AAton Cantar-X1 User Manual Download Page 15

Page: 15 / 64

+33 4 7642 9550

www.aaton.com

Cantar-X User Manual v2.44 (r14) 2013 March 21

p.15

Sampling Rate, Camera fps, Audio-TC

These parameters are not correlated.  The number of audio
samples  per  Earth  second  (Hertz),  is  called  the  sampling
rate,  e.g.  48000  is  48kHz.    The  camera  'frames  per
second' value should have no influence at all on the audio
sample  rate  frequency;  unfortunately,  some  vintage  NTSC
post-machines still impose the use of the disgraceful 48048
or 47952 sample rates to sync real-time audio with drifting-
time images.  The frame-rate used to express the audio time
can be 30, while images are filmed at 24 or 25fps.

On planet Earth

Timecode handling is simple at integer speeds such as 24,
25 and 30fps;

at 29.97DF (Drop-Frame) the frame count

jumps over two images every minute except for every tenth
minute to keep the count in line with Earth's time.

Stored in

the file metadata, the camera fps is a simple reminder and
can be changed later and then applied to the audio LTC
used in some audio-post sync operations.

'U'

Universal : 24.00, 25.00, 30.00, 29.97DF

In  a  file  digitized  at  48kHz,  the  Format stamp  indicates
the  number  of  samples  per  real-second,  i.e.  48000;  the
Time-stamp  carries  the  number  of  samples  since  midnight
using the sample rate value found in the Format stamp

(the

same logic is used in the Barebone mode-B for 'slow-planet'
recordings).

On the .1% 'Slow-Planet'

NTSC-compatible  23.976fps  cameras  cannot  use  the  no-
time-drift drop frame system invented for 29.97fps.  Nobody
found a workable counting algorythm.  In this environment
the  'slow-second'  (

the time it takes to shoot 30 frames of

NTSC video, i.e. 1.001 Earth sec.)

imposes its law.

As soon as you enter

Camera fps = 23.98NDF

29.97NDF

you are on the 'slow-planet', with its lazzy 24 (aka 23.976fps
on Earth) or lazzy 30 (aka 29.97fps on Earth). The Cantar's
TCXO clock is switched to a slower beat and TC separators
change from '

' to '

', e.g.12*45*36.

Many  nights  of  engineers'  time  have  been  spent  trying
to  find  a  universal  method  to  sync  sound  and  images  on
the slow-planet but nothing came out.  You must ask your
post-facility which one of the following

three sample-stamp

modes

is requested:

'A'

Avid v11 23.98NDF–A 29.97NDF–A

In a file digitized at 48kHz, the

Format-stamp

indicates the

number of samples digitized in a slow-second, i.e. 48048;
the

Time-stamp

noted at the file's head is the number of

samples since midnight, using the 48048 value of the
Format-stamp.

Introduced for the AvidMC v11.3.2, the 'A' mode is fading
away. See 'Avid-Recording-Rates' in Aaton tutorials

http://

www.aaton.com/documentation/

'B'

Barebone 23.98NDF–B 29.97NDF–B

In a file digitized at 48kHz, the

Format-stamp

indicates the

number of samples digitized in a real-second, i.e. 48000;
the

Time-stamp

is the number of samples since midnight

using the 48000 value of the Format-stamp.

While images and timecode are beating the slow-second,
the audio remains on Earth! Since modern post-machines
(e.g. Indaw) perform high quality real-time sample rate con-
version, 'Barebone' and 48kHz should have been THE only
choice but unfortunately some other manuafacturers decided
differently (see below).

'C'

Compensated ('F' Faux) 23.98NDF–C 29.97NDF–C

The 'C' mode is made for challenged post-machines. In a
file digitized at 48kHz, the

Format-stamp

value indicates

the real number of samples per real-second, i.e. 48000,
but the

Time-stamp

is the number of samples since midnight

as if

the digitization was done at 48048!

The '

' index (for Fake or Faux) is given by Fostex and FCP