue, 17 Mar 2020 04:38:02 -0700 (PDT), tabbypurr@gmail.com wrote:

Re colour setup, most PAL TVs were much the same but some sets were quite off. I never found out whether that was due to decoder or differing phosphors. I remember a minority of sets giving a rather yellowy green in lieu of green - the overall result was a quite pleasant warmer picture but it wasn't accurate.

I worked on TV when PAL was introduced in the UK. I set up a NTSC set
and a PAL set so they could be compared. PAL was clearly the better
system.

Steve

Hence the term
Never The Same Color

--
Monthly public talks on science topics, Hampshire , England
<http://diverse.4mg.com/scicaf.htm>

 

[Chuck]

On Tue, 17 Mar 2020 12:06:42 +0000, Stephen Wolstenholme
<steve@easynn.com> wrote:

On Tue, 17 Mar 2020 04:38:02 -0700 (PDT), tabbypurr@gmail.com wrote:

Re colour setup, most PAL TVs were much the same but some sets were quite off. I never found out whether that was due to decoder or differing phosphors. I remember a minority of sets giving a rather yellowy green in lieu of green - the overall result was a quite pleasant warmer picture but it wasn't accurate.

I worked on TV when PAL was introduced in the UK. I set up a NTSC set
and a PAL set so they could be compared. PAL was clearly the better
system.

Steve

When I came to the UK from the states, the 50hz flickering annoyed me.
Did natives notice the flicker?

 

[Silver Dream !]

On 2020-03-17 16:01:38 +0000, Chuck said:

Re colour setup, most PAL TVs were much the same but some sets were
quite off. I never found out whether that was due to decoder or
differing phosphors. I remember a minority of sets giving a rather
yellowy green in lieu of green - the overall result was a quite
pleasant warmer picture but it wasn't accurate.

I worked on TV when PAL was introduced in the UK. I set up a NTSC set
and a PAL set so they could be compared. PAL was clearly the better
system.

When I came to the UK from the states, the 50hz flickering annoyed me.
Did natives notice the flicker?

Normally no, we didn't notice that. Or noticed but didn't complain
about it because we didn't know that it could behave differently.

--
SD!

 

[Phil Allison]

tabb...@gmail.com wrote:

-----

'Never The Same Colour twice' was what I always heard.

** Never Twice the Same Colour" is the acronym.


SECAM = "Something Essentially Contrary to the American Method"

MCAS = "Money Counts Above Safety"



...... Phil

 

[Michael Terrell]

On Monday, March 16, 2020 at 11:00:36 PM UTC-4, Fred Smith wrote:

On 2020-03-17, Jeff Urban wrote:
In parts of Europe they used to call the NTSC system Never The Same Color

NTSC: Never Twice Same Color

SECAM: System Essentially Contrary to Americans

PAL: Perfect AT Last.

NTSC predates Color TV. It stands for 'National Television Standards Committee'. It is from the very early days of electronically scanned TV in the United States.

 

[Adam]

On 2020-03-17, Silver Dream ! <email@domain.com> wrote:

Yeah, thanks - the thing is though: I can measure this with somewhat
reasonable accuracy by sending pure R/G/B on to vectorscope and this
will tell me how _does_ the generator encode the signal. What I am
looking for though is the official/norm/standard specification telling
how it _should_ encode it.

Here you go:

https://www.govinfo.gov/content/pkg/CFR-2014-title47-vol4/pdf/CFR-2014-title47-vol4-sec73-682.pdf
https://www.govinfo.gov/content/pkg/CFR-2019-title47-vol4/pdf/CFR-2019-title47-vol4-sec73-699.pdf

But beware... My understanding is that the phosphors actually used in TV
sets don't conform to the standard.

-- Adam

 

[Michael Terrell]

On Tuesday, March 17, 2020 at 8:45:15 PM UTC-4, Adam wrote:

On 2020-03-17, Silver Dream ! <email@domain.com> wrote:
Yeah, thanks - the thing is though: I can measure this with somewhat
reasonable accuracy by sending pure R/G/B on to vectorscope and this
will tell me how _does_ the generator encode the signal. What I am
looking for though is the official/norm/standard specification telling
how it _should_ encode it.

