| Author |
Message |
Paul Keinanen
Guest
|
 Posted:
Sat Oct 30, 2004 12:10 pm Post subject:
Re: Origins of PAL: 1956 radio engeenering airticle from UK |
|
|
On Fri, 29 Oct 2004 23:19:46 +0000 (UTC), toor@iquest.net (John S.
Dyson) wrote:
| Quote: | In article <634e8e054d.news@lightfox.plus.com>,
Gareth Rowlands <gareth.see_sig_block@127.0.0.1> writes:
1) Of the distinction between electrical and optical gamut.
Actually, I have been fully aware of the lack of optical gamut on
many kinds of consumer grade display devices. It is interesting
that alot of people are stuck-up thinking about phosphors, when
many common devices don't use phosphors nowadays.
|
<clip>
| Quote: | Out of interest, does any reader know in which decade US manufacturers
stopped fitting tubes with 1953 NTSC primaries to their sets ?
Remember, high end TV sets in the US tend not to use phosphors at all.
|
Phosphors are quite narrow band emitters as such can be selected to
stimulate only the "right" kind of cone cells in the eye.
When using broadband emitters (e.g. when splitting white light into
blue, green and red in projectors and LCDs), the extremes of the
spectral response are also going to stimulate the non-intended cone
cells.
Aren't you going to suffer some colour purity problems ?
Paul
|
|
| Back to top |
|
 |
Roderick Stewart
Guest
|
| Posted:
Sat Oct 30, 2004 2:19 pm Post subject:
Re: Origins of PAL: 1956 radio engeenering airticle from UK |
|
|
In article <clv494$2msa$1@news.iquest.net>, John S. Dyson wrote:
| Quote: | Feed the same RGB values into either system
and you'll get the same RGB values out.
Actually, if you push the limits you'll NOT get exactly the same values.
If you stay very safe in the middle, then what you say is more true.
Trying to devolve the argument into 0-255 for each component (or whatever
legal range) ignores the fact that the colors are different, and ignores
the effects of the mapping between RGB <-> YIQ/YUV.
|
This is irrelevant. Whatever mapping takes place in the encoder between the
RGB input signals and YIQ or YUV is completely reversed in the decoder so you
get exactly the same voltage levels out. (This applies to still images of
course, as there are slight discrepancies as soon as anything moves, but this
applies to both systems).
| Quote: | What color
is your Green primary, assuming that you aren't limited to cheap
green phosphors (like many display devices in the US aren't limited.)
What happens when you aren't limited to a reddish orange 'red', and
actually have a better color for your color filter for your display?
|
The colours displayed by the display are a function of the display, not the
electronic encoding/decoding system. Feed the same RGB values into your
display from the output of *any* decoding system, and you will see the same
colours displayed. Your greens may be brighter than your neighbour's greens if
you use different types of display device, but this will apply even if you're
watching the same programmes from the same source and using the same type of
decoding system, so it is not the decofding system that is giving you brighter
greens.
This leaves aside the question of whether your brighter greens are more
accurate greens, or simply brighter. They can only be more accurate if the
signals came from a camera with the appropriate filters and matrixing.
| Quote: | Please refer to the effects of the matrix, the representable colors and the
nonlinear effects before making that claim. If you cannot represent
the color, then no matter how hard you try,
|
If you mean the linear matrix, then this is part of the camera, which still
produces RGB output signals with the same voltage range to the encoder. In
theory, if you can guarantee that the only programmes you watch on your
display device with its superior green have come from cameras with superior
green filtering and appropriate linear matrixing, then you will see more
accurate representations of green objects. But this is regardless of any
transparent RGB-to-RGB encoding/decoding that the signals have undergone along
the way, and how often is it really possible to make that guarantee nowadays?
How many production companies do you think even know what camera primaries or
linear matrixing coefficients they are using, never mind using different ones
for programmes destined for PAL and NTSC countries?
Rod. |
|
| Back to top |
|
|
Gareth Rowlands
Guest
|
| Posted:
Sun Oct 31, 2004 4:42 pm Post subject:
Re: Origins of PAL: 1956 radio engeenering airticle from UK |
|
|
In message <cluj6i$2hs2$1@news.iquest.net>
John S. Dyson wrote:
| Quote: | 1) Of the distinction between electrical and optical gamut.
