Glossary

• AES/EBU

  Professional digital audio interface.

• AIFF

  Audio file format developed by Apple.

• ANTI-CREASING

  A digital image processing method that removes aliasing and smoothes lines, in particular to prevent diagonals from appearing as stairs.

• ATSC (Advanced Television Systems Committee)

  Various proposals for US high-definition TV systems, with the common feature that the picture is encoded in MPG2 and the sound in Dolby AC-3.

• AVI (Audio Video Interleaved)

  Microsoft moving image format (the PC equivalent of Apple's Quick Time).

• ILO

  Minimal information in computing that can only receive 2 states: 0 or 1 (not to be confused with byte or octet).

• BLU-RAY OR BD

  Blu-ray or Blu-ray Disc (official abbreviation BD) is a digital medium, like DVD, enabling storage of very large amounts of information. It is ideal for broadcasting HD programs. Its name derives from the type of laser beam it uses, with a spectral color close to blue. The recordable version is called BD-R and the rewritable version BD-RE. Media size (12 cm and 8 cm diameter discs): - BD-R and BD-RE (rewritable): 25 GB (7.8 GB for 8 cm discs) - BD-R double-layer: 50 GB (15.6 GB for 8 cm discs) Supports MPEG2 and MPEG4 codecs. Recordable in resolutions ranging from 720×576 to 1920×1080. Includes an interactive mode called BD-J.

• BMP (Bitmap)

  File format developed by Microsoft for coding still images.

• BNC (Bayonett Conector)

  Bayonet connector used in particular for composite or component video connections (1 to 3 simultaneous connectors).

• BOBINE LABORATOIRE (Lab roll)

  It's the assembly of several camera reels placed end to end in preparation for the transition to telecine.

• BUS

  Internal link of a microcomputer, connecting the motherboard to various cards and other peripherals. This connection can be made over 8 wires (8 bits), 16 wires (16 bits), etc. There are several standards, such as Nubus, PCI, PCX...

• CACHE (Matte Key)

  Inlay from an external cutting signal, whether it comes from a separate video signal, or is included in a file.

• CANAL ALPHA (Alpha Channel)

  Complementary signal used to define an inlay. Usually integrated into a graphics file.

• CCIR

  International Radio Consultative Committee.

• CHROMINANCE

  The part of a component video signal that specifically contains color information.

• CUSTOMER

  An application that sends requests to a server to
  obtain information or perform a task on its behalf. Clients and servers communicate via special protocols, enabling them to be located on different platforms.

• CODEC (Compression / decompression software)

  Compensation is a way of creating a 50 Hz video signal from a film image shot at 24 fps. Telecine adds 1 frame every 12, so 1 frame every 24, creating a slight strobe effect. (In 60 Hz countries, you have to go from 24 fps to 30 fps, and the compensation is called "3:2", i.e. a frame is added every two images.

• COMPONENTS (component)

  One of the states of video coding after RGB, it is still in the form of three different signals: luminance (y) and chrominance (R-Y and B-Y, respectively called Cr and CB - the terminology "YUV" is in fact specific to Betacam).

• COMPOSITE

  One of the states of video coding, it takes the form of a single signal, in which luminance and chrominance are mixed, and exists in several standards: NTSC (30 i/s), PAL or SECAM (25 i/s).

• COMPOSITING

  A trick function that allows you to combine several image elements to create a new image that will be the assembly of the various integrated media or "layers".

• COMPRESSION

  An operation that reduces the amount of information needed to encode an image or sound. Compression can be achieved by simply removing redundant information, in which case the information can then be reconstituted without loss of information (lossless). This is particularly true of computer compression, where we work on a digitized stream to reduce its weight.
  Or use algorithms that take advantage of psychophysiological studies of human vision, but remove some of the original information (LOSSY).
  This can be done in a number of ways:
  - Spatial or intra-image: when there's a uniform area in an image, we don't have to preserve the information for each pixel individually. We therefore get rid of redundant information within the image itself - Temporal or inter-image: from one image range to the same range in the next image, there is sometimes no change and no need to repeat. We get rid of information that hasn't changed since the previous image.
  All digital systems use intra-frame compression, as do most new digital formats. The interframe method is used today by Digital Betacam and by broadcast systems in particular.
  Compression ratios are referred to as "2:1", "3:1" etc., which means that the number of bytes required to encode an image has been divided by 2, 3 etc. respectively....

