| ltc.h(3) | libltc | ltc.h(3) |
ltc.h - libltc - en+decode linear timecode
#include <endian.h>
#include <stddef.h>
struct LTCFrame
struct LTCFrameExt
struct SMPTETimecode
#define DEPRECATED_EXPORT
#define LTC_FRAME_BIT_COUNT 80
typedef unsigned char ltcsnd_sample_t
typedef long long int ltc_off_t
typedef struct LTCFrame LTCFrame
typedef struct LTCFrameExt LTCFrameExt
typedef struct SMPTETimecode SMPTETimecode
typedef struct LTCDecoder LTCDecoder
typedef struct LTCEncoder LTCEncoder
enum LTC_TV_STANDARD { LTC_TV_525_60,
LTC_TV_625_50, LTC_TV_1125_60, LTC_TV_FILM_24 }
enum LTC_BG_FLAGS { LTC_USE_DATE = 1, LTC_TC_CLOCK = 2,
LTC_BGF_DONT_TOUCH = 4, LTC_NO_PARITY = 8 }
void ltc_frame_to_time (SMPTETimecode *stime,
LTCFrame *frame, int flags)
void ltc_time_to_frame (LTCFrame *frame, SMPTETimecode
*stime, enum LTC_TV_STANDARD standard, int flags)
void ltc_frame_reset (LTCFrame *frame)
int ltc_frame_increment (LTCFrame *frame, int fps, enum
LTC_TV_STANDARD standard, int flags)
int ltc_frame_decrement (LTCFrame *frame, int fps, enum
LTC_TV_STANDARD standard, int flags)
LTCDecoder * ltc_decoder_create (int apv, int queue_size)
int ltc_decoder_free (LTCDecoder *d)
void ltc_decoder_write (LTCDecoder *d, ltcsnd_sample_t
*buf, size_t size, ltc_off_t posinfo)
void ltc_decoder_write_double (LTCDecoder *d, double *buf,
size_t size, ltc_off_t posinfo)
void ltc_decoder_write_float (LTCDecoder *d, float *buf, size_t
size, ltc_off_t posinfo)
void ltc_decoder_write_s16 (LTCDecoder *d, short *buf, size_t
size, ltc_off_t posinfo)
void ltc_decoder_write_u16 (LTCDecoder *d, unsigned short *buf,
size_t size, ltc_off_t posinfo)
int ltc_decoder_read (LTCDecoder *d, LTCFrameExt *frame)
void ltc_decoder_queue_flush (LTCDecoder *d)
int ltc_decoder_queue_length (LTCDecoder *d)
LTCEncoder * ltc_encoder_create (double sample_rate, double fps,
enum LTC_TV_STANDARD standard, int flags)
void ltc_encoder_free (LTCEncoder *e)
void ltc_encoder_set_timecode (LTCEncoder *e,
SMPTETimecode *t)
void ltc_encoder_get_timecode (LTCEncoder *e,
SMPTETimecode *t)
void ltc_encoder_set_user_bits (LTCEncoder *e, unsigned long
data)
unsigned long ltc_frame_get_user_bits (LTCFrame *f)
int ltc_encoder_inc_timecode (LTCEncoder *e)
int ltc_encoder_dec_timecode (LTCEncoder *e)
void ltc_encoder_set_frame (LTCEncoder *e, LTCFrame *f)
void ltc_encoder_get_frame (LTCEncoder *e, LTCFrame *f)
int ltc_encoder_get_buffer (LTCEncoder *e,
ltcsnd_sample_t *buf) DEPRECATED_EXPORT
int ltc_encoder_copy_buffer (LTCEncoder *e,
ltcsnd_sample_t *buf)
ltcsnd_sample_t * ltc_encoder_get_bufptr (LTCEncoder *e,
int *size, int flush) DEPRECATED_EXPORT
int ltc_encoder_get_bufferptr (LTCEncoder *e,
ltcsnd_sample_t **buf, int flush)
void ltc_encoder_buffer_flush (LTCEncoder *e)
size_t ltc_encoder_get_buffersize (LTCEncoder *e)
int ltc_encoder_reinit (LTCEncoder *e, double sample_rate,
double fps, enum LTC_TV_STANDARD standard, int flags)
void ltc_encoder_reset (LTCEncoder *e)
int ltc_encoder_set_bufsize (LTCEncoder *e, double sample_rate,
double fps) DEPRECATED_EXPORT
int ltc_encoder_set_buffersize (LTCEncoder *e, double
sample_rate, double fps)
double ltc_encoder_get_volume (LTCEncoder *e)
int ltc_encoder_set_volume (LTCEncoder *e, double dBFS)
double ltc_encoder_get_filter (LTCEncoder *e)
void ltc_encoder_set_filter (LTCEncoder *e, double rise_time)
