:warning: This file is part of the "archive" – it could contain statements or terms which are incompatible with the main body of the specification and the identity and timing model. :warning:

Examples of Timed Data Structures

Note: All the TDS examples given here are valid. However, in typical real-world usage, Timed Data Structures are likely to be defined with some constraints upon the nature of their Data Objects (e.g. only JSON payloads, only H.264 coded intra-frame video, only raw audio).

TDS Example 1

• Each Data Object is restricted to being a 10-bit raw video frame given what the active camera produces. The camera samples the scene at 25 frames per second.
• Data Objects in this TDS are chosen to be in v210 format only.

* TDS Entry
** Time Value: 0 milliseconds
** Data Object Contents: v210 Video Frame @ 1920x1080 4:2:2 10-bit

* TDS Entry
** Time Value: 40 milliseconds
** Data Object Contents: v210 Video Frame @ 1920x1080 4:2:2 10-bit

* TDS Entry
** Time Value: 80 milliseconds
** Data Object Contents: v210 Video Frame @ 1920x1080 4:2:2 10-bit

* TDS Entry
** Time Value: 120 milliseconds
** Data Object Contents: v210 Video Frame @ 1920x1080 4:2:2 10-bit

* TDS Entry
** Time Value: 180 milliseconds
** Data Object Contents: v210 Video Frame @ 1920x1080 4:2:2 10-bit

TDS Example 2

A camera exists at a point in space and captures video and audio. It provides access to this video and audio as SDI from which a TDS is produced. You could split the two transducers and represent their observations as two different TDSs if desired, but in this case we haven't.

• Each Data Object can contain both video and audio, with the audio covering samples for more than just one time instant. The video and audio rate may not align, and so there is no guaranteed mapping between the leading edge of the first video/audio sample in the 'frame' and the Time Value which the Data Object has in the TDS, but the encapsulation format defines the means to map to the Time Value with an offset.

* TDS Entry
** Time Value: 0 nanoseconds
** Data Object Contents: 29.97Hz SDI payload containing: Video Frame @ 1920x1080 + 1602 L24 audio samples

* TDS Entry
** Time Value: 33366700 nanoseconds
** Data Object Contents: 29.97Hz SDI payload containing: Video Frame @ 1920x1080 + 1601 L24 audio samples

* TDS Entry
** Time Value: 66733400 nanoseconds
** Data Object Contents: 29.97Hz SDI payload containing: Video Frame @ 1920x1080 + 1602 L24 audio samples

* TDS Entry
** Time Value: 100100100 nanoseconds
** Data Object Contents: 29.97Hz SDI payload containing: Video Frame @ 1920x1080 + 1601 L24 audio samples

* TDS Entry
** Time Value: 133466800 nanoseconds
** Data Object Contents: 29.97Hz SDI payload containing: Video Frame @ 1920x1080 + 1602 L24 audio samples

• Device(s) playing this out must be aware of the mapping between the encapsulated data and the Time Value in order to play it out correctly in time.

TDS Example 3

This describes a non-standard packing format for video, audio and data. In this specific TDS, video is sampled at 25Hz, audio at 48kHz, and GPS lat/lon co-ordinates at 16kHz. This results in a single Data Object every 1/48000 seconds, which consists of between one and three elements.

* TDS Entry
** Time Value: 0 nanoseconds
** Data Object Contents: v210 format video frame, 8 channels of L24 audio (1 sample per channel), and Lat/Lon co-ordinates

* TDS Entry
** Time Value: 20833 nanoseconds
** Data Object Contents: 8 channels of L24 audio (1 sample per channel)

* TDS Entry
** Time Value: 41666 nanoseconds
** Data Object Contents: 8 channels of L24 audio (1 sample per channel)

* TDS Entry
** Time Value: 62500 nanoseconds
** Data Object Contents: 8 channels of L24 audio (1 sample per channel), and Lat/Lon co-ordinates

* TDS Entry
** Time Value: 83333 nanoseconds
** Data Object Contents: 8 channels of L24 audio (1 sample per channel)

TDS Example 4

This TDS represents periodically sampled data from a temperature sensor. The sensor reflects the temperature at a point within a room.

• The system takes measurements from the sensor every 60 seconds and records this in a Data Object which is simply a number, known to be measured in degrees centigrade.

* TDS Entry
** Time Value: 60 seconds
** Data Object Contents: 16

* TDS Entry
** Time Value: 120 seconds
** Data Object Contents: 17

* TDS Entry
** Time Value: 180 seconds
** Data Object Contents: 19

* TDS Entry
** Time Value: 240 seconds
** Data Object Contents: 24

* TDS Entry
** Time Value: 300 seconds
** Data Object Contents: 26

TDS Example 5

This TDS is similar to the example in TDS 4, but instead Data Objects are recorded in an interrupt-driven fashion when the value of the temperature sensor changes by a degree. TDS 5 is a more accurate representation of the changing temperature than TDS 4. Both TDS 4 and TDS 5 are associated with the same TVI.