• CRENELING (Aliasing)

  Defect in the image that gives it a crenellated appearance. (also known as the
  staircase").

• CSDI (Compressed Serial Digital Interface)

  Interface used to communicate the DV signal in its compressed digital form with audio, time code and commands from one piece of equipment to another. This signal is put into serial form so that it can be transmitted over a single BNC-type cable, enabling transfers at 4 times the nominal speed (DVC pro).

• D1

  Digital VCR format that records an uncompressed 4:2:2 signal. Despite its high quality, this is a relatively little-used format because of its price.

• D2

  Digital VCR format that records a digital composite signal (PAL or NTSC). Designed to replace the old 1-inch VTRs, the potential quality of digital is limited by the use of a composite signal, with all the problems that this poses (color sequencing, cross-color, etc.), and so this format is not widely used.

• D3

  Digital VCR format that records a digital composite signal (PAL or NTSC). The same problems were encountered as with the D2, but the smaller cassettes (1/2 inch) made it possible to develop a camcorder.

• D5

  Digital VCR format that records an uncompressed 4:2:2 signal, and is capable of replaying D3 cassettes.

• D6

  Digital video recorder format that records an uncompressed high-definition digital signal in both European and Japanese formats.

• DCT (Discrete Cosine Transform)

  This is the first step in almost all compression systems: the image is divided into blocks of 8 x 8 pixels, which are transformed into frequency variation information. Instead of having information on the pixels themselves, we have information on their variations from one to another. The process does not cause any loss of original information (lossless compression): the image can be completely reconstructed.

• DEFORMATION (Warp)

  Function available on trick systems that allows part of an image to be distorted according to various constraints.

• HORIZONTAL SLIDING (Crawl)

  Regular horizontal scrolling of a title.

• VERTICAL ROLLER (Roll)

  Regular vertical scrolling of a title.

• DOLBY E

  Dolby E is a technology developed by the Dolby© laboratories to compress 8 discrete audio tracks into a single stereo stream, i.e. a 5.1 and a stereo pair. This encoding is now mandatory for the delivery of High Definition PADs (HDCAM etc...) to television channels, even for stereo broadcasting.

• DIFFUSABLE (Broadcast)

  Image quality sufficient for use at the time of program broadcast (synonymous with ONLINE).

• DVD Video (Digital Versatile Disk)

  A compact disc with a capacity of 4.34 to 8.6 GB on one or both sides, featuring not only multi-channel sound (AC 3 or DTS) but also MPG2-coded picture. This standard supports 4:3 and 16:9 formats. Video DVDs come in 4 different sizes: DVD 5, with one layer of 4.34 GB; DVD 9, with 4.34 GB multiplied by two layers on one side; DVD 10, with one layer of 4.34 GB per side on two sides; and DVD 18, with two layers per side on two sides. (the latter is not used because of the technical constraints involved in its production).

• DLT

  Cartridge-based recorder for delivery to the DVD glassmaster press, as well as for high-volume backup of media files. With the DLT 35, you can fit 35 GB on a cartridge. Transfer time is around 16 GO/h.

• DV (Digital Video)

  DV is a compressed (5:1) digital video tape recording format jointly defined by a number of video equipment manufacturers.
  It is available in several formats:
  - Consumer DV.
  - DVCam, a professional format from Sony
  - DVCPro, a professional format from Panasonic.
  But in all cases, the name DV, whether for the general public or professionals, refers to the way in which video and audio signals are recorded on tape.

• DVCAM

  Format developed by Sony from DV. Uses 4:2:0 compressed to 5:1 for 25 fps, and 4:1:1 for 30 fps.

• DVCPRO

  Format developed by Panasonic from DV. DVCPro 25 uses 4:2:2 compressed to 3:1.

• DVCS (DV Compression Serial)

  Low bit-rate signal, 36 or 72 Mb/s compressed DV, audio, timecode and controls, serialized for transmission over a single BNC-type cable.