int ltc_encoder_encode_byte (LTCEncoder *e, int byte, double
speed)
int ltc_encoder_end_encode (LTCEncoder *e)
void ltc_encoder_encode_frame (LTCEncoder *e)
void ltc_encoder_encode_reversed_frame (LTCEncoder *e)
void ltc_frame_set_parity (LTCFrame *frame, enum
LTC_TV_STANDARD standard)
int ltc_frame_parse_bcg_flags (LTCFrame *frame, enum
LTC_TV_STANDARD standard)
ltc_off_t ltc_frame_alignment (double samples_per_frame, enum
LTC_TV_STANDARD standard)
libltc - en+decode linear timecode
Linear (or Longitudinal) Timecode (LTC) is an encoding of timecode data as a Manchester-Biphase encoded audio signal. The audio signal is commonly recorded on a VTR track or other storage media.
libltc facilitates decoding and encoding of LTC from/to timecode, including SMPTE date support.
Author
Copyright
Copyright (C) 2006-2022 Robin Gareus robin@gareus.org
Copyright (C) 2008-2009 Jan Weiß jan@geheimwerk.de
Inspired by SMPTE Decoder - Maarten de Boer mdeboer@iua.upf.es
This program is free software; you can redistribute it and/or modify it under the terms of the GNU Lesser General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.
This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more details.
You should have received a copy of the GNU Lesser General Public License along with this library. If not, see http://www.gnu.org/licenses/.
Raw 80 bit LTC frame
The datastream for each video frame of Longitudinal Timecode consists of eighty bit-periods.
At a frame-rate of 30 fps, the bit-rate corresponds to 30 [fps] * 80 [bits/f] = 2400 bits per second. The frequency for a stream of zeros would be 1.2 kHz and for a stream of ones it would be 2.4 kHz. With all commonly used video-frame-rates and audio-sample-rates, LTC timecode can be recorded easily into a audio-track.
In each frame, 26 of the eighty bits carry the SMPTE time in binary coded decimal (BCD).
These Bits are FRAME-UNITS, FRAME-TENS, SECS-UNITS, SECS-TENS, MINS-UNITS, MINS-TENS, HOURS-UNITS and HOURS-TENS. The BCD digits are loaded 'least significant bit first' (libltc takes care of the architecture specific alignment).
32 bits are assigned as eight groups of four USER-BITS (also sometimes called the 'Binary Groups'). This capacity is generally used to carry extra info such as reel number and/or date. The User Bits may be allocated howsoever one wishes as long as both Binary Group Flag Bits are cleared.
The function ltc_frame_to_time can interpret the user-bits as SMPTE Date+Timezone according to SMPTE 309M-1999. similarly ltc_time_to_frame will do the reverse.
The last 16 Bits make up the SYNC WORD. These bits indicate the frame boundary, the tape direction, and the bit-rate of the sync tone. The values of these Bits are fixed as 0011 1111 1111 1101
The Bi-Phase Mark Phase Correction Bit (Bit 27 or 59) may be set or cleared so that that every 80-bit word contains an even number of zeroes. This means that the phase of the pulse train in every Sync Word will be the same.
Bit 10 indicates drop-frame timecode. The Colour Frame Flag col.frm is Bit 11; if the timecode intentionally synchronized to a colour TV field sequence, this bit is set.
Bit 58 is not required for the BCD count for HOURS-TENS (which has a maximum value of two) and has not been given any other special purpose so remains unassigned. This Bit has been RESERVED for future assignment.