Here you go:

https://www.govinfo.gov/content/pkg/CFR-2014-title47-vol4/pdf/CFR-2014-title47-vol4-sec73-682.pdf
https://www.govinfo.gov/content/pkg/CFR-2019-title47-vol4/pdf/CFR-2019-title47-vol4-sec73-699.pdf

But beware... My understanding is that the phosphors actually used in TV
sets don't conform to the standard.

There are none that are an exact natch, but they have come a long way in the past 50+ years. The first could not render an accurate flesh tone, and the image was cartoonish. The first 'Rare Earth' phosphors helped a lot, and they have been tweaked over the following years.

 

[Ian Jackson]

In message <5fa87dad-beed-49db-9008-4c6af41ffb1f@googlegroups.com>,
Michael Terrell <terrell.michael.a@gmail.com> writes

On Monday, March 16, 2020 at 11:00:36 PM UTC-4, Fred Smith wrote:
On 2020-03-17, Jeff Urban wrote:
In parts of Europe they used to call the NTSC system Never The Same Color

NTSC: Never Twice Same Color

SECAM: System Essentially Contrary to Americans

PAL: Perfect AT Last.


NTSC predates Color TV. It stands for 'National Television Standards
Committee'. It is from the very early days of electronically scanned TV
in the United States.

Indeed. PAL and SECAM are acronyms for the description of how the
systems work, and not for the organisation that devised the system.
--
Ian

 

[Rob]

Silver Dream ! <email@domain.com> wrote:

It was better in terms of colour fidelity and resolution. It was worse
in terms of flickering, due to lower refresh rate and some el cheapo
decoders being employed too often.

The refresh rate has nothing to do with the color standard (NTSC vs PAL),
it is because of the different mains frequency in the countries that used
the standard, and the advantage of approximately synchronizing the vertical
refresh with the mains frequency on older receivers.

In Brazil, PAL is (was) used with the same refresh rate as was used
for NTSC in the USA.

 

[Rob]

Jeff Urban <jurb6006@gmail.com> wrote:
> From what I've read PAL is very similar to NTSC. It has a scheme with a comb filter that makes a phase (tint) control unnecessary, 50Hz frame rate rather than 60 and a higher horizontal rate resulting in more lines in the raster. (your spelling says you are across the pond)

The different frame rate is not an attribute of PAL vs NTSC, but of the
underlying B/W standard that was optimized for the local mains frequency.

PAL is used with ~60Hz frame rate in Brazil. NTSC with 50 Hz frame rate
is possible, but there just happened to be no country where a 50 Hz mains
frequency was in use and where NTSC color was selected.

 

[Phil Allison]

Rob wrote:
--------------

The refresh rate has nothing to do with the color standard (NTSC vs PAL),
it is because of the different mains frequency in the countries that used
the standard, and the advantage of approximately synchronizing the vertical refresh with the mains frequency on older receivers.

** Early tube based ( B&W ) receivers were locked to the supply frequency as were the local TV transmitters. This simplified PSU design and prevented hum bars rolling down the screen. Later this was changed to non locked.

60Hz was had an inherent problem cos it conflicted with frames rates used by film cameras. 50 Hz was far better since is was close to double the 24 fps rate used for movies - so on TV they ran slightly faster than in a cinemas.

60Hz TV has to use tricks (repeated frames) to play 24fps movies and can look jerky - even today.


..... Phil

 

[Silver Dream !]

On 2020-03-18 19:56:39 +0000, Rob said:

It was better in terms of colour fidelity and resolution. It was worse
in terms of flickering, due to lower refresh rate and some el cheapo
decoders being employed too often.

The refresh rate has nothing to do with the color standard (NTSC vs PAL),
it is because of the different mains frequency in the countries that used
the standard, and the advantage of approximately synchronizing the vertical
refresh with the mains frequency on older receivers.

True - refresh rate is defined in other standards, which define also
number of lines.

see Rec. ITU-R BT.470-6 p. 2
That was an inaccurate mental shortcut from my side. Caused by the fact
that I live on the other side of the pond, where "M" norm (525/60) you
refer to wasn't used. OTOH NTSC isn't used with anything else than "M".
This opens the trap I fell into and inaccurately combined PAL
characteristics with non-M refresh rate/resolution, even if I myself
corrected other people on occasions before.

In Brazil, PAL is (was) used with the same refresh rate as was used
for NTSC in the USA.

Right - PAL/M

--
SD!