• ETHERNET

  LAN type using RJ 45 connectors :
  10 Base-T: 10 Mb/s.
  100 Base-T: 100 Mb/s .
  1000 Base-T: 1 Gb/s..

• FLV

  Video format used by Flash.

• FRAGMENTATION

  A characteristic frequently observed on discs when a large number of master clips are recorded and then erased, without the possibility of completely emptying the discs.
  In this case, the media files are spread all over the surface of the disk, and the next recording will be made using several "holes", which will split it into several fragments. A highly fragmented disc may prevent images from being read, especially when using the highest resolutions. We get rid of them by consolidating, or by making a simple partition-to-partition copy.

• GENERATION

  This term refers to the number of copies a video signal undergoes. A signal coming from a video source (camera) and recorded for the first time on a tape is said to be generation 1; a tape onto which this generation 1 tape has been copied is said to be generation 2; and so on.
  In general, analog video does not exceed 4 or 5 generations.
  In practice, uncompressed digital video on tape produces around twenty generations.
  With uncompressed digital video on disc, we could achieve millions of generations (disc-to-disc copies), but this no longer makes sense, since one of the advantages of nonlinear video is that it avoids copies.

• GIGA OCTET

  1024 Megabytes.

• GOP (Group of Picture)

  Number of frames encoding an MPEG sequence.

• HARMONIZATION (Stretch)

  Sound processing that allows the original frequencies to be recovered when a sound is lengthened or shortened.

• IEEE-1394 (Firewire)

  IEEE-1394 is a connection standard between a computer and its various peripherals. Designed in 1986 by Apple, this interface is now used by many manufacturers. It has the advantage of :
  - offer a bandwidth of 100 to 800 Mb (as an example, a DV device requires around 30 Mb/s)
  - connect some 63 peripherals together without the need for an identification number (SCSI ID)
  - enable these peripherals to communicate with each other without going through the computer;
  - be able to plug and unplug them while you're working, without having to shut down and restart the computer or peripherals: just plug in, and the peripheral is immediately operational. (hot disconnect).
  As manufacturers of digital video equipment discovered the advantages of this interface, they integrated it into their equipment (camcorders, televisions, etc.), enabling them to be easily connected to the computer world. For example, some camcorders equipped with this connector can directly transfer a still image on a few seconds of tape to the computer's hard disk. IEEE-1394 can therefore transmit a signal from a DV VTR, as well as many other formats.
  However, as IEEE-1394 was available and chosen as the interface when the DV was designed, there is often confusion between the two terms. IEEE-1394 could well be the future of audio connection, as it allows both MIDI commands and 32 channels of sound to pass through the same target.

• DOWNSTREAM Keyer

  Inlaying is performed on top of all other effects, i.e. at the last stage of the trick.

• CHROMINANCE INCREASE (Chroma Key)

  Inlay that replaces a colored background (often blue or green) with another image.

• INSERTION (Insert)

  1/ Linear video editing mode: unlike "assemble" mode, we record on previously recorded tracks without touching the servo and synchronization tracks.
  2/ In non-linear editing, an editing method that allows a new element to be inserted in the middle of a sequence and the end of the sequence to be postponed without modifying it - something that typically cannot be done as an "insertion" in linear editing.

• INTER-IMAGE (Interframe)

  A type of compression that doesn't just work frame by frame, but is spread over a sequence of images - like MPEG.

• INTERPOLATION (Interpolate)

  Digital image processing, where pixels are inserted or created by calculation using neighboring pixels. This process is used in particular for zooming in on an image, for variable speed effects, and whenever image size or frequency is changed.

• INTRA-IMAGE (Intraframe)

  A type of compression that treats each image separately.

• ITU (International Telecommunication Union)

  Formerly "CCIR". United Nations organization that defines various telecommunications standards, including 4:2:2.

