[net.video] Help w/VHS slow dubbing

d@alice.UucP (Dan Rosenberg) (02/25/86)

I have a slight video problem that I need a quickie fix for:
	I am, along with six other students, producing a video tape on various
subjects in physics. Several scenes MUST be in slow motion. All video was shot
on a 1/2" VHS deck (Panasonic PV-8000) then dubbed up to 3/4" U-matic format.
The scenes that had to be in slow motion were then dubbed up using the frame by
frame "slow" advance feature of the PV-8000. In viewing the 3/4" tape, and the
3/4" composite master, we were quite pleased. Cheap, crisp, clear slow motion
video. 
	Problem: The tape is to be distributed on 1/2" (i.e. for our home
collection, school's collection, and so forth and so on...) VHS. When we dub
down to 1/2" (not to the PV-8000, but just a 3-4 year old Industrial Quality SP
only Panasonic VHS) things begin to look funny. All of the standard play video
dubs fine, but the slow motion "rolls" a bit. It is not the type of rolling
that you get when you have the Vert. Hold set wrong on a monitor, but each
frame (of slow motion original video, not each frame of Composite 3/4") moves
down the screen as it is being played.
	I realize that our 3/4" editors clean up the video a bit, but things
look very bad on when played back on 1/2". There is no "video cheese" (our name
for cruddy looking video w/bad sync, or bad what have you...) on the 1/2" tape,
things hold together ... but there is the annoying roll.
	The project is wrapping up, and we will be ready to go in a few weeks,
so what I really need is a fairly easy (and CHEAP, i.e. no major outlay at all)
way to stabilize the video. We have a good deal of equipment, but don't know
exactly what to do.
	Idea: We have a sync generator. BB from this to 3/4" deck. Set 3/4" deck
to external sync. I would rather not do this unless I had a fairly good idea
that it would work. It would involve disruption of normal studio 
though.
	Please mail replys directly to:
			..!ihnp4!charm!kalucki
		or if you are on an Bell Labs machine in NJ:
			charm!kalucki

p.s. I am posting this through a friend.

brown@nicmad.UUCP (02/27/86)

In article <5031@alice.uUCp> d@alice.UucP (Dan Rosenberg) writes:
>I have a slight video problem that I need a quickie fix for:
>	I am, along with six other students, producing a video tape on various
>subjects in physics. Several scenes MUST be in slow motion. All video was shot
>on a 1/2" VHS deck (Panasonic PV-8000) then dubbed up to 3/4" U-matic format.
>The scenes that had to be in slow motion were then dubbed up using the frame by
>frame "slow" advance feature of the PV-8000. In viewing the 3/4" tape, and the
>3/4" composite master, we were quite pleased. Cheap, crisp, clear slow motion
>video. 
>	Problem: The tape is to be distributed on 1/2" (i.e. for our home
>collection, school's collection, and so forth and so on...) VHS. When we dub
>down to 1/2" (not to the PV-8000, but just a 3-4 year old Industrial Quality SP
>only Panasonic VHS) things begin to look funny. All of the standard play video
>dubs fine, but the slow motion "rolls" a bit. It is not the type of rolling
>that you get when you have the Vert. Hold set wrong on a monitor, but each
>frame (of slow motion original video, not each frame of Composite 3/4") moves
>down the screen as it is being played.

I believe you have what is known as a time-base error.  Someone on the net
will correct me if I am wrong, but basically what is the matter is that the
slow motion stuff that you get recorded on the 3/4" machine isn't exactly
NTSC standard anymore.  When the machine is placed in slow motion, the
vertical sync area has a little something added to it.  You're lucky that
the 3/4" deck will record it, the one I have doesn't always.  When you played
it back to the other VHS machine, the bastardized vertical sync probably
caused the recorder to not record correctly.  Try going from the 3/4" deck
to the newest VHS table-top unit you can find.  They tend to be more tollerant
of bad vertical sync.

To fix the problem, you need to send the output of the VHS machine through
a TBC unit, that will sync itself up to the incoming video, then record
the slow motion stuff on 3/4".

I don't do any slow-motion recording/editing at home, because of the above
reasons.
-- 

              ihnp4------\
            harvard-\     \
Mr. Video      seismo!uwvax!nicmad!brown
              topaz-/     /
             decvax------/

dsi@unccvax.UUCP (Dataspan Inc) (02/27/86)

     The problem is that during slow advance on your PV8000, the relationship
of the equalisation pulses to the vertical sync pulse is destroyed. I doubt
that your sync generator will genlock to the vertical sync pulse which is 
provided during slow motion and stop frame reproduction. The reason that this
is done is to essentially force the synchronisation at the monitor to 
"noninterlaced" (only one field is being displayed at a given time) and because
there is no control track to provide reconstitution of the vertical sync
during normal playback.  (Note also that the "dead zone" occurs during the
vertical blanking interval).

     The solution is to get a time base corrector. Most of them which are
designed for use with the heterodyne VTR send an advanced V-sync tach signal
to the VTR.  The advanced signal is used to precorrect the headwheel phase
and the control track is simply ignored.  Remaining timebase errors are 
removed by digital techniques.  Now, when you go into slow/stop motion,
the sync which will be on the TBC out will have the correct synchronizing
waveforms impressed on it.  However, still frame reproduction will actually
"jitter" now because the TBC out signal is always interlaced.