Actually, I have been fully aware of the lack of optical gamut on
many kinds of consumer grade display devices.
|
I don't have a problem with your concepts of the 'purity' of display
colour emitters. Where I think you are selling yourself short are those
statements you make where you appear to be confusing the realms of
electrical and optical gamut such as in:
| Quote: | and PAL does impose such a limit, but so do low end (too-common)
CRT phosphors.
|
and
| Quote: | Geesh, a high quality SMPTE phosphor can reproduce the limited PAL
gamut just fine.
|
and
| Quote: | (along with a TV system that can provide the wider gamut?)
|
because you are making it seem as if the properties of electrical
gamut are the same as those of optical gamut.
| Quote: | There are definitely differences in terminology across the pond.
|
How about "incomplete decoding" ? Why leave the reader struggling to
comprehend about what is 'incomplete' when what is really being talked
about is really 'unwanted additions', or artefacts ?
If you are seeing something that shouldn't really be there, such as
luminance being mistaken by the decoder for chrominance information,
have the courage to stand up and say so.
Cheers,
Gareth.
--
http://www.rat.org.uk gareth at lightfox dot plus dot com |
|
| Back to top |
|
|
Gareth Rowlands
Guest
|
| Posted:
Sun Oct 31, 2004 4:48 pm Post subject:
Re: Origins of PAL: 1956 radio engeenering airticle from UK |
|
|
In message <clv56n$2n5t$1@news.iquest.net>
John S. Dyson wrote:
| Quote: | PAL did an excellent job of reproducing the color errors
as created in the PAL studio. The same problem would happen on NTSC.
It is much more likely to have technical errors in the studio
than for an NTSC signal hue only to be corrupted.
|
Out of interest, where would you think the most likely place(s) for an
NTSC hue error to happen is/are ?
Cheers,
Gareth.
| Quote: | It is more likely that there is a problem with studio equipment than
only the hue being distorted on NTSC.
|
--
http://www.rat.org.uk gareth at lightfox dot plus dot com |
|
| Back to top |
|
|
Gareth Rowlands
Guest
|
| Posted:
Sun Oct 31, 2004 5:17 pm Post subject:
Re: Origins of PAL: 1956 radio engeenering airticle from UK |
|
|
In message <clva5s$2o9g$1@news.iquest.net>
John S. Dyson wrote:
| Quote: | You aren't really answering the comment about CPA -- do you know what
it is?
|
'Color Phase Alternation' - from what I understand, it was a field-by-
field method of axis swiching as distinct from the PAL line-by-line
V axis switch.
| Quote: | Hint: it is essentially the major difference between PAL and NTSC,
and was EXPLICITLY declined by the NTSC team.
|
I think they declined the use of field-by-field CPA because of flicker
effects. Is it true the guys at Philco came up with the idea of 525 PAL
in 1951 but it was never tried out in practice ?
Cheers,
Gareth.
--
http://www.rat.org.uk gareth at lightfox dot plus dot com |
|
| Back to top |
|
|
Doug McDonald
Guest
|
| Posted:
Sun Oct 31, 2004 7:49 pm Post subject:
Re: Origins of PAL: 1956 radio engeenering airticle from UK |
|
|
Gareth Rowlands wrote:
| Quote: | How about "incomplete decoding" ? Why leave the reader struggling to
comprehend about what is 'incomplete' when what is really being talked
about is really 'unwanted additions', or artefacts ?
|
Incomplete decoding means literally what it says.
A truly complete decode retrieves the original
luma and chroma components uncontaminated by cross-effects.
There is no struggle to comprehend what it means for
by any ordinary human.
An incomplete decode leaves in artefacts.
Doug McDonald |
|
| Back to top |
|
|
Doug McDonald
Guest
|
| Posted:
Sun Oct 31, 2004 7:51 pm Post subject:
Re: Origins of PAL: 1956 radio engeenering airticle from UK |
|
|
Gareth Rowlands wrote:
| Quote: |
Out of interest, where would you think the most likely place(s) for an
NTSC hue error to happen is/are ?
|
There is zero question about that ... in the
post-production process at the hands of the producer,
or, sometimes, on the stage itself at the hands of
the lighting people, who are told by the producer
to put color gels on the lighting. The producer
gets what they want.
Doug McDonald |
|
| Back to top |
|
|
Paul Keinanen
Guest
|
| Posted:
Sun Oct 31, 2004 9:14 pm Post subject:
Re: Origins of PAL: 1956 radio engeenering airticle from UK |
|
|
On Sun, 31 Oct 2004 08:49:50 -0600, Doug McDonald
<mcdonald@SnPoAM_scs.uiuc.edu> wrote:
| Quote: | Incomplete decoding means literally what it says.
A truly complete decode retrieves the original
luma and chroma components uncontaminated by cross-effects.
|
I agree that you may completely separate the interleaved luminance and
chrominance sidebands for a stationary image. However, when there is a
movement the position of the sidebands change and on the lower
sideband of the chrominance signal, the sidebands move in the opposite
direction relative to the luminance sidebands.