• JPEG (Joint photographic experts group)

  Image compression format based on studies into how humans interpret the images they see, so as to make information losses as unnoticeable as possible.
  JPEG is one of the standards based on DCT. But then, from one manufacturer to another, the JPEG standard will itself be used according to proprietary algorithms.
  This compression consists of 6 steps:
  1> The image is broken down into 8 x 8 pixel blocks
  2> Application of DCT, which transforms the previous block into an 8 x 8 matrix, still representing frequency variations in the image.
  3> Quantification: we divide the matrix according to a chosen factor, in order to reduce the amount of information. It is only at this stage that image information is lost (LOSSY COMPRESSION).
  4> Zigzag sweep of the matrix, starting with low frequencies and ending with the highest.
  5> Variable-length coding: this essentially consists of getting rid of unnecessary zeros.
  6> Entropic coding: elements are coded according to their frequency, with the most frequent elements having the shortest codes.

• KILO-OCTET (Kilo Byte)

  1024 bytes.

• KILO-OCTET PER SECOND (Kilo Byte per Second)

  Throughput in digital transfer.

• KINESCOPE

  Device for transcribing an image from video tape to film. (KINESCOPAGE).

• LTC (longitudinal Time Code)

  Timecode signal recorded in a linear VCR track, similar to an audio track (it cannot be heard as a crackling sound).

• LUMINANCE

  The part of a component video signal that specifically contains light intensity information, without the color (this is the "black" and "white" part of the video signal).

• MEGAOCTETS (Mega Byte)

  1024 Kilobytes.

• RAM (Random Access Memory)

  The memory a computer uses to temporarily store the information it processes.

• MIDI (Musical Instrument Digital Interface)

  Protocol that tells a device how to play a piece of music.

• MJPEG (Motion JPEG)

  MJPEG is the name of a JPEG standard adapted to interlaced video images. It's not a different standard and, contrary to popular belief, it's not interframe coding: each image is processed independently of the previous and subsequent ones.

• EXTERNAL MODULE (Plugin)

  Software extension that complements the functions of a system

• Off-Line Editing

  First edit made with lower image quality, either to use cheaper equipment or to work with a resolution that takes up less space on the virtual editing computer's hard disks.

• NON-LINEAR EDITING

  Also known as virtual editing. Once images have been recorded on computer disks, editing operations can be as simple as issuing playback commands, eliminating the need to copy images as in video, or move them as in film.
  Editing therefore has no physical reality, and any part of the rushes or sequences can be accessed at any time (as opposed to linear systems, which require the unwinding of a tape to achieve this). Similarly, you can have as many versions of a sequence as you like, since they all use the same media.
  The quality of digitization now enables us to work directly on-line with compression ratios of 1:4 in 360 kb/i or 2:1 in lossless on Media 100 XS. on Media 100 844X the compression ratio is zero, and the image is coded on 10 bits instead of the usual 8 bits.

• MP@ML (Main Profile @ Main Level)

  One of the MPEG2 sub-standards used in broadcasting, for DVDs for example.

• MPEG (Motion Pictures Experts Group)

  Image and sound compression standard.
  Compression is performed on a sequence of images (interframe or GOP), for example 12. The system first stores a complete image, then only what has changed in the second image. The system regularly stores a complete image to get rid of any errors that may have propagated.
  If we go into a little more detail, MPEG compression is made up of 3 types of images:
  1> I (intraframe coded frame) images are intraframe coded like a JPEG image. They are independent of each other.
  2> P-frames (predicted frames) are predicted from an I-frame or another P-frame. For example, if we observe the sequence of numbers 3-6-9-12-15..., we can predict that the following numbers will be 18-21-24.... When we recognize the variations that occur from one image to the next, we can predict those that will follow. The compression ratio of P images is greater than that of I images.
  3> B-frames (bi-directional frames): the image is interpolated from the surrounding I or P frames. B images have the highest compression ratio.
  Typically, a 12-frame MPEG sequence (GOP 12) consists of : I-B-B-P-B-B-P-B-B-P-B-B-I-B…
  There are several standards: MPEG1, MPEG2..... MPEG1 was originally designed to work at 1.2 MB/s, so it could be read from a CD-ROM, but the quality is still poor and encoding is only done on a single frame of the two used in video. What's more, the encoding size is standardized at 352×288, i.e. half the size of a normal D1 video format.
  MPEG2 was designed to cover a wide range of qualities, from VHS to high-definition television (hdtv). It is available in a series of algorithms (profiles) and resolutions (levels).
  Two combinations are actually encountered: MPEG2 MP@ML (bitrate of 15 MB/s and GOP 12 in PAL) which can only be used for broadcasting, and MPEG2 422@ML (bitrate of 50 MB/s and gop2) which is used for high-quality editing or archiving applications.