The Binary Group Flag Bits (bits 43 and 59) are two bits indicate the format of the User Bits data. SMPTE 12M-1999 defines the previously reserved bit 58 to signals that the time is locked to wall-clock within a tolerance of ± 0.5 seconds.
SMPTE 12M-1999 also changes the numbering schema of the BGF. (BGF1 was renamed to BGF2 and bit 58 becomes BGFB1)
To further complicate matters, the BGFB assignment as well as the biphase_mark_phase_correction (aka parity) bit depends on the timecode-format used.
25 fps 24, 30 fps
BGF0 27 43
BGF1 58 58
BGF2 43 59
Parity 59 27
The variable naming chosen for the LTCFrame struct is based on the 24,30 fps standard.
The Binary Group Flag Bits should be used only as shown in the truth table below. The Unassigned entries in the table should not be used, as they may be allocated specific meanings in the future.
BGF0 BGF1 BGF2
user-bits timecode Bit 43 Bit 58 Bit 59 (30fps, 24 fps)
| | Bit 27 Bit 58 Bit 43 (25fps)
No User Bits format specified | ? | 0 0 0
Eight-bit character set (1) | ? | 1 0 0
Date and Timezone set | ? | 0 0 1
Page/Line multiplex (2) | ? | 1 0 1
Character set not specified | clk | 0 1 0
Reserved | ? | 1 1 0
Date and Timezone set | clk | 0 1 1
Page/Line multiplex (2) | clk | 1 1 1
(1) ISO/IEC 646 or ISO/IEC 2022 character set. If the seven-bit ISO codes are being used, they shall be converted to eight-bit codes by setting the eighth bit to zero. 4 ISO codes can be encoded, user7 and user8 are to be used for the first code with LSB 7 and MSB in 8. the remaining ISO codes are to be distributed in the same manner to user5/6 user3/4 and user1/2 accordingly.
(2) The Page/Line indicates ANSI/SMPTE-262M is used for the user-bits. It is multiplex system that can be used to encode large amounts of data in the binary groups through the use of time multiplexing.
libltc does not use any of the BGF - except for the Parity bit which can be calculated and set with ltc_frame_set_parity. Setting and interpreting the BGF is left to the application using libltc. However libltc provides functionality to parse or set date and timezoe according to SMPTE 309M-1999.
further information: http://www.philrees.co.uk/articles/timecode.htm and http://www.barney-wol.net/time/timecode.html
Data Fields:
unsigned int binary_group_flag_bit1:1 indicate timecode is
local time wall-clock, see table above - bit 58
unsigned int binary_group_flag_bit2:1 indicate user-data
char encoding (or parity with 25fps), see table above - bit 59
unsigned int biphase_mark_phase_correction:1 see note on
Bit 27 in description and ltc_frame_set_parity .
unsigned int col_frame:1 colour-frame: timecode
intentionally synchronized to a colour TV field sequence
unsigned int dfbit:1 indicated drop-frame timecode
unsigned int frame_tens:2 SMPTE framenumber BCD tens 0..3.
unsigned int frame_units:4 SMPTE framenumber BCD unit 0..9.
unsigned int hours_tens:2 SMPTE hours BCD tens 0..2.
unsigned int hours_units:4 SMPTE hours BCD unit 0..9.
unsigned int mins_tens:3 SMPTE minutes BCD tens 0..6.
unsigned int mins_units:4 SMPTE minutes BCD unit 0..9.
unsigned int secs_tens:3 SMPTE seconds BCD tens 0..6.
unsigned int secs_units:4 SMPTE seconds BCD unit 0..9.
unsigned int sync_word:16
unsigned int user1:4
unsigned int user2:4
unsigned int user3:4
unsigned int user4:4
unsigned int user5:4
unsigned int user6:4
unsigned int user7:4
unsigned int user8:4
Extended LTC frame - includes audio-sample position offsets, volume, etc
Note: For TV systems, the sample in the LTC audio data stream where the LTC Frame starts is not necessarily at the same time as the video-frame which is described by the LTC Frame.
off_start denotes the time of the first transition of bit 0 in the LTC frame.