With suitable movement, some of the luminance and chrominance
sidebands fall exactly on each other. How do you expect to extract
this information completely ?
While you might be able to make some educated guesses about these
destroyed sidebands based on other non-overlapped sidebands, these are
still guesses and some information is lost. Even if this information
is used to suppress those overlapped sidebands, so this is not a
complete decoding.
Paul |
|
| Back to top |
|
|
Doug McDonald
Guest
|
| Posted:
Mon Nov 01, 2004 12:56 am Post subject:
Re: Origins of PAL: 1956 radio engeenering airticle from UK |
|
|
Paul Keinanen wrote:
| Quote: | On Sun, 31 Oct 2004 08:49:50 -0600, Doug McDonald
mcdonald@SnPoAM_scs.uiuc.edu> wrote:
Incomplete decoding means literally what it says.
A truly complete decode retrieves the original
luma and chroma components uncontaminated by cross-effects.
I agree that you may completely separate the interleaved luminance and
chrominance sidebands for a stationary image. However, when there is a
movement the position of the sidebands change and on the lower
sideband of the chrominance signal, the sidebands move in the opposite
direction relative to the luminance sidebands.
With suitable movement, some of the luminance and chrominance
sidebands fall exactly on each other. How do you expect to extract
this information completely ?
|
If the motion is changing, you can't do it easily.
If the motion is constant and slow, it can be done
very will with the commonplace Faroujda chips in
most medium to high end TVs. These use the information
from three points in time.
They work very well indeed.
Doug McDonald |
|
| Back to top |
|
|
Jim Easterbrook
Guest
|
| Posted:
Mon Nov 01, 2004 10:44 am Post subject:
Re: Origins of PAL: 1956 radio engeenering airticle from UK |
|
|
In sci.engr.television.advanced John S. Dyson <toor@iquest.net> wrote:
| Quote: |
Which ones are those? There are definitely differences in terminology
across the pond. For example, the patent that I heard about was a
PAL decoder, but the patent that I read was a chroma demodulator.
|
I thought it was for a chroma/luma separator, as that's the hard bit,
and is the inventive step the patent covers. I didn't attempt to patent
the rest of a decoder that is well known to those skilled in the art.
The patent does include a simple block diagram of a complete decoder
embodying the invention.
I'm pleased to read (in another of your posts) that you have got an NTSC
decoder that's as good as the PAL decoder described in my patent. I look
forward to seeing your patent application.
| Quote: | And finally -- PAL is still not nearly fully decoded in the consumer
(and even most often in the professional environment.) I was amazed
to see the cross color on the DTV broadcast (equivalent to a low end
NTSC decoder in quality.) Admittedly, in the US, there are TV stations (and
even networks) that sometimes use low end decoders, but at least there
is the commonly accessable choice to fully decode NTSC.
|
It is hard to make comparisons across systems, countries, TV stations
and programme genre like this. You really have to seek out the most
testing material for each system and then try to decode it. For example,
in PAL we have found the "mobile and calendar" sequence used in the MPEG
community to be really testing as it contains objects moving at the PAL
critical speeds. By comparison, I can spend an entire evening watching
PAL TV at home and not see a single artefact. A lot of contemporary
material is just too soft to cause cross colour. I expect this is not
the case with NTSC as the sub carrier has a lower spatial frequency.
This might be why advanced decoders are so much more common in NTSC
domestic equipment.
--
Jim Easterbrook
BBC Research & Development <http://www.bbc.co.uk/rd/>
*** All opinions are mine and might not be shared by the BBC *** |
|
| Back to top |
|
|
R. Mark Clayton
Guest
|
| Posted:
Mon Nov 01, 2004 11:53 am Post subject:
Re: Origins of PAL: 1956 radio engeenering airticle from UK |
|
|
"Jim Easterbrook" <jim.easterbrook@rd.bbc.co.uk> wrote in message
news:cm5435$a4r$1@nntp0.reith.bbc.co.uk...
| Quote: | In sci.engr.television.advanced John S. Dyson <toor@iquest.net> wrote:
snip |
| Quote: | I'm pleased to read (in another of your posts) that you have got an NTSC
decoder that's as good as the PAL decoder described in my patent.
|
What number is that? Any chance of an abstract?
| Quote: | I look forward to seeing your patent application.
|
Probably the same as yours. You see in the UK any prior art invalidates a
petent, whereas in the USA it has to be over a year before.
| Quote: | --
Jim Easterbrook
BBC Research & Development <http://www.bbc.co.uk/rd/
*** All opinions are mine and might not be shared by the BBC *** |
|
|
| Back to top |
|
|
Jim Easterbrook
Guest
|
| Posted:
Mon Nov 01, 2004 12:28 pm Post subject:
Re: Origins of PAL: 1956 radio engeenering airticle from UK |
|
|
In sci.engr.television.advanced R. Mark Clayton <nospamclayton@btinternet.com> wrote:
| Quote: |
"Jim Easterbrook" <jim.easterbrook@rd.bbc.co.uk> wrote in message
news:cm5435$a4r$1@nntp0.reith.bbc.co.uk...
