color subcarrier

D

Dan

Guest
Hi,

is it possible to synchronize in phase an 3.579545MHz oscillator *only* from
one edge of the horizontal syncs to regenerate the color subsucarrier
instead of using the color burst?
I've read that horizontal syncs are in phase with the subcarrier every two
video lines (NTSC) so the syncs could be used every two lines. Assuming
these syncs are clean of course.
I wonder if it is conceivable in practice or only in theory?

Dan



---
Outgoing mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.637 / Virus Database: 408 - Release Date: 2004-03-20
 
<phil-news-nospam@ipal.net> a écrit dans le message de
news:c47gt511e84@enews2.newsguy.com...
On Sun, 28 Mar 2004 14:04:11 -0500 Dan <dan@nonono.no> wrote:

| is it possible to synchronize in phase an 3.579545MHz oscillator *only*
from
| one edge of the horizontal syncs to regenerate the color subsucarrier
| instead of using the color burst?
| I've read that horizontal syncs are in phase with the subcarrier every
two
| video lines (NTSC) so the syncs could be used every two lines. Assuming
| these syncs are clean of course.
| I wonder if it is conceivable in practice or only in theory?

Yes, it can be done. You just have a somewhat narrow range for it.
This will let you sync the frequency, but not the phase. The exact
phase relationship between the sync edge and the subcarrier is not
defined. Since the sync pulses are derived from a common frequency
as the subcarrier (often 4x), the relationship should be fixed, so
whatever phase you get, it should at least be stable. When color TV
first came out, locking phase was not all that good so there was a
"tint" control which biased the subcarrier phase. You'll have to
make similar adjustments to get the right phase relationship, and
it will vary per channel.
Click to expand...
So it's worth the try afterall.
If I can lock to any stable phase, it's good enough. I expected anyway to
add a phase (tint) ajustment as the logic circuitry will also add to the
unpredictable phase shift.

Regards!
Dan

--
--------------------------------------------------------------------------
---
| Phil Howard KA9WGN | http://linuxhomepage.com/
http://ham.org/ |
| (first name) at ipal.net | http://phil.ipal.org/
http://ka9wgn.ham.org/ |
--------------------------------------------------------------------------
---
Click to expand...
 
This is possible, but complex. In modern NTSC video systems (I know that
there is also a spec for PAL, but I never worked in PAL systems), there is a
fixed relationship between the subcarrier and the horizontal sync pulses.
Leitch Video used to make sync generators that were SCH locked. They would
compare the leading edge of the horizontal sync pulses, in relation to the
subcarrier phase to determine the proper locking. This was referred to an
SCH genlock. The Leitch Frame Synchronizers (analog type) are using SCH
reference locking for the external gen lock.

Before 1970, the SCH was not considered an important spect of video. When
very precise colour frame video editing came along, this became an important
issue.

I cannot remember the exact waveform display specifications for this in the
terms of the subcarrier to sync relation ship. It has something to do with
taking reference to the first positive cycle of burst that is at 50% or
more of the total burst amplitude, and looking back the equivalent time base
of 19 burst cycles. The leading edge of the horizontal sync pulse should be
within +-15 degrees of phase to be considered within specs of the SCH
phase. Do not quote me on this, I am going from memory from a long time
ago!

There are TV waveform scopes that will take the SCH measurement, and display
it in a vector format to show the SCH phase position. Tektronix, Leader, and
HP have Vector/Waveform monitors that will take these readings. Leitch Video
also made a 19 inch, 1 rackspace high, rackmount SCH meter with an LED
bargraph, to read out the SCH phase.

--

Greetings,

Jerry Greenberg GLG Technologies GLG
=========================================
WebPage http://www.zoom-one.com
Electronics http://www.zoom-one.com/electron.htm
=========================================


"Dan" <dan@nonono.no> wrote in message
news:AkF9c.20851$KQ1.299676@weber.videotron.net...
Hi,

is it possible to synchronize in phase an 3.579545MHz oscillator *only* from
one edge of the horizontal syncs to regenerate the color subsucarrier
instead of using the color burst?
I've read that horizontal syncs are in phase with the subcarrier every two
video lines (NTSC) so the syncs could be used every two lines. Assuming
these syncs are clean of course.
I wonder if it is conceivable in practice or only in theory?

Dan



---
Outgoing mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.637 / Virus Database: 408 - Release Date: 2004-03-20
 
What I would like to try is to sync an oscillator with a brief pulse (half
3.58MHz cycle or 130nS) from the leading edge of horz sync, every two lines.
If the oscillator does sync (it remains to be seen), it would most probably
not be in phase with the wanted subcarrier.
The oscillator could be, say 59 degres out of phase (or any other value) but
always the same out of phase value for a given video source. This phase
could then be compensated by an adjustment like a variable low-pass. Of
course, this is only theory, I still have to try this...