For 525/60 Television systems, the first transition shall occur at the beginning of line 5 of the frame with which it is associated. The tolerance is ± 1.5 lines.
For 625/50 systems, the first transition shall occur at the beginning of line 2 ± 1.5 lines of the frame with which it is associated.
Only for 1125/60 systems, the first transition occurs exactly at the vertical sync timing reference of the frame. ± 1 line.
Examples
ltcdecode.c.
Data Fields:
LTCFrame ltc the actual LTC frame. see
LTCFrame
ltc_off_t off_end the sample in the stream
corresponding to the end of the LTC frame.
ltc_off_t off_start the approximate sample in the
stream corresponding to the start of the LTC frame.
int reverse if non-zero, a reverse played LTC frame was
detected. Since the frame was reversed, it started at off_end and finishes
as off_start (off_end > off_start). (Note: in reverse playback the
(reversed) sync-word of the next/previous frame is detected, this offset is
corrected).
ltcsnd_sample_t sample_max the maximum input sample
signal for this frame (0..255)
ltcsnd_sample_t sample_min the minimum input sample
signal for this frame (0..255)
double volume the volume of the input signal in dbFS
Human readable time representation, decimal values.
Examples
example_encode.c, ltcdecode.c, and ltcencode.c.
Data Fields:
unsigned char frame sub-second frame 0..(FPS - 1)
unsigned char hours hour 0..23
unsigned char mins minute 0..60
unsigned char months valid months are 1..12
unsigned char secs second 0..60
char timezone[6] the timezone 6bytes: '+HHMM' textual
representation
unsigned char years LTC-date uses 2-digit year 00.99.
sample-count offset - 64bit wide
Opaque structure see: ltc_decoder_create, ltc_decoder_free
Opaque structure see: ltc_encoder_create, ltc_encoder_free
see LTCFrame
see LTCFrameExt
default audio sample type: 8bit unsigned (mono)
see SMPTETimecode
encoder and LTCframe <> timecode operation flags
Enumerator
the standard defines the assignment of the binary-group-flag bits basically only 25fps is different, but other standards defined in the SMPTE spec have been included for completeness.
Enumerator
Create a new LTC decoder.
Parameters
Returns
Examples
ltcdecode.c.
Release memory of decoder.
Parameters
Examples
ltcdecode.c.
Remove all LTC frames from the internal queue.
Parameters
Count number of LTC frames currently in the queue.
Parameters
Returns
Decoded LTC frames are placed in a queue. This function retrieves a frame from the queue, and stores it at LTCFrameExt*
Parameters
Returns
Examples
ltcdecode.c.
Feed the LTC decoder with new audio samples.
Parse raw audio for LTC timestamps. Once a complete LTC frame has been decoded it is pushed into a queue (ltc_decoder_read)
Parameters
Examples
ltcdecode.c.
Wrapper around ltc_decoder_write that accepts 64-bit floating point audio samples. Note: internally libltc uses 8 bit only.
Parameters
Wrapper around ltc_decoder_write that accepts 32-bit floating point audio samples. Note: internally libltc uses 8 bit only.
Parameters
Wrapper around ltc_decoder_write that accepts signed 16 bit audio samples. Note: internally libltc uses 8 bit only.
Parameters
Wrapper around ltc_decoder_write that accepts unsigned 16 bit audio samples. Note: internally libltc uses 8 bit only.
Parameters
reset the write-pointer of the encoder-buffer
Parameters
Copy the accumulated encoded audio to the given sample-buffer and flush the internal buffer.
Parameters
Returns
Examples
example_encode.c.
Allocate and initialize LTC audio encoder.
calls ltc_encoder_reinit internally see, see notes there.
Parameters
Examples
example_encode.c, and ltcencode.c.
Move the encoder to the previous timecode frame. This is useful for encoding reverse LTC. uses ltc_frame_decrement() internally.
Generate LTC audio for given byte of the LTC-frame and place it into the internal buffer.
see ltc_encoder_copy_buffer and ltc_encoder_get_bufferptr
LTC has 10 bytes per frame: 0 <= bytecnt < 10 use SMPTESetTime(..) to set the current frame before Encoding. see tests/encoder.c for an example.