I'm pleased to read (in another of your posts) that you have got an NTSC
decoder that's as good as the PAL decoder described in my patent.
What number is that? Any chance of an abstract?
|
The European patent number is EP1175102. You can read the whole thing
at http://l2.espacenet.com/espacenet/viewer?PN=EP1175102
--
Jim Easterbrook
BBC Research & Development <http://www.bbc.co.uk/rd/>
*** All opinions are mine and might not be shared by the BBC *** |
|
| Back to top |
|
|
Robert Orban
Guest
|
| Posted:
Thu Nov 11, 2004 4:09 am Post subject:
Re: Origins of PAL: 1956 radio engeenering airticle from UK |
|
|
In article <10mrgmgcfatne65@news.supernews.com>, pete@rgb.com says...
| Quote: |
"R. Mark Clayton" <nospamclayton@btinternet.com> wrote in message
news:ckkcun$av4$1@sparta.btinternet.com...
Like [the UK] had quadriphonic radio broadcasts in the seventies.
And that's someting to be proud of?
Did you ever listen to them?
|
You are misquoting Mr. McDonald, I do believe. The _US_ had
a quadraphonic broadcasting system approved in the 1970s after an FCC
rulemaking procedure that depended on third-party test results of the
various contenders' systems. IIRC, there were four contenders, including
Dorren and Nippon Columbia (I forget who the other two were, although
one may have been CBS Laboratories).
These were not matrix (like SQ and QS) systems -- they were entirely
discrete, using a second subchannel (at 76 kHz, phase-locked to the 19
kHz pilot tone) with two carriers in quadrature to support the two extra
audio channels. The main difference between the systems was how stereo
and mono compatibility were handled, which depended on the encoding
equations maping the audio channels to the four discrete physical
channels in the FM baseband. The Nippon Columbia system used Duane
Cooper's system of heirarchical directional coding (described by Dr.
Cooper in a number of AES Journal papers at the time), while the Dorren
system was simpler, encoding the two rear channels on the extra
subcarrier.
The FCC finally chose the Dorren system, but, by then, the quad fad has
passed and nothing ever came of the system commercially. But it sounded
just fine, thank you. |
|
| Back to top |
|
|
Aztech
Guest
|
| Posted:
Sat Nov 13, 2004 6:03 pm Post subject:
Re: Origins of PAL: 1956 radio engeenering airticle from UK |
|
|
"Robert Orban" <donotreply@spamblock.com> wrote in message
news:trCdnXtJRZglTg_cRVn-uw@giganews.com...
| Quote: | In article <10mrgmgcfatne65@news.supernews.com>, pete@rgb.com says...
The FCC finally chose the Dorren system, but, by then, the quad fad has
passed and nothing ever came of the system commercially. But it sounded
just fine, thank you.
|
Given the way most US broadcasters (and now everyone else) like to misuse
certain equipment *cough*, it's apparent quality isn't exactly their primary
concern.
However, Bayern Radio and Swedish Radio send out very nice DD5.1 streams via
DVB-s, and also a 1.5Mbps DTS stream in the latter case, which is all very nice.
Az. |
|
| Back to top |
|
|
Robert Orban
Guest
|
| Posted:
Wed Nov 17, 2004 6:05 am Post subject:
Re: Origins of PAL: 1956 radio engeenering airticle from UK |
|
|
In article <2fqld.2885403$yk.448898@news.easynews.com>, az@tech.com
bloviated...
| Quote: |
"Robert Orban" <donotreply@spamblock.com> wrote in message
news:trCdnXtJRZglTg_cRVn-uw@giganews.com...
In article <10mrgmgcfatne65@news.supernews.com>, pete@rgb.com says...
The FCC finally chose the Dorren system, but, by then, the quad fad has
passed and nothing ever came of the system commercially. But it sounded
just fine, thank you.
Given the way most US broadcasters (and now everyone else) like to misuse
certain equipment *cough*, it's apparent quality isn't exactly their primary
concern.
|
In fact, the transparency of the quadraphonic channel was PRECISELY the FCC's
concern during the rulemaking. |
|
| Back to top |
|
|
|
|
|
|
FAQ
Memberlist
Register
Profile
Log in to check your private messages
Log in