Glossary of Terms

The most common HD format is the 1,080-line system. Unlike standard definition systems, HD systems quote the number of actual visible lines used for the picture. A 1080-line frame is 1920 pixels across. This format has five times the resolution of the 625-line picture currently in use. A 1080-line picture can run at:

23.98 fps used for film effect transfers to be transmitted at 29.97 fps
24 fps used for transfer to 35mm film
25 fps used for European film look transmitted in PAL
29.97 fps used in the NTSC areas of the world (particularly USA) for HD transmission
30 fps used occasionally

All of these frame rates are progressive and give film like motion. Progressive segmented frame (Psf) is used for television

The 1080-line system also supports interlace (see below) images at:

50 fields per second
59.94 fields per second
60 fields per second

SOne of the systems proposed for American high definition transmission is the 720-line system.

The 720-line system is 1280 pixels across, so the resolution is 1280 x 720 pixels, just over twice the resolution of a 625-line picture and also results in a 16/9 aspect ratio. 720-line images are usually Progressive, and have frame rates up to 60 frames per second.

One of the major HD camera manufacturers (e.g. Panasonic) make the Varicam line of cameras that shoot at a variable frame rate in 720p from 4 ~ 60 fps Progressive and is useful for the type of slow motion shooting used by nature cameramen.

The 2K-transfer process was originally developed to transfer 35mm film material into graphics computers for visual effects work.

2K refers to the number of pixels scanned across a frame, 2K is actually 2048x 1556 pixels, the number of vertical pixels depends on the aspect ratio; a 16:9 frame would be 2048 x 1152 pixels.

There are however many variations, just as there are many different film aspect ratios, however most work is done at full aperture to provide the maximum useable image so as to be able to correct camera shake etc. without having to crop the image.

Visual effects specialists and animators now use 2K, 4K and even 8K resolutions. 2K is often used for DI (Digital Intermediate) production of film prints for theatrical release, although high end movies are starting to move to 4k scans (4096 x 3112 pixels) which preserve even more of the detail and range captured by modern film stocks which currently have a resolution of approximately 6K

AVC Intra 100 is an advanced intra-frame codec for HD Video. Capable of encoding 1080 and 720 frame sizes, it is a Full Raster 4:2:2 10-bit codec. Each frame is individually compressed, but only the differences between frames are recorded, making it an efficient codec. AVC Intra 100 can be edited natively in AVID systems with MX hardware, but currently has to be transcoded to ProRes HQ for Final Cut Pro.  The data rate is 100 Mb/s, so for 1080 HD 1 minute of recording requires 1GB of storage.

AVC Intra 100 is recorded onto P2 Media cards in Panasonic HD camcorders or onto the standalone Panasonic AG-HPG20 (P2 Gear) miniature field recorder.

A method of reproducing colour images in a single sensor electronic camera.

Devised during the early development of digital stills cameras in the 1970s by Dr Bryce Bayer, of Eastman Kodak, who employed an array of red green and blue filters on a matrix or “mosaic” of photosensors.

Once recorded this raw data must be processed before viewing. This process is known as “de-mosaicing” or “de-Bayering” and can be performed at various resolution and quality settings. To emulate the inherent sensitivity of the human eye to the visible spectrum twice as many Green sensors as Red or Blue are generally used.

Bit rate is measured as "bits per second" (bps) and refers to the rate at which data is transmitted. As a general rule the more data the better the resulting pictures/sound.

Compression / Decompression also Coder / Decoder.

This has come to mean both a device and piece of software for converting one digital format to another or a method of reducing the amount of data/information in a digital signal so that it can be recorded or broadcast more efficiently. High Def transmission is highly compressed as there is simply too much information in a full High Def image to transmit over the limited bandwidth available. Viewers watching High Def broadcasts at home are in fact watching a very highly compressed version of the original, this can result in artifacts such as ‘blockiness’ or ‘mosquito noise’ in uniform or smoky areas and in areas of fast motion. Codecs are generally defined as “lossless” or “lossy” – this refers to the amount of information that is discarded irretrievably during the compression process. Because of the high data rates produced in HD TV and Digital Cinematography and the real-time recording constraint, most codecs used are lossy.

