Description:SAME AS UPLOADED , Claims:1. An apparatus for decoding a media signal from a data stream, the apparatus comprising:
a processor; and
a memory storing instructions that, when executed by the processor, cause the processor to:
sequentially decode a sequence of samples (13) which describe the media signal by:
selecting (54), for a current sample (13’), a set (48) of reconstruction levels out of a plurality (50) of reconstruction level sets (52) depending on quantization indices (58) decoded from the data stream (14) for previous samples of the sequence of samples,
entropy decoding a quantization index (56) for the current sample (13’) from the data stream (14), wherein the quantization index (56) indicates one reconstruction level out of the selected set (48) of reconstruction levels for the current sample, and
dequantizing (62) the current sample (13’) onto the one reconstruction level of the selected set (48) of reconstruction levels that is indicated by the quantization index (56) for the current sample.

2. The apparatus of claim 1,
wherein the media signal comprises a picture (212) and the sequence of samples representing transform coefficients of a transform coefficient block (10), scanned along a predetermined coefficient scan (14) so that the decoding the sequence of samples yields the transform coefficient block (10),
wherein the apparatus is configured to subject (36) the transform coefficient block (10) to an inverse transformation to obtain a picture block (284) of the picture (212).

3. The apparatus of claim 1,
wherein the media signal comprises a picture (212),
wherein the apparatus is configured to predict the picture content of the picture (212) within a picture block (284) of the picture,
wherein the sequence of samples (13) represents picture samples of a prediction residual of the prediction of the picture content of the picture within the picture block so that the decoding the sequence of samples yields the picture samples of the prediction residual,
wherein the apparatus is configured to combine the picture samples of the prediction residual with the prediction of the picture content of the picture within the picture block to obtain a block of reconstructed samples for the picture.

4. The apparatus of claim 1,
wherein the media signal comprises a picture (212),
wherein the apparatus is configured to predict the picture content of the picture (212) within a picture block (284) of the picture,
wherein the sequence of samples (13) is formed by transform coefficients of a transform coefficient block (10), scanned along a predetermined coefficient scan (14), so that the decoding the sequence of samples yields the transform coefficient block (10), and
wherein the apparatus is configured to:
subject (36) the transform coefficient block (10) to an inverse transformation to obtain a block of residual samples, and
combine the block of residual samples with the prediction of the picture content of the picture within the picture block to reconstruct the picture within the picture block.

5. The apparatus of claim 4, wherein the apparatus is configured to:
perform the combination by adding the block of residual samples and the prediction of the picture content of the picture within the picture block.

6. The apparatus of one of claims 3 to 5, wherein the apparatus is configured to:
perform the prediction of the picture content of the picture within the picture block by:
intra- prediction, or
inter prediction.

7. The apparatus of claim 6, wherein the data stream includes an indication indicating whether intra- or inter prediction is used.

8. The apparatus of one of claims 1 to 7, wherein the number of reconstruction level sets (52) of the plurality (50) of reconstruction level sets (52) is two.

9. The apparatus of one of claims 1 to 8, wherein the apparatus is configured to:
parametrize the plurality (50) of reconstruction level sets (52) by way of a predetermined quantization step size, and
derive information on the predetermined quantization step size from the data stream (14).

10. The apparatus of one of claims 1 to 9,
wherein the media signal comprises a picture (212), and
wherein the apparatus is configured to:
partition the picture (212) into picture blocks (284),
derive information on a predetermined quantization step size from the data stream (14) in a manner varying among the picture blocks, and
parametrize the plurality (50) of quantization level sets (52) by way of the predetermined quantization step size.

11. The apparatus of any of claims 1 to 10,
wherein the media signal comprises a picture (212), and
wherein the apparatus is configured to:
partition the picture (212) into picture blocks (284),
reconstruct each of a subset of the picture blocks (284) of the picture (212) by way of an inverse transformation of a transform coefficient block (10), wherein the transform coefficients of a predetermined transform coefficient block, scanned along a predetermined coefficient scan, form the sequence of samples,
derive information on a predetermined quantization step size from the data stream (14) in a manner varying among subset of transform blocks, and
parametrize the plurality (50) of quantization level sets (52) by way of the predetermined quantization step size.

12. The apparatus of claim 10 or 11,
wherein the predetermined quantization step size is defined by a quantization parameter that applies to a single picture block or a group of picture blocks and the quantization parameter is derived by:
predicting a quantization parameter for a predetermined picture block based on quantization parameters of neighboring picture blocks;
entropy decoding a quantization parameter difference for the predetermined picture block or group of picture blocks from the data stream; and
adding the quantization parameter difference to the prediction of the quantization parameter to obtain the quantization parameter for the predetermined picture block or group of picture blocks.

13. The apparatus of claims 11 or 12,
wherein the sequence of samples are transform coefficients of a transform coefficient block, scanned along a predetermined coefficient scan (14), and
wherein the apparatus is configured to:
derive a base quantization step size for the transform coefficient block from the data stream,
derive scale information from the data stream which defines as to how the base quantization step size is scaled in order to obtain quantization step sizes for the transform coefficients of the transform coefficient block (10) so that the quantization step sizes vary across the transform coefficient locations inside the transform coefficient block (10), and
parametrize the plurality (50) of quantization level sets (52) by a predetermined quantization step size obtained by scaling the base quantization step size according to the scale information.

14. The apparatus of one of claims 8 to 12,
wherein each of the plurality (50) of reconstruction level sets (52) for the current sample consists of integer multiples of a predetermined quantization step size, and
wherein the quantization step size is the same for all reconstruction level sets of the plurality (50) of reconstruction level sets (52) for the current sample.