• MPEG4

  Unlike MPEG-2, which was aimed solely at digital TV and DVD applications, MPEG-4 encompasses all new video applications such as downloading and streaming on the Internet and cell phones, digital radio, video games, TV and high-definition media.
  MPEG-4 is broken down into a series of standards, the parts, which specify a particular type of coding. In each part, several profiles (collection of algorithms) and levels (quantitative constraints) are defined. For example, H264 or MPEG-4 AVC (Advanced Video Coding) enables compression at a rate 2 times lower than MPEG-2 or even higher, with no loss of quality.
  In Europe, the Digital Video Broadcast (DVB) standardization organization approved H.264/AVC for television broadcasting in Europe at the end of 2004. In France, it has been adopted to equip the receivers of all HD televisions, and is compulsorily broadcast on DTT for all pay-TV programs.

• NTSC (National Television System Committee)

  Signal Composite mainly used in North America, Japan...

• DIGITALIZATION

  This operation transforms an analog signal into a signal consisting solely of bytes (0 and 1). At a fixed sampling frequency, according to a scale defined by the number of increments, the analog signal is measured and expressed as a sequence of numbers.

• OCTET (Byte)

  Computer unit consisting of 8 bits.

• OMF (open media framework interchange)

  File format for exchanging digital media between heterogeneous platforms, from Macintosh to PC, or Silicon Graphics for example.

• PAL (Phase Alternate Line)

  Composite signal used in Europe for production and postproduction. There is a "PAL plus" project which, within the bandwidth of a normal PAL signal, accommodates a 16:9 format signal, while still allowing compatibility with 4:3 monitors.

• PARTITION

  When we format a disk, we can divide it into several logical units for personal organization. You can change the size of the partitions, but in this case you lose everything that was saved on the disk.

• PIXEL (Picture Element)

  Elementary point of the image. This notion is valid for digital images. In chemical imaging, we tend to speak of the number of strokes per millimetre. There are two types of pixel: square and non-square. Video screens, like computer monitors, have a 4/3 format, i.e. 1.33. When computer manufacturers designed their work surfaces, they naturally chose values that were multiples of 4/3.

• PIXELISATION

  Defects in the digital image that appear, for example, when the image is magnified or the resolution used is too low.

• PRO TOOLS

  Digidesign 24-bit audio recording, editing, mixing and mastering system.

• QUICK TIME

  Macintosh architecture for creating files containing video and audio. (the PC equivalent of AVI).

• RAID (Redundant Array of Independent Disks)

  Special formatting of a group of disks to increase data throughput and/or security.

• RAID 0

  Common formatting of a set of disks to increase total volume throughput, but offering no redundancy.

• RAID 2

  Set of 9 disks, where each bit of each byte is recorded on the first 8 disks, and a parity bit on the 9th.

• RAID 3

  A set of disks plus an additional disk, where information is stored on defined sectors across the main disks, with a parity sector on the additional disk. If one of the main disks fails, it can be replaced by a new one, and the information it contained is reconstructed by calculation between the other disks and the parity disk. This is the system best suited to applications transferring large amounts of information, such as video applications.

• RAID 4

  Same principle as raid 3, but suitable for applications that randomly transfer small volumes of information, such as database applications.

• RAID 5

  Like raid 4, but the role of the parity disk is distributed across all disks.

• NOISE REDUCTION

  Operation that removes all or part of audio or video noise, to improve signal quality and compression.

• WIDE AERA NETWORK (WAN)

  A network of computers located relatively far apart, in different cities for example.

• LOCAL AERA NETWORK (LAN)

  A network of computers in relatively close proximity, for example in several offices of a company.

• ROTOSCOPYING (Rotoscoping)

  Frame-by-frame work in which a real-life shot is used as a reference and an animated graphic is added. Can be used either for special effects or to correct image defects. (Rotoscoping is often combined with functions for automating repetitive tasks).

• RED GREEN BLUE or RGB (Red Green Blue - RGB)

  The 3 primary signals created by the camera, which are returned to the video monitor for display. Each of these electrical signals is proportional to the amount of the corresponding color in the light.