4:4:4 and 4:2:2 Scary looking numbers you might hear bandied about by people who don’t necessarily know what they mean either!

Due to storage and transmission limitations, there has been, until very recently, a need to reduce or compress the signal. Since the human eye is much more sensitive to variations in brightness than colour, a video system can afford to devote less ‘space’ or ‘bandwidth’ to the part of the signal that deals with colour than to the part that deals with brightness.

A system that records in 4:2:2 devotes twice as much space to brightness versus the colour; it requires two-thirds the bandwidth of the full 4:4:4 signal and the human eye can’t tell the difference. The reason why 4:4:4 is used is that there is more colour and exposure latitude stored in the image for a greater range of post production grading and visual effects. Chroma keying (blue or green screen work) is also easier and cleaner in 4:4:4 because there is no loss of colour data.  HDCAM SR can record in 4:4:4 or 4:2:2, whilst HDCAM and DVCPro HD record in 4:2:2 whilst HDV only records in 4:2:0, the further reduction in colour information making it unsuitable for chroma key work.

Cineform Inc. has developed a family of precision codecs for acquisition and DI workflows. Capable of encoding a range of frame sizes and frame rates. Cineform is usually recorded on a Windows platform but is compatible in post with AVID, FCP and Adobe Premiere. The SI-2K camera system uses Cineform as its acquisition format

Humans are able to perceive a range of luminance values, colours and saturations.
Display devices can only partially represent this range or ‘colour space’

P3 is currently the widest colour space or Gamut which can be electronically reproduced by DLP (Digital Light Processing) projectors. LCD screens and CRT monitors have a narrower colour space and are unable to accurately reproduce certain colours, especially in the cyan/green part of the visible spectrum.
Film for Projection has a considerably wider colour gamut than any electronically generated image, but can have reduced contrast ratios when compared to DLP projectors.

As technology improves, possibly with the development of laser projectors, a far wider range of colours will be capable of being reproduced with higher contrast ratios as true “blacks” will be reproduced by switching the laser off.

Often used for any information that is not in standard video format Usually in .dpx file format (Digital Picture Exchange) This is becoming widely used as it can contain any metadata (timecode, colourspace information, camera details etc.) as well as the image. This is uncompressed and can be accessed directly by other computers.

Other data formats can include Quick Time for Final Cut Pro or MXF for Avid editing which can be compressed.

Data from some digital cameras can be recorded directly onto HDCam SR tape. However they can all be recorded as data files onto a variety of media including the S-Two data recorder, Codex data recorder, or on-camera, solid state data magazines.

Digital Cinema Initiative. – a working group set up by the major Hollywood Studios to define standards for Digital Cinema (d-cinema) data and projection.

The process where film grain is reduced to enhance picture quality and aid acceptability for HD transmission. New techniques preserve even more of the detail and range captured by modern film stocks which currently have a resolution of approximately 6K

Generally a digital intermediate is a digital file or files resulting either from a scan of a negative film original or ingest of digital files from digital acquisition, which is used for editing, effects and colour grading. It is the material that is used in DI labs and constitutes the whole film. As such it should carry all the useful information that is contained in the original camera negative to provide both the latitude and sharpness of the original.

Digital Imaging Technician.  

This grade came about (originally in America where it is a union grade) due to the increasingly complex menus found in HD cameras. Setting up, running and monitoring an HD camera can be more technical than a film trained Cinematographer or crew might feel comfortable with. Some DITs promote themselves as being able to provide the ‘look’ the DoP wants through manipulating the menus in-camera. As described elsewhere, this can be a risky practice as it often involves ‘throwing’ away data, although it is now possible to manipulate the ‘look’ without throwing anything away through the use of a ‘Look up Table’ which can be applied to the image when viewed on set or by the editing and visual effects departments. Newer cameras tend to have much simpler menus, and should require less in-field maintenance. Current practice is to capture as much information as possible, lighting and composing the scene as you wish it to look, leaving any final manipulation of the image until the on-line grade, just as one would when shooting film. A competent film crew will have no problem transitioning to HD / digital in terms of setting up and running the cameras, but having a DIT with specialist knowledge may prove useful on larger productions using multiple cameras.
As recording formats transition from HD to Data, the DIT’s job is expanding to encompass logging, metadata, archive and backup (eg to LTO 4 tape or external hard drives) exporting compressed HD (pro res 422/Avid codec) for editing as well as producing DVD and Blueray discs for rushes viewing. This is a complex and time consuming job that absolutely cannot be done by a trainee! A good video assist technician is also becoming more essential at every level of production.

