Tag : Conversion

ARC

Aspect Ratio Converters change picture aspect ratio, usually between 16:9 and 4:3. Other aspect ratios may be also allowed for, such as 14:9, and custom values are often available. Technically, the operation involves independent horizontal and vertical resizing and there are a number of choices for the display of 4:3 originals on 16:9 screens and vice versa (e.g. letterbox, pillar box, full height and full width). Whilst changing the aspect ratio of pictures, the objects within should retain their original shape with the horizontal and vertical axes expanded or contracted equally.

See also: Aspect ratio

Down-res

Decreasing the size of video images to fit another format. Typically this reduces an HD format to an SD format and, as the input images represent over-sampled versions of output, the final quality should be excellent; often better than an SD-shot original. Moving from 1080/60I to 480/60I is down-resing. Technically the process involves: spatial interpolation to reduce size while retaining quality, color correction to compensate for the difference in HD and SD color standards and possibly re-framing to fit 16:9 HD onto 4:3 SD. Note that down-res does not include any change of frame rate.

See also: Down conversion, Format conversion, Standards conversion, Up-res, Versioning

Format conversion

Changing the format of television or other moving media images without changing the vertical (frame or field) refresh rate. Starting with 1920 x 1080/50I and converting it to 720 x 576/50I is a format conversion. This only alters the format spatially, changing the vertical and horizontal size which technically is a relatively straightforward task. Note that this may include aspect ratio conversion (ARC) typically to account for the use of 16:9 and 4:3 displays.

See also: ARC, Down-res, Standards conversion, Up-res

Frame-rate conversion

The process of changing the vertical rate of moving pictures. This has become far more important with the use of HD and UHD with their inclusion of many frame rates in the standards with 23.98P, 24P, 25P, 29.97P, 30P, 50I, 50P, 59.94I, 59.94P, 60I and 60P Hz. Changing from one rate to another, including between interlaced and progressive scans, requires considerable technology to produce high quality results. Further frequencies are used in other media markets, such as mobile and some web-based areas with more restricted bandwidth, where lower rates such as 15 f/s and others are found.

Conversions that simply drop or repeat fields or frames produce poor results – especially where the video includes a lot of motion, e.g. sports. For better results some form of motion compensation is used that analyzes the motion and uses this information to create new images in between those that are coincident between the input and output.

Interpolation (spatial)

Defining the value of a new pixel from those of its near neighbors. For example, when re-positioning or re-sizing a digital image, for dramatic effect or to change picture format, more, fewer or different pixels are required from those in the original image. Simply repeating or removing pixels causes unwanted artifacts. For far better results the new pixels have to be interpolated, calculated by making suitably weighted averages of adjacent input pixels, to produce a more accurate result. The quality will depend on the techniques used; bi-cubic interpolation is generally accepted as being good, and the number of pixels (points) taken into account (hence 16-point interpolation), or area of original picture that is used to calculate the result all affect the quality of the result.

See also: Anti-aliasing, Interpolation (temporal), Sub-pixel

Interpolation (temporal)

Interpolation between the same point in space (pixel) on successive frames. It can be used to provide motion smoothing and is extensively used in standards converters to reduce the judder caused by changes of field, or frame, rates such as between 50 and 60 Hz. The technique can also be adapted to create frame averaging for special effects and slow motion. Various qualities of processing are used. It can be very complex, attempting to work out how each element in successive pictures is moving in order to synthesize ‘between’ images (e.g. to convert 50 pictures into 60 pictures while still showing smooth motion).

Motion Adaptive Processing

Motion adaptive processing is responsive to the output of a motion detector.  The motion detector may work on a global or a local basis in the picture, and may be binary or may measure on a continuous scale the amount of motion, or the confidence that an object or region is moving.  Motion adaptive processing controls a mix between processing that is optimized for static detail and processing that is optimized for moving areas.

Motion Compensated Processing

Motion compensated processing is responsive to the output of a motion estimator.  A motion estimator usually works on a local basis in the picture and measures not only the existence of motion but also its speed and direction.  Motion compensated processing typically controls spatiotemporal filters that track the motion of each part of the picture.

Standards conversion

Changing the standard of existing television material that may involve two processes (four if going from and to analog coded systems such as PAL and NTSC). The two main processes are format conversion to change the spatial (horizontal and vertical) sizes of the pictures and changing the vertical scan rate – the number of pictures per second. For broadcast applications this needs to be completed while retaining the maximum possible fidelity of the input content. The re-sizing process involves the relatively straightforward task of spatial interpolation – spreading the information from the original pixel array over a different pixel structure. Note that the crude method of dropping or repeating lines/pixels will give very poor results and the detail of the interpolation process used is important for best results.

The second process is far more complex as, changing the number of frames or fields per second (temporal conversion) means creating new ones or removing some – preferably without upsetting any movement shown in the pictures, so simply repeating or dropping fields or frames will not do. For this the movement within the pictures has to be analyzed so that ‘in-between’ pictures can be synthesized. This is a very specialized area and there are highly developed techniques used on the best modern standards converters that do this very well, but never perfectly.

See also: Format (conversion), Frame-rate conversion

Synchronization

Temporal conversion without the use of interpolation. If input and output field or frame rates are not identical then field or frame drops or repeats must occur.