     In short, when you go into slow/stop motion, the signal isn't NTSC 
anymore, and your 1/2 inch machines (utilizing digital techniques to find
sync pulses, etc) don't know what to do with non NTSC sync pulses. Try 
taping a cheap "Pong" game or something to test the effect.

     What this world needs is 1) a GOOD but CHEAP time base corrector with,
say, a 32 line window and 2) VHS and Beta machines with provisions to 
disconnect the control track servo and provide advanced Vsync tach signal
to the headwheel servo comparator.  Everyone seems to be using video these
days, but no one seems to have TBC's.  You don't need a $15,000 Tektronix
110 TBC (which is extremely nice, by the way) to do what you want to do,
but the manufacturers haven't gotten the message yet.

David Anthony
DataSpan, Inc

brown@nicmad.UUCP (03/02/86)

In article <451@unccvax.unccvax.UUCP> dsi@unccvax.UUCP (Dataspan Inc) writes:
>
>     The problem is that during slow advance on your PV8000, the relationship
>of the equalisation pulses to the vertical sync pulse is destroyed. I doubt
>that your sync generator will genlock to the vertical sync pulse which is 
>provided during slow motion and stop frame reproduction. The reason that this
>is done is to essentially force the synchronisation at the monitor to 
>"noninterlaced" (only one field is being displayed at a given time) and because
>there is no control track to provide reconstitution of the vertical sync
>during normal playback.  (Note also that the "dead zone" occurs during the
>vertical blanking interval).

I beg to differ.  You have the right idea, but some bad information.  The
funny vertical sync pulse, that is provided by the VHS VCR during special
effects playback, is there to help mask some of the noise that can occur.
The added sync pulse helps the tv from 'rolling' during special effects.
But, this added sync makes it different from the normal NTSC sync.  VCRs
don't really care if the information being recorded is fully interlaced
or not.  All they are looking for is the vertical sync, in order the phase
the heads correctly, during recording.

The control track is ALWAYS recorded on the tape and it is ALWAYS referenced
during playback.  That is how the machine knows where to place the rotating
heads.  During special effects, the head placement can be even more critical.
Though some machines are better than others.  The ones that provide the
'clear picture' special effects use the control track to place the heads
more accurately.

The vertical sync is not reconstituted by the VCR, during playback.  It is
part of the original signal that was recorded.  Because the vertical sync
was damaged during special effects, following VCRs may have trouble
recording it, which is exactly what happened.

If VCRs recreated the vertical sync (or even the complete vertical blanking
area), then we wouldn't have tape duplicators recording "copy guard" stuff
on the tape, as the VCR would throw it out.  I wish that the VCRs did do
that, but they don't.

>     The solution is to get a time base corrector. Most of them which are
>designed for use with the heterodyne VTR send an advanced V-sync tach signal
>to the VTR.  The advanced signal is used to precorrect the headwheel phase
>and the control track is simply ignored.  Remaining timebase errors are 
>removed by digital techniques.  Now, when you go into slow/stop motion,
>the sync which will be on the TBC out will have the correct synchronizing
>waveforms impressed on it.  However, still frame reproduction will actually
>"jitter" now because the TBC out signal is always interlaced.

Sending a VCR sync information, makes that VCR give the TV signal in phase
with the incoming sync.  Again the control track is used.  But this time,
the VCR speeds up or slows down the tape, so that it matches the incoming
sync.

>     In short, when you go into slow/stop motion, the signal isn't NTSC 
>anymore, and your 1/2 inch machines (utilizing digital techniques to find
>sync pulses, etc) don't know what to do with non NTSC sync pulses. Try 
>taping a cheap "Pong" game or something to test the effect.

Not all VCRs have trouble recording non NTSC standard sync.  One of my
machines will record the information correctly, as long as the sync
information looks even close.  Which is why the old style "copy guard"
doesn't work on my machine.  The new "macrovision" does, but that is
another story.

>     What this world needs is 1) a GOOD but CHEAP time base corrector with,
>say, a 32 line window and 2) VHS and Beta machines with provisions to 
>disconnect the control track servo and provide advanced Vsync tach signal
>to the headwheel servo comparator.  Everyone seems to be using video these
>days, but no one seems to have TBC's.  You don't need a $15,000 Tektronix
>110 TBC (which is extremely nice, by the way) to do what you want to do,
>but the manufacturers haven't gotten the message yet.

As I said above, the CONTROL TRACK IS ALWAYS used.  The VCR NEEDS to know
where the video information was laid down on the tape.  The control track
provides that information.  External sync just makes the VCR change tape
speed to match the incoming sync.  If you were to tell the VCR, ANY VCR,
to ignore the control track, you WILL end up with garbage on your TV.

Last point to remember,; if you don't give a VCR external sync, it uses
an internal referenced sync.

I suggest you read the ads in 'Videography', as there are TBCs for less
than $4000.00.  True, for me that is still too expensive, for home use.
-- 

              ihnp4------\
            harvard-\     \
Mr. Video      seismo!uwvax!nicmad!brown
              topaz-/     /
             decvax------/