The process of changing high definition pictures to standard definition (e.g 625-line) for transmission or inclusion in standard definition programmes or for monitoring or recording on set. There are specialist down conversion units that can change the aspect of HD pictures or even pan and scan wide screen for 4:3 delivery. Most broadcast HD VTRs and some cameras have built in down converters and these are often used for direct transfers of 16:9 HD to 16:9 DigiBeta or DV tapes.

DVCProHD is a codec developed by Panasonic from the Standard Definition DVCPro 50. It is similar in concept to Sony’s HDCam format. It was first employed in the 720p 4:2:2 HD 1/4-inch tape Varicam, and has a 100Mb/sec recording capability. It can be recorded to HD videotape or to P2 Media cards, and used in camcorders (e.g. Varicam) or as a studio / OB record tape for HD in both 720 and 1080 formats.
Varicam allows 4 – 60 fps recording.

Digital Video Interface. A high-bandwidth digital connection between a video source and a display device.

A type of connection which modulates a laser to carry a signal (such as a single or dual link HDSDI signal) over a fine glass core contained within a robust sheath.  Allows much longer cable lengths and much thinner cables than conventional copper cables, as well as being immune from electrical interference. However, fibre optic connections can be susceptible to dust and dirt and may require specialist cleaning.

There have always been three main frame rates:

24 frames per second used for film acquisition and cinema projection
25 frames per second for European television (actually 50i fields/sec)
30 frames per second for US and Japanese television (actually 60i fields/sec)

The rate of 30fps is always quoted for convenience; it is actually 29.97 frames (or 59.94 fields) per second. For HDTV productions at a nominal 24fps, 23.98fps is employed, as it is a divisible of the 59.94 standard. As technology improves, 50P and 60P (Progressive) will become a standard for HDTV and cinema digital projection. These formats will enable flickerless display without the artifacts of an interlaced image.

These are solid state (no moving parts) recording devices that can record the full HD signal.

At present the flash mags used with the Arriflex D-21, Panvision Genesis or the Thomson Viper can record 10 minutes in 4:4:4 or 15 minutes in 4:2:2 at 24 or 25fps. The data must then be transferred to tape or to a hard drive for more permanent storage.

P2 is used by Panasonic P2 HD cameras and AG-HPG20 miniature field recorders, using the AVC Intra or DVCProHD codecs. 1 Gb per minute at AVC Intra 100 and 0.5 GB per minute at AVC Intra 50.

P2 cards are currently available up to 64Gb, a figure likely to increase. They can be imported directly into non-linear editing programmes such as Avid and Final Cut Pro .

Roughly speaking, a description of the relative contrast range of an image capture device or an image display.

Film, CMOS and CCD imagers all have different sensitivities as the light falling on them increases or decreases in level. With electronic capture sensors the relative sensitivities for low, mid and highlights can be electronically corrected to closer emulate the linear characteristics of film.

Commonly used settings for electronic cameras include log(s) which emulates the “S-curve” of film and Rec 709 used in HD TV broadcasting

Sony 4:2:2 HD 1/2-inch tape format recorded to a cassette developed from DigiBeta. Always 8 bit, the colour sampling is 4:2:2 in studio recorders, but 3:1:1 in HDCAM Camcorders with significant compression applied. Can be used in camcorders (e.g. HDW-900R) or as a studio / OB record tape for HD. Copying one tape to another will result in generational quality loss unlike HDCAM SR.

Sony 1/2-inch tape format with full colour resolution (i.e. 4:4:4) and less compression.  Both the Arri D-21 and the Panavision Genesis often use this format to record in.

It can record Single Stream 4:2:2 (YUV) or Dual Stream 4:4:4 (RGB) 1080 or 720 HD and at 880 Mb/s (1.6:1 compression) or 440 Mb/s (3.2:1 compression).