Dan

"Jerry G." <jerryg50@hotmail.com> a écrit dans le message de
news:c48crq$m5l$4@news.eusc.inter.net...
This is possible, but complex. In modern NTSC video systems (I know that
there is also a spec for PAL, but I never worked in PAL systems), there is
a
fixed relationship between the subcarrier and the horizontal sync pulses.
Leitch Video used to make sync generators that were SCH locked. They would
compare the leading edge of the horizontal sync pulses, in relation to the
subcarrier phase to determine the proper locking. This was referred to an
SCH genlock. The Leitch Frame Synchronizers (analog type) are using SCH
reference locking for the external gen lock.

Before 1970, the SCH was not considered an important spect of video. When
very precise colour frame video editing came along, this became an
important
issue.

I cannot remember the exact waveform display specifications for this in
the
terms of the subcarrier to sync relation ship. It has something to do with
taking reference to the first positive cycle of burst that is at 50% or
more of the total burst amplitude, and looking back the equivalent time
base
of 19 burst cycles. The leading edge of the horizontal sync pulse should
be
within +-15 degrees of phase to be considered within specs of the SCH
phase. Do not quote me on this, I am going from memory from a long time
ago!

There are TV waveform scopes that will take the SCH measurement, and
display
it in a vector format to show the SCH phase position. Tektronix, Leader,
and
HP have Vector/Waveform monitors that will take these readings. Leitch
Video
also made a 19 inch, 1 rackspace high, rackmount SCH meter with an LED
bargraph, to read out the SCH phase.

--

Greetings,

Jerry Greenberg GLG Technologies GLG
=========================================
WebPage http://www.zoom-one.com
Electronics http://www.zoom-one.com/electron.htm
=========================================
Click to expand...
 
On Mon, 29 Mar 2004 18:16:05 -0500 Dan <dan@nonono.no> wrote:

| What I would like to try is to sync an oscillator with a brief pulse (half
| 3.58MHz cycle or 130nS) from the leading edge of horz sync, every two lines.
| If the oscillator does sync (it remains to be seen), it would most probably
| not be in phase with the wanted subcarrier.
| The oscillator could be, say 59 degres out of phase (or any other value) but
| always the same out of phase value for a given video source. This phase
| could then be compensated by an adjustment like a variable low-pass. Of
| course, this is only theory, I still have to try this...
|
| Dan

How about this:

Syncronize with the vertical sync pulses, and produce a gating waveform
that removes the vertical sync pulses, and the horizontal equalizing
pulses that come just before and after the vertical sync pulses. Then
clip off just the sync pulse tips, gate off the vertical parts of that
waveform, and feed it through a narrow crystal filter at 7.159090909
MHz. Then do the PLL thing from that filter output. As long as the
video you get us very well timed to begin with (plus or minus 15 Hz),
this should give you the 455th harmonic of the horizontal sync pulse.
Now divide by 2 to get 3.579545454 MHz. The narrower your filter is,
the more stable this would be, but more limited is the range that can
be accepted, too.

--
-----------------------------------------------------------------------------
| Phil Howard KA9WGN | http://linuxhomepage.com/ http://ham.org/ |
| (first name) at ipal.net | http://phil.ipal.org/ http://ka9wgn.ham.org/ |
-----------------------------------------------------------------------------
 
On Mon, 29 Mar 2004 18:16:05 -0500, Dan wrote:

What I would like to try is to sync an oscillator with a brief pulse (half
3.58MHz cycle or 130nS) from the leading edge of horz sync, every two lines.
If the oscillator does sync (it remains to be seen), it would most probably
not be in phase with the wanted subcarrier.
The oscillator could be, say 59 degres out of phase (or any other value) but
always the same out of phase value for a given video source. This phase
could then be compensated by an adjustment like a variable low-pass. Of
course, this is only theory, I still have to try this...

Dan
Click to expand...
One more thing to watch out for, is your video source stable? If it is
comming from a consumer-type video tape machine, the sync pulses will
have a "wobble" due to slight variations in tape speed and drum speed.
The VTR processes the color information to make it stable enough for a
monitor, but then there is no relationship between the sync pulses and
the color information. Most monitors (and TVs) can handle a fair
amount of instability in the sync timing, but the color subcarrier has
to be kept fairly close to specification.

--

Tony Newman <tonyn@efn.org>
Springfield, OR
USA
 
You must log in or register to reply here.

Welcome to EDABoard.com

Sponsor

Back
Top