Maker Pro
Maker Pro

Interesting pic !

P

Phil Allison

Jan 1, 1970
0
Hi to all,


ever wondered how Dolby Digital 5.1 sound is ACTUALLY carried by a 35mm
cinema film ?

Turns out the 320 kbit/s DD data stream is encoded onto squares optical
grids of about 60 x 60 bits & very cunningly fitted in * between* the
sprocket holes !!

http://upload.wikimedia.org/wikipedia/commons/1/1d/35mm_film_audio_macro.jpg

The embedded Dolby logo is cute dontcha think ?

A tiny CCD camera in the projector captures each 2.4mm square image, 96 of
them per second, so it can be processed into 5 channels of audio.

The other tracks you see are Sony SDDS, L/R stereo analogue and timing
pulses for DTS so the CD can be synched with the movie.

Such a multi-format movie can be shown by any cinema in the world and exact
copies are made by ordinary, chemical film processing.

Betcha you ALL knew that already ....




........ Phil
 
B

Bob Parker

Jan 1, 1970
0
Hi to all,


ever wondered how Dolby Digital 5.1 sound is ACTUALLY carried by a 35mm
cinema film ?

Turns out the 320 kbit/s DD data stream is encoded onto squares optical
grids of about 60 x 60 bits & very cunningly fitted in * between* the
sprocket holes !!

http://upload.wikimedia.org/wikipedia/commons/1/1d/35mm_film_audio_macro.jpg

The embedded Dolby logo is cute dontcha think ?

A tiny CCD camera in the projector captures each 2.4mm square image, 96 of
them per second, so it can be processed into 5 channels of audio.

The other tracks you see are Sony SDDS, L/R stereo analogue and timing
pulses for DTS so the CD can be synched with the movie.

Such a multi-format movie can be shown by any cinema in the world and exact
copies are made by ordinary, chemical film processing.

Betcha you ALL knew that already ....


....... Phil


Interesting and educational. There's lots more photos and info at
http://www.maniacpony.pwp.blueyonder.co.uk/onscreen/sound.htm

Bob
 
P

Phil Allison

Jan 1, 1970
0
"Bob Parker"
Interesting and educational. There's lots more photos and info at
http://www.maniacpony.pwp.blueyonder.co.uk/onscreen/sound.htm



** I would still like to know how the CCD camera reliably acquires 3500
bits of data from 96 tiny squares every second.

Does the film come to a halt 96 times per second to give the camera a
stationary image to snap ?

Or, is the CCD exposure time so short that relative motion is effectively
frozen ?

Maybe clever spinning optics are used to make the image appear stationary to
the camera ?




......... Phil
 
N

Nicholas Sherlock

Jan 1, 1970
0
Phil said:
** I would still like to know how the CCD camera reliably acquires 3500
bits of data from 96 tiny squares every second.

Does the film come to a halt 96 times per second to give the camera a
stationary image to snap ?

Or, is the CCD exposure time so short that relative motion is effectively
frozen ?

Maybe clever spinning optics are used to make the image appear stationary to
the camera ?

Perhaps the exposure time is short, but to make it even finer, a flash
is fired 96 times a second. This technique is used in high speed
photography.

Cheers,
Nicholas Sherlock
 
P

Phil Allison

Jan 1, 1970
0
"Nicholas Sherlock"
Perhaps the exposure time is short, but to make it even finer, a flash is
fired 96 times a second. This technique is used in high speed photography.


** I suppose a high brightness LED could be pulsed at 96 Hz continuously.

The optical data bits are approx 1/800 of an inch square and the film speed
is about 18 inches per second, ie it moves the width of a data bit in
1/15,000 of a second.

The light pulse would need to last 20 uS at most to stop action !!



........ Phil
 
C

Clifford Heath

Jan 1, 1970
0
Phil said:
Does the film come to a halt 96 times per second to give the camera a
stationary image to snap ?

I thought that happened with all film projectors, using, for example, the
Maltese cross mechanism: <http://en.wikipedia.org/wiki/Maltese_Cross_mechanism>.
But IIRC on normal film, the audio is displaced by a few frames and recovered
from continuously moving film.

Clifford Heath.
 
B

Bob Parker

Jan 1, 1970
0
I thought that happened with all film projectors, using, for example, the
Maltese cross mechanism:
<http://en.wikipedia.org/wiki/Maltese_Cross_mechanism>.
But IIRC on normal film, the audio is displaced by a few frames and
recovered
from continuously moving film.

Clifford Heath.

One of those articles mentions the digital area of the film being
scanned 24 frames away from the gate, where it's moving continuously not
being stopped for projection of each frame.
Maybe they pulse high intensity LEDs for a couple of microseconds to
grab each optical digital block?
There's probably info on some of the professional sites... I'll have
a look when I get time.

Bob
 
P

Phil Allison

Jan 1, 1970
0
"Clifford Heath"
I thought that happened with all film projectors, using, for example, the
Maltese cross mechanism:
<http://en.wikipedia.org/wiki/Maltese_Cross_mechanism>.


** Did you even read your own URL ?

Says the film stops 24 times per second - not 96.


But IIRC on normal film, the audio is displaced by a few frames and
recovered
from continuously moving film.


** Bit hard to recover any analogue audio from a stationary film.



....... Phil
 
P

Phil Allison

Jan 1, 1970
0
"ER"
Phil Allison wrote:

The film runs at 24fps but there are 4 sprocket holes per frame



** Is this profound - or just silly ?





...... Phil
 
B

Bob Parker

Jan 1, 1970
0
"Nicholas Sherlock"



** I suppose a high brightness LED could be pulsed at 96 Hz continuously.

The optical data bits are approx 1/800 of an inch square and the film speed
is about 18 inches per second, ie it moves the width of a data bit in
1/15,000 of a second.