The Dual Stream record function can also be employed to record the two separate images of a 3D stereo camera system.

Well established 4:2:2 HD recording format developed by Panasonic. Due to its early availability widely used for recording telecine transfers of film elements for HD post. Particularly popular in USA but under a threat for supremacy by HDCAM SR. Panasonic has never made an attempt to develop companion HD-D5 camcorder.

High Definition Video. A 1/4 inch 4:2:0 JVC / Sony / Canon prosumer HD format, not to be confused with DVCProHD. Not acceptable for full HD production and no good for Chromakeying.

High Definition Multimedia Interface. A high-bandwidth digital connection for both video and audio data. HDMI ensures the best video signal is sent from the source (HDTV signal) to the display (LCD, Plasma, DVL). It does this by sending video and multi-channel audio data to the display through one single cable.

The digital signal format that allows HD images to be sent over a single BNC cable (in the case of 4:2:2 HD) or over a pair of BNC cables (‘dual link’ for 4:4:4 HD).  With a typical bandwidth of 1.85 Gb/sec for 4:2:2 the cable length is limited to around 50m with typical cable and to a maximum of about 100m with high quality cable without any joints.  The use of fibre optic links allows much longer transmission paths without quality loss, although care must be taken to keep connections free of dust and dirt.

When shooting digital, one problem to watch out for is dark and black tones not being rendered accurately in daylight conditions. Many HD / Digital cameras are over sensitive to the Infra Red component in daylight, and this can show up in certain black and dark green dyes and fabrics (especially man made fibres) as a reddish muddy brown or magenta cast. For instance, an all black costume made with a mix of natural and man made fabrics, may show up as a mix of black and reddish brown on camera. This effect can be more apparent when ND filters are used on the camera (especially ND.9 or above). Only testing in bright daylight or under powerful HMI (daylight balanced) lights will show this up. Camera Filters are now available to reduce this effect such as Schneider True Cut IR & Tiffen Full Spectrum IR ND and Tiffen Clear Hot Mirror for CMOS chips. This colour cast is difficult and time consuming to correct in post. 

An interlaced picture is an image separated into two fields consisting of its even and uneven lines. This was originally done because there isn’t enough bandwidth (space) to send the picture through ‘complete’ in one go, so the picture is divided into odd and even lines and combined later to make a frame. Interlace records and transmits less information than Progressive but at a faster rate. Interlace is used for sport, news current affairs etc. as it is good for slow motion and fast moving objects as it displays less strobing and motion blur. It is also partially responsible for the ‘video’ look and the feeling of immediacy that it engenders.

A term loosely used to describe a type of image suited to direct viewing on a conventional HD monitor.

Grading linear images is generally preferred as the results are more controllable and predictable than dealing with Log since they can more easily be viewed on standard video displays without the need for correction. If you are shooting HD for television broadcast it is most likely you will shoot linear.

Industry standard for robust data backup to magnetic tape.  It is increasingly popular as a delivery format for DI, such as dpx. There are multiple generations of LTO, released at approx 2 yearly intervals, each improving upon the capacity and data transfer rate of the last. Backwards compatibility is not fully supported at each new introduction, so re-archiving should be considered. Perversely, the tape itself has a stable lifetime of 30 years.

The latest, LTO-4, can hold 800GB of data (1.6TB if compressed), and has a maximum data transfer rate of 120MB/s.

It is important to note that LTO tape drives are compatible with IT infrastructure but not with video. Also the relatively slow read/write characteristics make it impractical for use in the field.

‘LOG’ reproduces how the human eye perceives luminance, giving more data to the shadows (where the human eye can differentiate more detail) than to the highlights. An image recorded in ‘LOG’ will appear to have low contrast and lifted black levels if viewed directly on a conventional monitor, but can be corrected for viewing with a suitably programmed preview processor (or “LUT box”).  Shooting ‘ LOG’ can have advantages as a recording mode for feature film work, where post production will follow a DI route.  Film projects going through a DI process conventionally use images scanned in a LOG mode so as to capture as much of the original film image as possible.  In certain HD cameras, shooting in LOG may achieve slightly increased shadow and / or highlight detail. Anything shown on television will have to be converted back to linear.