• RS 422

  Serial communications standard, allowing connections of up to 300 m, and widely used in professional video to transfer commands between VTRs, editing controllers, mixers, etc. Usually uses 9-pin connectors.

• S-VIDEO

  One of the coding states after components (yc), where video is in the form of two different signals: luminance (y) , and chrominance (c). At this stage, color coding to PAL, SECAM or NTSC is already done.

• SDDI (Serial Digital Data Interface)

  Signal transmitting compressed 4:2:2 video, audio, time code and commands in serial form for transmission over a single BNC-type cable. The SDDI standard is used to transmit information other than video.

• SDI (Serial Digital Interface)

  Signal transmitting uncompressed 10-bit 4:2:2 video, 4 audio channels and time code, serialized for transmission over a single BNC 75 ohm cable. Throughput 270 mb/s.

• SECAM (Sequential Color Memory)

  Composite signal used in France, Russia and some African countries only for final broadcasting and special reception.

• SERVER (Server)

  Either the computer or the application that responds to a client's requests by sending information or performing a task on its behalf.

• OPERATING SYSTEM (OS)

  First software layer on a computer that manages the basic functions of other applications. On the Macintosh, you'll find Mac OS 9 or X; on the PC, Windows NT, 2000 or XP.

• VARIABLE BIT RATE (VBR)

  Compression that uses a variable number of kilobytes per image, depending on its complexity, such as Quick time Sorenson or Mepg 2 - this type of encoding maximizes the image's quality/weight ratio.

• TCP/IP (Transmission Control Protocol and Internet Protocol)

  A protocol that enables data to be transmitted from one computer to another, both on a local area network (LAN) and over the Internet.

• HIGH-DEFINITION TELEVISION

  16:9 format and double the vertical and horizontal definition.
  - The European format is based on 4:2:2, at 50 Hz, but in 1280 lines.
  - Japanese format at 60 Hz in 1125 lines.
  - The American NTSC format actually covers several different proposals.

• DIGITAL TELEVISION

  All digital broadcasts, whatever the medium (cable, satellite, terrestrial).

• REAL TIME

  Possibility of reading an effect on a non-linear system without having to calculate it beforehand.

• TERA OCTET (Tera Byte)

  1024 Gigabytes.

• TGA (targa)

  Image file format developed by Truevision, used on PCs and MACs, particularly for computer-generated images.

• TIFF (targed image file format)

  Bitmap image file format developed by Aldus, used on PCs and MACs for still images.

• TIMECODE

  Timecode is a numbering system that is used for both video and sound, but is increasingly used in film systems too. Timecode is based on time counting. It is generated at the time of shooting, can be modified at a later date in some cases, and can only be read by machines. The timecode is made up of two distinct numbers: the timecode itself (time code) and the user's bits (personal code).
  In the time code word, 8 digits define a number in hours, minutes, seconds and frames, for example 12:35:48:05.

• TRANSFORMATION (Morph)

  Function available on trick systems that allows you to start from one image and transform it into another.

• TRUE TYPE

  Vector font format, used on both PC and MAC.

• HD TV

  The term HD covers all digital audiovisual technologies such as HDV (consumer), HDCam (professional), high-definition television (DVB-T, DVB-S, DVB-C, DSL TV...) and Blu-ray Disc media.
  Each HDTV standard can exploit a display variant that conforms to historical analog video standards: 50 Hz (conforming to 25 frames per second PAL or SECAM) and 60 Hz (conforming to 30 frames per second NTSC).
  Resolution 1080 lines of 1920 dots in interlaced display (1080i) at 50i for Europe or 60i for North America, South America and Japan.
  Resolution 720 lines of 1280 dots in progressive display (720p) at 25p for Europe or 30p for North America, South America and Japan.
  This second version was not selected by the HDFORUM, contrary to the EBU's opinion, and is therefore not available for distribution in France.

• TWISTER

  3D object rendering engine with interactive focus, including the new version of Mental Ray. Works with Softimage 3D and any other 3D software on the market.

• WAVE

  Audio format developed by Microsoft, which can transmit sound in many different ways, and which contains a description of how the sound has been encoded.