A Huffman tree possesses the block property.

A Huffman tree possesses the sibling property.

A Huffman tree does not possess the sibling property.

The function of the update procedure is to preserve the sibling property.

The function of the decoding procedure is to preserve the sibling property.

The function of the update procedure is to leave the sibling property.

Arithmetic coding is not especially useful when dealing with sources with small alphabets.

Arithmetic coding is especially useful when dealing with sources with large alphabets.

Arithmetic coding is especially useful when dealing with sources with small alphabets.

To distinguish a sequence of symbols from another sequence of symbols, we need not tag it with a unique identifier.

To distinguish a sequence of symbols from another sequence of symbols, we need to tag it with an identifier.

To distinguish a sequence of symbols from another sequence of symbols, we need to tag it with a unique identifier.

In Arithmetic coding, we need a function that will map a sequence of symbols into a large interval.

In Arithmetic coding, we need a function that will map a sequence of symbols into the unit interval.

In Arithmetic coding, we need a function that will map a sequence of symbols into a decimal interval.

The procedure for generating the tag works by reducing the size of interval in which the tag resides.

The procedure for generating the tag works by keeping the size of interval unchanged in which the tag resides.

The procedure for generating the tag works by increasing the size of interval in which the tag resides.

Arithmetic coding, although less complicated than Huffman coding, allows us to code sequences of symbols.

Arithmetic coding, although more complicated than Huffman coding, allows us to code sequences of symbols.

Huffman coding, although more complicated than arithmetic coding, allows us to code sequences of symbols.

One of the more common forms of static dictionary coding is digram coding.

One of the more common forms of adaptive dictionary coding is digram coding.

One of the less common forms of static dictionary coding is digram coding.

The search buffer contains a portion of the recently encoded sequence.

The search buffer contains the next portion of the recently encoded sequence.

The look-ahead buffer contains a portion of the recently encoded sequence.

The distance of the pointer from the look-ahead is called the match pointer.

The distance of the pointer from the search buffer is called the offset.

The distance of the pointer from the look-ahead buffer is called the offset.

The Achilles’ heel may occur in LZW algorithm.

The Achilles’ heel may occur in LZ77 algorithm.

The Achilles’ heel may occur in LZ78 algorithm.

In the LZ78 algorithm, only the index to the dictionary is sent.

In the LZ77 algorithm, the element pair <i, c> is sent.

In the LZW algorithm, only the index to the dictionary is sent.

The LZ78 algorithm drops the reliance on the search buffer and keeps an explicit dictionary.

The LZ78 algorithm keeps reliance both on the search buffer and an explicit dictionary.

The LZ78 algorithm keeps reliance on the search buffer.

Continuous processes are modelled with continuous distributions because it can simplify the design process considerably.

Discrete processes are modelled with continuous distributions because it can simplify the design process considerably.

Continuous processes are modelled with discrete distributions because it can simplify the design process considerably.

Usually we use a midrise quantizer if the number of levels is large, and a midtread quantizer if the number of levels is small.

Usually we use a midrise quantizer if the number of levels is even, and a midtread quantizer if the number of levels is odd.

Usually we use a midrise quantizer if the number of levels is odd, and a midtread quantizer if the number of levels is even.

Nonuniform sources are often modeled by pdfs.

Nonuniform sources are often modeled by using transforms.

Nonuniform sources are often modeled by cdfs.

The unbounded error is called granular noise, and the bounded error is called overload noise.

The bounded error is called quantization noise, and the unbounded error is called granular noise.

The bounded error is called granular noise, and the unbounded error is called overload noise.

The Jayant quantizer uses quantization with two-word memory.

The Jayant quantizer uses quantization with one-word memory.

The Jayant quantizer uses quantization with three-word memory.

The adaptivity of the Jayant quatizer depends on the value of the multipliers.

The adaptivity of the Jayant quatizer depends on the representation levels.

The adaptivity of the Jayant quatizer depends on the value of the step sizes.

The amount of compression is described in terms of frequency and measured in cycles per sample.

The amount of compression is described in terms of bits and measured in samples per bits.

The amount of compression is described in terms of rate and measured in bits per sample.

The vector quantization coding scheme is very attractive as it uses a lookup table for encoding.

The vector quantization coding scheme is very attractive as it uses a lookup table for decoding.

The vector quantization coding scheme is less attractive as it uses a lookup table for decoding.

The less popular approach to designing vector quantizers is a clustering procedure.

The most popular approach to designing vector quantizers is a l-means clustering procedure.

The most popular approach to designing vector quantizers is a clustering procedure.

Use of LZW for image compression is very popular.

Use of LBG for audio compression is very popular.

Use of LBG for image compression is very popular.

In vector quantizer design, a simple initial codebook is the set of all output points.

In vector quantizer design, a simple initial codebook is the set of output points from the corresponding scalar quantizers.

In scalar quantizer design, a simple initial codebook is the set of output points from the corresponding vector quantizers.

In uniform quantization, the step size is inversely proportional to the number of levels.

In uniform quantization, the step size is directly proportional to the number of levels.

In uniform quantization, the step size is inversely proportional to the maximum input signal.