Look Up Table: not a piece of furniture, but a way of changing the look of an image in contrast and/or colour without affecting the original (non-destructive grading). If one attempts to control the image in terms of contrast or colour ‘in camera’ using the in-built menus, one is potentially throwing data away. e.g. If you make something monochrome in camera, you will never get the colour back later. Applying a LUT using appropriate equipment (e.g. Truelight, LUTher, Speedgrade etc.) allows the DoP to demonstrate a particular look to the director on-set as well as with the editing and visual effects departments later on, without changing or throwing away any of the recorded image data. A 1D LUT controls contrast, brightness and gamma but not colour. A 3D LUT effectively controls all that a 1D LUT does per colour. Hence adding control of colour saturation and hue as well as secondary colour correction.

Metadata can cover any information that the user wishes to be carried with the image, e.g timecode, lens type and focus/iris info. When & where it was shot, who owns the rights, how it should be processed in post etc.

Motion Picture Experts Group

MPEG2 was the first practical compression scheme for HD images.

MPEG4 is an efficient mechanism for compressing data. H264 is a popular variant used in HD TV in the UK.

A normal ‘Academy frame of 35mm film, is ‘pulled’ through the camera gate four perforations at a time. This will allow for all aspect ratios including anamorphic. To save film stock (more than 25%), the camera can be modified to ‘pull’ 3 perforations through at a time. This still works for the normal cinema screen aspect ratio of 1:1.85 and also works for the 16:9 aspect ratio currently used by television broadcasters. A recently revived option is where the camera only pulls through 2-perfs at a time, saving over 50% in stock and processing and allowing for 20 minutes of filming per 1000 feet of stock. This is only suitable if you want to shoot the widescreen aspect ratio of 2.39:1 for feature projection, however, by cropping the sides, it can also be used in 16/9 mode for possible television transmission. This results in less grain than Super 16mm and all the benefits of 35mm in terms of stability, range of lenses etc.. Another benefit of 3 and 2 perf. modified cameras is that they run more quietly as less film is passing through the gate every second. On the minus side there is less space between frames to hide a hair in the gate.

The basic “element” of a picture. The resolution of a television picture can be defined (horizontally and vertically) in pixels just like a computer image. Rows of horizontal pixels making up a television picture are more usually known as lines.

Standard Definition TV is 720 pixels x 576 (active) lines. High Definition TV is 1920 x 1080 or 1280 x 720.

‘Stuck’ and ‘Hot’ pixels are not recalcitrant or sexy! They are visible as bright blue, green, red or white static dots on the screen and are caused by a faulty pixel in the sensor (you can also get these on LCD screens, which can be a problem if using an HD LCD monitor as it’s hard to tell if it’s the camera or the monitor at fault). Stuck/hot pixels are not usually fixed, but can be ‘mapped out’ (hidden) so that the camera no longer sees them. They can often be removed in post in a similar manner to dust and dirt removal on film

The complete frame exposed in one go – all the lines scanned from top to bottom, like a snapshot rather than the normal TV system of interlaced capture (see above). Progressively scanned pictures have around thirty percent more resolution than an interlaced picture and appear less busy, as they don’t have the problem of interlace line twitter with strong horizontal lines i.e. Venetian blinds.

Shooting in Progressive mode gives HD video a more filmic, less ‘video’ look that may be more suited to a drama production aesthetic. Progressive images are also more suited to viewing on large screens and displays. Progressive frames are preferred by visual effects departments as they have a smaller file size, and are easier to handle within software packages.

Raw image data is acquired when no processing or compression is carried out in-camera and only the pure data from the sensor is recorded. This data requires subsequent processing outside the camera in order to both view and use the material, but can result in better image quality.

A proprietary codec employed in the RED camera system.

Raw Bayer data from the 4K single CMOS sensor is compressed before recording using a relatively lossy wavelet codec.

RED camera can record at two quality settings, REDcode 28 (28MB/s approx 12.5:1 compression) and REDcode 36 (36MB/s approx 9.5:1 compression)
See also Bayer Encoding.

