Nokia 9290 Video Style Guide - Page 9

Nokia Mobile Phones Nokia 9290 Communicator Video Editing for the Nokia 9290 Communicator 9(12) Multimedia converter provides two options to control video compression: • Smooth motion, small image size. • Normal motion, normal image size. When the first option is used, the maximum amount of pixels in the compressed image is 8960, i.e., the image width can be 112 pixels and height 80 pixels, for example. The maximum width of the picture is 128 pixels and the maximum height is 96 pixels. The actually coded image size depends on the aspect ratio of the input images. Ten frames per second are coded. When the second option is used, the maximum image size is 128x96. The image size that is actually coded depends on the aspect ratio of the input images. 7.5 frames per second are coded. The resulting file size and bit-rate depend largely on the contents of the image sequence. Typically, video bit rates vary from 50 kbps to 100 kbps. "Smooth motion, small image size" option tends to reserve slightly more bits than " Normal motion, normal image size". 5.3 Basics of Audio Coding Arbitrary sounds can be represented as a sum of waves having different frequencies and amplitudes. In other words, any sound is an amplitude waveform as a function of time. Sounds can be digitised when samples of the corresponding waveform are taken frequently enough. For arbitrary sounds and music, a 44.1 kHz sampling frequency is considered to provide high quality. For speech, an 8 kHz sampling frequency is adequate for most applications. Typically, 16 bits is enough to represent one sample. Digitised audio can be compressed in various ways. A simple coding method is to use an adaptive step size to quantise audio samples. Such a technique is used in IMA ADPCM audio coding standard that reserves 4 bits per sample. Consequently, if the sampling frequency is 8 kHz, IMA ADPCM coded audio takes 32 kbps. Another simple audio coding method is A-law PCM, which uses a logarithmic quantisation step size and reserves 8 bits per sample. More advanced audio coding methods take advantage on the human psychoacoustic model. Parts of the audio signal are barely audible and can be discarded or compressed. Typically, the advanced coding audio methods are categorised into generic audio coding and speech coding techniques. Generic audio coding algorithms are targeted for music and sound as well as human voices, whereas speech coding algorithms are aimed at speech only and perform relatively poorly when music is coded. One of the most advanced speech coding standards today is the adaptive multi-rate (AMR) speech codec, which was developed by the European Telecommunications Standards Institute (ETSI). It includes eight speech coding modes, whose bit rates range from 4.75 to 12.2 kbps. Some of the modes are speech codecs specified for other standards. For example, AMR at 12.2 kbps is the same speech codec as GSM enhanced full-rate codec. 5.4 Audio Coding for the Nokia 9290 Communicator The Nokia 9290 Communicator supports an output of 8 kHz mono sounds. Therefore, the following procedure is used to compress high-quality audio tracks in Multimedia converter: • The audio track is extracted from the input file and decompressed if necessary. Copyright  Nokia Corporation 2001-2002. All rights reserved.