Geoffrey S. Mendelson <
g...@mendelson.com> writes:
> By the time the US went to color [...] the system that won the "color
> wars" was color on top of an existing black and white signal, so it
> had to use the same specs as it.
>
> They both used a 15.7kHz scan oscilator, so that gave PAL TV sets
> 625 lines of resolution and the US 525. There was a problem with
> interference between the 60 Hz field rate and the 15.7kHz scan, so
> it was moved to 3000/1001 or 29.97 frames per second.
I work in television program production in the USA. Here's something I
posted to a private newsgroup - I can only post my part here, but I was
reposnding to a question about why we're shooting at 23.976fps. As can
be seen, the interference mentioned above concerned the horizontal sync
and color subcarrier frequencies.
-----------------------------------------------------------------------
Perhaps someone is wondering, where the heck did such an odd
frame rate (as opposed to plain old 24.000 frames per second)
come from?
24p was originally suggested for digital television because it's
easy to convert to other formats. But then we have yet to escape
the past when the sync frequencies were tweaked to accomodate tv
as it was broadcast. When you're done reading this short excerpt,
you'll know more about it than 99.9% of the people I have to put
up with in my work -
| *Television Broadcasting: Equipment,*
| *Systems and Operating Fundamentals*
|
| By Harold E. Ennes (Howard W. Sams & Co, 1978)
|
| [Pages 72-74]
|
| 2-10. LUMINANCE AND CHROMINANCE MODULATION LEVELS
|
| The word "compatability" implies that conventional monochrome
| receivers be able to reproduce a color telecast in black and white.
| This requirement immediately fixes the bandwidth for colorcasting
| as no more than the established 6 MHz per television channel.
| Therefore, the addition of the color information must be by a
| method that will not add to the required bandwidth.
|
| The pickup tube in a monochrome camera is scanned at the line
| frequency of 15,750 Hz. When the scanning beam sweeps across the
| target at this frequency, the beam is changed in amplitude in
| accordance with the charge pattern that corresponds to the focused
| image. The resultant rate of voltage changes (frequency of voltage)
| will always be some multiple of the initial scanning rate of 15,750
| Hz. (Fig. 2-17A) [ See
http://MIX.COM/Interleaving.jpeg ] There-
| fore, the major signal components lie at integral multiples of the
| line-scanning rate.
|
| These clusters of signal information decrease in amplitude as
| they get farther from the carrier frequency. Gaps formed between
| the signal components contain no great amount of information at all.
| These _gaps_ occur at _odd multiples of one-half the line frequency._
| Thus, if we generate a subcarrier at some odd multiple of one-half
| the line frequency, subcarrier sidebands will lie in the gaps formed
| by the harmonics of the line frequency (Fig. 2-17B). This process
| is termed "interleaving" in color-TV systems. The color subcarrier
| is placed high in the band, and its maximum-amplitude sidebands
| occur where the monochrome sidebands are small. As the color
| sidebands get farther from the subcarrier and nearer the monochrome
| (main visual) carrier, they decrease in amplitude and produce
| minimum effect in that region, where the monochrome sidebands are
| larger.
|
| Since there are an odd number of lines (525), an odd multiple of
| one-half the line frequency is also an odd multiple of one-half the
| frame rate. A frame is composed of two (interlaced) fields. For
| each field, a point on one line that is made brighter by the color-
| subcarrier sidebands lies directly above a point on the succeeding
| line, which is made darker. When the viewer is far enough from the
| screen that the lines are not distinguishable, this "space-integration"
| effect cancels any brightness variations caused by the presence of
| color sidebands in the composite signal. Also, a "time-integration"
| occurs since the brightness variations in corresponding lines of
| successive frames are 180 degrees out of phase, provided the color
| sidebands fall at exactly one-half multiples of the frame frequency.
| (Since during a frame the color frequency passes throught a whole
| number of cycles plus one-half cycle, the patterns produced by the
| subcarrier in corresponding lines successive frames are one-half
| cycle, or 180 degrees, out of phase.) Since interference between
| the two carriers cancels, it is only necessary for the receiver to
| demodulate each signal with respect to its own carrier.
|
| The frequency of the chrominance subcarrier is 3.579545 MHz. This
| is an odd multiple of half the line frequency for proper multiplexing
| (interleaving) with the luminance signal, as pointed out above. This
| frequency is not so high as to cause interference with the sound
| carrier or restriction of the chrominance bandwidth. At the same
| time, it is not so low that conventional monochrome receivers will be
| visibly affected by chrominance information in nonlinear circuits.
| At frequencies above 3.5 MHz, most monochrome receivers have atten-
| uation of 10 to 25 dB or more.
|
| The chrominance subcarrier frequency was also chosen so as to result
| in minimum beat-frequency interference with the sound carrier. To
| achieve this result, it is necessary that the frequency offset between
| the sound and chrominance carriers also be an odd multiple of one-half
| the line frequency.
|
| Thus, the necessary choice of the chrominance subcarrier frequency
| has resulted in a slight change in actual line and frame frequencies
| from those of previous monochrome standards. To make the chrominance
| carrier frequency an odd multiple of half the line rate, the _new_
| line rate becomes 15,734.26 Hz. This is a reduction of about 0.1
| percent, and is well within the range of existing monochrome-receiver
| hold controls. The new field rate then is:
|
| 15,734.26
| --------- = 59.94 Hz
| 525/2
|
| Thus, stations no longer use "line-lock" circuits to lock the field
| rate (and, hence, all synchronizing signals) to the power line frequency
| (60 Hz). The color subcarrier frequency is generated by a crystal
| oscillator, and counters are used to obtain a driving signal for the
| conventional sync generator [...]
A 59.94Hz field rate is a 29.97Hz frame rate. 24p was chosen because
it's easy to convert to other frame rates, but then it got bastardized
down to 23.976 so it can easily go to 29.97. Which is (if not 30.00
itself) where we should be shooting, because it looks better.
But, at 23.976 one can cram an extra ten minutes on a 40 minute tape.
Not at all unlike the way we used to shoot 35mm film using just three
perforations per frame, instead of four. The frame size is smaller,
yes, but no one much cared because they got an extra seven minutes on
a 2000 foot magazine of film.
This is, of course, to say nothing of our still even using tape. That
is almost history now, but only because stock is getting really hard to
even find, heh...
Billy Y..
--
sub #'9+1 ,r0 ; convert ascii byte
add #9.+1 ,r0 ; to an integer
bcc 20$ ; not a number