* TDS Entry
** Time Value: 52 seconds
** Data Object Contents: 16

* TDS Entry
** Time Value: 115 seconds
** Data Object Contents: 17

* TDS Entry
** Time Value: 153 seconds
** Data Object Contents: 18

* TDS Entry
** Time Value: 174 seconds
** Data Object Contents: 19

* TDS Entry
** Time Value: 191 seconds
** Data Object Contents: 20

TDS Example 6

This is a duplicate of TDS 5. It uses Data Objects which are byte-identical, in the same sequence, but the TDS id is different as the Time Values differ. It is associated with the same TVI, but using a different Time Context. Here the Time Context uses an origin set to 01/01/1970 with a 'TAI' scale, along with a global reference.

* TDS Entry
** Time Value: 1520423775 seconds
** Data Object Contents: 16

* TDS Entry
** Time Value: 1520423838 seconds
** Data Object Contents: 17

* TDS Entry
** Time Value: 1520423876 seconds
** Data Object Contents: 18

* TDS Entry
** Time Value: 1520423897 seconds
** Data Object Contents: 19

* TDS Entry
** Time Value: 1520423914 seconds
** Data Object Contents: 20

TDS Example 7

A user wants to set the temperature of a heater. They choose five temperatures and choose times in the future for each of these temperatures to take effect. This is a TDS with Data Objects containing temperatures.

The user later decides to move all of these Time Values by one hour to account for a daylight savings change they had forgotten about. This results in new Time Values and as such creates a new TDS whose Data Objects are each offset from each other by the same time period which they were previously. In this instance the same Time Context is used in both cases, but there may be a mapping defined externally to indicate that the two TDSs are in some way related.

TDS Example 8

A user has a raw video TDS containing v210 format Data Objects. The user creates a matching H264 intra-coded sequence of Data Objects representing the same content. This is necessarily a different TDS with a different id, but may have the same Time Values as the original v210 TDS. Both of these Timed Data Structures are associated with the same Time Varying Information.

A further TDS is created using the original v210 Data Objects in the same sequence, but with the time period between each Time Value halved. This produces a double frame rate video representation. This is a new TDS with new Time Values and is associated with a different Time Varying Information, but uses the same Data Objects.

TDS Example 9

A camera is creating 50Hz 10-bit raw 1920x1080 4:2:2 video output, producing Data Objects and a resultant Timed Data Structure. The camera is reconfigured to output at half the rate. This can be modelled as the same TDS (associated with the same TVI) but consists of a lower frequency of generation of Data Objects and Time Values after this point.

TDS Example 10

The output of a vision mixer represents the bringing together of several TDSs (each of which is associated with a different TVI). This 'mix' operation may be performed live with streams of data coming in, or may be performed in a non-linear editor. In either case, the new output TDS will be a mix of identical Data Objects to those provided to the inputs, and some new Data Objects which may be a composite of multiple input Data Objects.s

Depending on the application, Time Values assigned to output Data Objects may match the values used at inputs or may be coined fresh in order to associate this composition with a particular point in time.

TDS Example 11

A TDS is stored on disk with its corresponding Data Objects and Time Values. It is played out onto a network using RTP. The RTP payload type and "maximum transmission unit" mean that the data cannot be carried exactly as it is laid out in the existing Data Object. As such a new set of Data Objects and as such a new TDS must be created for transmission via RTP, however the Time Values for the RTP encapsulated Data Objects may be chosen to remain the same as previous. Once data reaches a receiving system, it may be possible to reconstruct an identical TDS with identical Data Objects and Time Values to the one which was originally stored on disk. This entity could as a result use the same TDS id.

TDS Example 12

This example shows a TDS which is very unlikely to ever exist in reality. It is intended to show that at this level of modelling it doesn't matter what the data is, but that it is binary data which has had a Time Value associated with it. This differs from a playlist in that a playlist alone does not usually define precisely when each item must play out.

* TDS Entry
** Time Value: 92785 seconds
** Data Object Contents: Still image of an elephant in JPEG format

* TDS Entry
** Time Value: 104829 seconds
** Data Object Contents: WAV format audio file containing 5 seconds worth of 1kHz tone

* TDS Entry
** Time Value: 104934 seconds
** Data Object Contents: Extract of Lorem Ipsum covering 500 characters in UTF8

* TDS Entry
** Time Value: 582056 seconds
** Data Object Contents: 2 lines extracted from an Apache web server log file

* TDS Entry
** Time Value: 983632 seconds
** Data Object Contents: AVI format video file covering 1 minute at 25fps