The default output signal is @ -3dBFS (38..218 at 8 bit unsigned). see also ltc_encoder_set_volume
if speed is < 0, the bits are encoded in reverse. slowdown > 10.0 requires custom buffer sizes; see ltc_encoder_set_buffersize
Parameters
Returns
Examples
example_encode.c.
Encode a full LTC frame at fixed speed. This is equivalent to calling ltc_encoder_encode_byte 10 times for bytes 0..9 with speed 1.0.
Note: The internal buffer must be empty before calling this function. Otherwise it may overflow. This is usually the case if it is read with ltc_encoder_copy_buffer after calling this function.
The default internal buffersize is exactly one full LTC frame at speed 1.0.
Parameters
Examples
example_encode.c, and ltcencode.c.
Encode a full LTC frame at fixed speed -1. This is equivalent to calling ltc_encoder_encode_byte 10 times for bytes 9..0, rolling in reverse at speed 1.
Note: The internal buffer must be empty before calling this function. Otherwise it may overflow. This is usually the case if it is read with ltc_encoder_copy_buffer after calling this function.
Parameters
Terminate encoding and add final transition
Refer to the image at LTCFrame. In this example, the encoded data starts and ends with a rising edge. The transition at the start of tne next frame marks the end of the previous frame. This transition is encoded at the beginning of a frame. However if there is no additional frame to be encoded, a final terminating transition has to be added.
Since LTC is usually sent as continuous stream, this is of no concern. However for a fixed, finite duration to be encoded, this method adds a terminating transition to the buffer.
After this one must either call ltc_encoder_reset() or ltc_encoder_free.
Parameters
Returns
Release memory of the encoder.
Parameters
Examples
example_encode.c, and ltcencode.c.
Copy the accumulated encoded audio to the given sample-buffer and flush the internal buffer.
Parameters
Returns
Deprecated
Retrieve a pointer to the accumulated encoded audio-data.
Parameters
Returns
Examples
example_encode.c, and ltcencode.c.
Query the length of the internal buffer. It is allocated to hold audio-frames for exactly one LTC frame for the given sample-rate and frame-rate. ie. (1 + sample-rate / fps) bytes
Note this returns the total size of the buffer, not the used/free part. See also ltc_encoder_get_bufferptr
Parameters
Returns
Examples
example_encode.c.
Retrieve a pointer to the accumulated encoded audio-data.
Parameters
Returns
Deprecated
Get encoder signal rise-time / signal filtering
Parameters
Returns
Low-level access to the encoder internal LTCFrame data
Parameters
Query the current encoder timecode.
Note: the decoder stores its internal state in an LTC-frame, this function converts that LTC-Frame into SMPTETimecode on demand. see also ltc_encoder_get_frame.
Parameters
Query the volume of the generated LTC signal
Parameters
Returns
Move the encoder to the next timecode frame. uses ltc_frame_increment() internally.
Examples
example_encode.c, and ltcencode.c.
Change the encoder settings without re-allocating any library internal data structure (realtime safe). changing the fps and or sample-rate implies a buffer flush, and biphase state reset.
This call will fail if the internal buffer is too small to hold one full LTC frame. Use ltc_encoder_set_buffersize to prepare an internal buffer large enough to accommodate all sample_rate, fps combinations that you would like to re-init to.
The LTC frame payload data is not modified by this call, however, the flag-bits of the LTC-Frame are updated: If fps equals to 29.97 or 30000.0/1001.0, the LTCFrame's 'dfbit' bit is set to 1 to indicate drop-frame timecode.
Unless the LTC_BGF_DONT_TOUCH flag is set the BGF1 is set or cleared depending on LTC_TC_CLOCK and BGF0,2 according to LTC_USE_DATE and the given standard. col_frame is cleared and the parity recomputed (unless LTC_NO_PARITY is given).
Parameters
Examples
ltcencode.c.
reset ecoder state. flushes buffer, reset biphase state
Parameters
Configure a custom size for the internal buffer.
This is needed if you are planning to call ltc_encoder_reinit() or if you want to keep more than one LTC frame's worth of data in the library's internal buffer.