The amount of detail a device can capture or show. This depends on the number of pixels in a frame. The resolution of a frame is described as: the number of pixels horizontally across the picture by the number of pixels (or lines) vertically. HD has 1920x1080 (2,073,600 pixels) or 1280x720 (921,600 pixels). In a film image the resolution is measured in lines per millimetre – current medium speed filmstocks are roughly 6144x4668.

There are a variety of standards. You have a choice of a number of line, field and frame rates. With careful discussion before you start shooting your production, everything will work fine.

Examples:
720 line, 50 Frames, Progressive (labeled as - 720p50)
1080 line, 50 Frames, Interlaced (labeled as - 1080i50)
As time progresses we would ideally master on
1080 line, 50 Frames, Progressive (labeled as - 1080p50)

In the UK our current Standard Definition television system is known as the 625-line system. Not all the lines are used for the picture though. Only 576 lines are actually used for picture information. So a Standard Definition 625 line picture is 576 active lines or pixels high. A standard definition digital picture is 702 pixels across, so the resolution of the picture is 702 x 576 pixels. There are the same numbers of pixels across the frame for both 4:3 and 16:9 images. HD is always 16:9.

The BBC considers the following formats to be standard definition:

1. All Standard Definition video formats
2. HDV from all manufactures
3. Cameras with image sensors under 1/2”
4. Super16mm film whether transferred to tape in high definition or not.

Many people debate and refute this point. Super 16mm film telecined in High Definition IS unquestionably HD. The present challenge lies with compressing it sympathetically enough for broadcast. Various de-graining technologies exist that combined with better codecs than the BBC currently use could allow Super 16mm film to be broadcast successfully.

5. 35mm film transferred to standard definition tape formats
6. Non-linear editing codecs with bit rates below 160Mb
7. Live contributions via links at less than 60Mbs (MPEG2)

The BBC will accept High Definition programme material acquired using either of the following:

1920 x 1080 interlace at 25 frames a second (now called 1080i25) or
1920 x 1080 progressive at 25 frames a second (now called 1080p25)

All delivered high definition master tapes must be 1080i25 (whether acquired 1080p25 or 1080i25).
Rollers and moving captions must be added in 1080i25 to prevent unacceptable judder.
High Definition Film Effect. Most High Definition cameras can capture in both Interlace and Progressive modes. It is not acceptable to add film effect to high definition images for high definition delivery.

Electronic cameras usually have electronic shutters. These are, by default, at 1/25th second averaged over two fields so each field is exposed for 1/50th second. Shutters are important for low frame rate progressive images in HD i.e. 24 and 25 fps. As you’re only exposing one frame rather than two fields the exposure time for your progressive image is 1/25th sec, which will cause movement in the frame to smear. To overcome this you have to set the shutter to 1/50th sec. A shutter is used to shorten the time the camera has to capture the frame without changing the frame rate. Film cameras have a mechanical shutter, which mostly rely on a spinning disc with a variable slot cut in it.

The length of time a shutter is open can be measured either as a speed (e.g. 1/50th of a second) or as an angle. The term ‘angle’ refers to the mechanical shutters in film cameras and certain digital cameras. A 180° shutter is open for half the frame rate - i.e. at 25 frames a second a 180° shutter is open for 1/50th of a second.

The process of changing standard definition (576-line) to high definition. Although some broadcast VTRs have ‘up converters’ built in, currently these are not recommended for the high quality images HD broadcasters require. Specialist ‘up converters’ are available that can deal with some of the problems (e.g. dirt, video noise, grain etc) that can make ‘up conversion’ look poor.

Often asked for or included in HD kit. This can either be a small on-camera monitor, a stand-alone unit or a feature on the main monitor. It allows the DoP to correctly expose the image by checking for clipped highlights (visible as peaks with the top cut off) or clipped blacks (visible as a flat line at the bottom of the scale). The waveform monitor provides the DoP with an absolute indicator of how the camera is responding to light levels, even if the HD monitor has not been correctly set up. If it is clipped on the waveform, it won’t be recorded and cannot be rescued later on.

This Sony system records 1080i 4:2:0 HD onto disc with a recording time of up to two hours at 18Mb/s down to 60 minutes at 35Mb/s.