The light pulse would need to last 20 uS at most to stop action !!



....... Phil

I found a projector setup procedure
(http://www.film-tech.com/warehouse/manuals/CENMR3.pdf) which seems to
indicate that the Dolby digital pickup is illuminated by continuous LED
light and is positioned where the film's moving at a constant speed.
Apparently the CCD is functioning as a very fast electronic shutter.

Bob
 
P

Phil Allison

Jan 1, 1970
0
" Exasperating Ratbag "


Neither really, just pointing out that although the film runs at 24fps
there are 4 sprocket holes per frame with 4 gaps filled with DD5.1 data
hence the 96 times per second figure for the number of images camera
takes.


** What crapology !

Worse than merely silly - it is completely STUPID.

Another MORON who cannot read or follow a simple thread !!


PISS OFF - IDIOT !!!





........ Phil
 
C

Clifford Heath

Jan 1, 1970
0
Phil said:
** Did you even read your own URL ?
Says the film stops 24 times per second - not 96.

Nothing stopping them using 4 CCDs to capture 4 audio
frames simultaneously. It's not as though very many pixels
are required in each one. They could even be optically
routed to a single CCD using prisms.
 
S

Simon

Jan 1, 1970
0
" Exasperating Ratbag "





** What crapology !

Worse than merely silly - it is completely STUPID.

Another MORON who cannot read or follow a simple thread !!


PISS OFF - IDIOT !!!





....... Phil

wow, whered that come from Phil?

Do you suffer from tourettes or something?
 
P

Phil Allison

Jan 1, 1970
0
"Simple Simon"


= Another Exasperating Ratbag


PISS OFF - IDIOT !!!





...... Phil
 
A

atec 77

Jan 1, 1970
0
Simon said:
wow, whered that come from Phil?

Do you suffer from tourettes or something?
He suffers several diseases , closely akin to tourettes but with a bad
case of a philthy past bad foot in mouth .
 
J

jasen

Jan 1, 1970
0
I found a projector setup procedure
(http://www.film-tech.com/warehouse/manuals/CENMR3.pdf) which seems to
indicate that the Dolby digital pickup is illuminated by continuous LED
light and is positioned where the film's moving at a constant speed.
Apparently the CCD is functioning as a very fast electronic shutter.

That makes sense, they need a constant speed portion for the analogue pickup
anyway.
 
J

jasen

Jan 1, 1970
0
Phil Allison wrote:
Neither really, just pointing out that although the film runs at 24fps
there are 4 sprocket holes per frame with 4 gaps filled with DD5.1 data
hence the 96 times per second figure for the number of images camera takes.

that's if the invest in a raster type sensor.

I expect they just have a linear sensor (like fax machines and scanners use)
and simply read the film as it moves past, hmm, about 15000 rows per second
(during the data blocks) sounds possible, i expect the blue data (sony?) is
read the same way.
 
B

Bob Parker

Jan 1, 1970
0
that's if the invest in a raster type sensor.

I expect they just have a linear sensor (like fax machines and scanners use)
and simply read the film as it moves past, hmm, about 15000 rows per second
(during the data blocks) sounds possible, i expect the blue data (sony?) is
read the same way.

That would seem to make the most sense. Just capture each row of
data bits as it flies past...
 
P

Phil Allison

Jan 1, 1970
0
"Bob Parker"
That would seem to make the most sense. Just capture each row of data
bits as it flies past...



** The DD5.1 35mm film system is clearly STATED to use a CCD camera whose
images are processed - so not a line scanner.

High effective shutter speeds ( even 10uS ) are well possible with CCD
technology, as is operation with a few uS pulse of light.

Processing the "checkerboard" images into a data stream is then a simple DSP
problem.

Let Occam's razor apply.




........ Phil
 
B

Bob Parker

Jan 1, 1970
0
** The DD5.1 35mm film system is clearly STATED to use a CCD camera whose
images are processed - so not a line scanner.

High effective shutter speeds ( even 10uS ) are well possible with CCD
technology, as is operation with a few uS pulse of light.

Processing the "checkerboard" images into a data stream is then a simple DSP
problem.

Let Occam's razor apply.

....... Phil


Google might be my friend, but searching for technical info on the
actual Dolby 701/702 digital track readers and their CCD sensors ain't
easy! I found heaps of info about how to install them into various
models of projectors and connect and align them, though.
Anyhow, http://www.hps4000.com/pages/digital/dolby_sr_d_digital.pdf
says...

The data block is read by a 512 element charged coupled device (CCD),
which is, in essence, a television camera. The output of the CCD is an
analog video "picture" of the spots on the film. This is converted into
the digital video domain at a rate which tracks the speed of the film.
This video data is scanned for synchronization information and to
determine where the four corners of the data block are located. If two
of the corners are known, the position of the entire block is known and
the proper location of the data bits is known. Determining the whether
or not a real data spot exists and preventing false data from entering
the process is best understood by imagining a piece of window screen
being placed directly over the data block. The corner bits verify the
proper Position for the screen. Looking at the "screen", one sees in the
small square openings exactly where the spots should be. Using a digital
version of such a screen, the data spots are identified and then sent to
a "thresholding" stage for further enhancement. Based on the average
density of the spots, a threshold value is determined, above which the
spot is recognized as a "1" and below which, a "0". This stage reduces
the possibility that a scratch could be mistaken for actual data and
establishes what, on the film, is a spot and what is not.

I'm not sure if that clarfies anything or not. :)

Bob
 
Top