The buffer-size is (1 + sample_rate / fps) bytes. resizing the internal buffer will flush all existing data in it - alike ltc_encoder_buffer_flush.
Parameters
Returns
Examples
ltcencode.c.
Configure a custom size for the internal buffer.
This is needed if you are planning to call ltc_encoder_reinit() or if you want to keep more than one LTC frame's worth of data in the library's internal buffer.
The buffer-size is (1 + sample_rate / fps) bytes. resizing the internal buffer will flush all existing data in it - alike ltc_encoder_buffer_flush.
Parameters
Returns
Deprecated
Set encoder signal rise-time / signal filtering
LTC signal should have a rise time of 40us +/- 10 us. by default the encoder honors this and low-pass filters the output depending on the sample-rate.
If you want a perfect square wave, set 'rise_time' to 0.
Note ltc_encoder_reinit resets the filter-time-constant to use the default 40us for the given sample-rate, overriding any value previously set with ltc_encoder_set_filter
Parameters
Examples
ltcencode.c.
Low-level access to the internal LTCFrame data.
Note: be careful to about f->dfbit, the encoder sets this [only] upon initialization.
Parameters
Set the encoder LTC-frame to the given SMPTETimecode. The next call to ltc_encoder_encode_byte or ltc_encoder_encode_frame will encode this time to LTC audio-samples.
Internally this call uses ltc_time_to_frame because the LTCEncoder operates on LTCframes only. see als ltc_encoder_set_frame
Parameters
Examples
example_encode.c, and ltcencode.c.
Set the user-bits of the frame to the given data.
The data should be a 32-bits unsigned integer. It is written LSB first continiously int the eight user fields.
Parameters
Set the volume of the generated LTC signal
typically LTC is sent at 0dBu ; in EBU callibrated systems that corresponds to -18dBFS. - by default libltc creates -3dBFS
since libltc generated 8bit audio-data, the minimum dBFS is about -42dB which corresponds to 1 bit.
0dB corresponds to a signal range of 127 1..255 with 128 at the center.
Parameters
Returns
Examples
ltcencode.c.
LTCFrame sample alignment offset.
There is a relative offset of the LTC-Frame start and the TV-frame. The first bit of a LTC frame corresponds to a specific line in the actual video frame. When decoding this offset needs to be subtracted from the LTC-frame's audio-sample-time to match the TV-frame's start position.
For film frames or HDV the offset is zero.
Parameters
Returns
Decrement the timecode by one Frame (1/framerate seconds) and set the Frame's parity bit accordingly (see ltc_frame_set_parity)
Parameters
Returns
Get the 32-bits unsigned integer from the user-data bits. The data should be written LSB first in the frame
Parameters
Increment the timecode by one Frame (1/framerate seconds) and set the Frame's parity bit accordingly (see ltc_frame_set_parity)
Parameters
Returns
Parse Binary Coded Group Flags into standard independent format: bit 0 (1) - BGF 0, bit 1 (2) - BGF 1, bit 2 (4) - BGF 2
Parameters
Returns
Reset all values of a LTC FRAME to zero, except for the sync-word (0x3FFD) at the end. The sync word is set according to architecture (big/little endian). Also set the Frame's parity bit accordingly (see ltc_frame_set_parity)
Parameters
Set the parity of the LTC frame.
Bi-Phase Mark Phase Correction bit (bit 27 - or 59) may be set or cleared so that that every 80-bit word contains an even number of zeroes. This means that the phase in every Sync Word will be the same.
This is merely cosmetic; the motivation to keep the polarity of the waveform constant is to make finding the Sync Word visibly (on a scope) easier.
There is usually no need to call this function directly. The encoder utility functions ltc_time_to_frame, ltc_frame_increment and ltc_frame_decrement include a call to it.
Parameters
Convert binary LTCFrame into SMPTETimecode struct
Parameters
Examples
ltcdecode.c.
Translate SMPTETimecode struct into its binary LTC representation and set the LTC frame's parity bit accordingly (see ltc_frame_set_parity)
Parameters
Generated automatically by Doxygen for libltc from the source code.
| Sun Sep 4 2022 | Version 1.3.2 |