15. The apparatus of claims 1 to 14,
wherein all reconstruction levels of all reconstruction level sets represent integer multiples of a predetermined quantization step size, and
wherein the apparatus is configured to dequantize the samples by:
deriving, for each sample, an intermediate integer value depending on the selected reconstruction level set for the respective sample and the entropy decoded quantization index for the respective sample, and
multiplying, for each sample, the intermediate value for the respective sample with the predetermined quantization step size for the respective sample.

16. The apparatus of claim 15,
wherein the number of reconstruction level sets (52) of the plurality (50) of reconstruction level sets (52) is two, and
wherein the apparatus is configured to derive the intermediate value for each sample by:
if the selected reconstruction level set for the respective sample is a first set, multiply the quantization index for the respective sample by two to obtain the intermediate value for the respective sample; and
if the selected reconstruction level set for a respective sample is a second set and the quantization index for the respective sample is equal to zero, set the intermediate value for the respective sample equal to zero; and
if the selected reconstruction level set for a respective sample is a second set and the quantization index for the respective sample is greater than zero, multiply the quantization index for the respective sample by two and subtract one from the result of the multiplication to obtain the intermediate value for the respective sample; and
if the selected reconstruction level set for a current sample is a second set and the quantization index for the respective sample is less than zero, multiply the quantization index for the respective sample by two and add one to the result of the multiplication to obtain the intermediate value for the respective sample.

17. The apparatus of one of claims 1 to 16, wherein the apparatus is configured to:
select (54), for the current sample (13’), the set (48) of reconstruction levels out of the plurality (50) of reconstruction level sets (52) depending on a LSB portion or previously decoded bins of a binarization of the quantization indices (58) decoded from the data stream (14) for previous samples of the sequence of samples.

18. The apparatus of any of claims 1 to 16, wherein the apparatus is configured to:
select (54), for the current sample (13’), the set (48) of reconstruction levels out of the plurality (50) of reconstruction level sets (52) depending on the results of a binary function of the quantization indices (58) decoded from the data stream (14) for previous samples of the sequence of samples.

19. The apparatus of one of claims 1 to 18, wherein the apparatus is configured to:
select (54), for the current sample (13’), the set (48) of quantization levels out of the plurality (50) of quantization level sets (52) depending on a parity of the quantization indices (56) decoded from the data stream (14) for previous samples of the sequence of samples.

20. The apparatus of one of claims 1 to 19, wherein the apparatus is configured to:
decode the quantization index (56) for the current sample (13’) in form of an absolute value which is indicative of the absolute of:
a rank distance between a rank of zero and a rank of the one reconstruction level when ordering the selected set of quantization levels according to their values and, if zero is not included in the selected set of quantization levels, a rank distance between the rank of the one level and a rank of a smallest level of equal sign, when ordering the set of quantization levels according to their values, plus one, or
a rank distance between a rank of a predetermined level in the set of quantization levels which is of minimum absolute value, and a rank of the one reconstruction level when ordering the set of quantization levels according to their values, and if the absolute value is greater than zero, a sign value which is indicative of the sign of the one reconstruction level.

21. The apparatus of one of claims 1 to 20,
wherein the number of reconstruction level sets (52) of the plurality (50) of reconstruction level sets (52) is two, and a first reconstruction level set of the plurality (50) of reconstruction level sets (52) comprises zero and even multiples of a predetermined quantization step size, and a second reconstruction level set of the plurality (50) of reconstruction level sets (52) comprises zero and odd multiples of the predetermined quantization step size, and
wherein the apparatus is configured to:
decode the quantization index (56) for the current sample (13’) in form of an absolute value and, if the absolute value is greater than zero, a sign value which is indicative of the sign of the one reconstruction level, and
determine the reconstruction level from the absolute value and the sign value to be a first function applied to the absolute value and the sign if the selected reconstruction level set is the first reconstruction level set and a second function applied to the absolute value and the sign, if the selected reconstruction level set is the second reconstruction level set, with the first and second functions being symmetric with respect to the absolute value and reconstruction level.

22. The apparatus of any of claims 1 to 21, wherein the apparatus is configured to:
decode an absolute value of the quantization index (56) for the current sample (13’) from the data stream using a binarization of the absolute value which comprises a first bin which specifies whether the absolute value is greater than zero or not.

23. The apparatus of claim 22, wherein the binarization of the absolute value further comprises a second bin (100) which specifies whether the absolute value is greater than one or not, where the second bin is only included in the data stream if the first bin indicates that the absolute value is greater than zero.

24. The apparatus of claims 22 or 23, wherein the binarization of the absolute value further comprises a further bin (100) which specifies a parity of the absolute value, where the further bin is only included in the data stream if the first bin indicates that the absolute value is greater than zero

25. The apparatus of claim 23, wherein the binarization of the absolute value further comprises a further bin (100) which specifies a parity of the absolute value, where the further bin is only included in the data stream if the second bin is included in the data stream and the second bin indicates that the absolute value is greater than one.

26. The apparatus of one of claims 1 to 35, wherein the apparatus is configured to:
decode an absolute value of the quantization index (56) for the current sample (13’) from the data stream using binary arithmetic decoding by:
entropy decoding (85’) a first bin of a bin string onto which the absolute value is binarized, which first bin specifies whether the absolute value is greater than zero or not, using a first adaptive probability model,
wherein the probability model is selected among a set of adaptive probability models and the selection depends on the set (48) of reconstruction levels selected for the current sample (13’).

27. The apparatus of one of claims 20 to 25, wherein the apparatus is configured to:
decode an absolute value of the quantization index (56) for the current sample (13’) from the data stream using binary arithmetic decoding by:
entropy decoding (85’) a first bin of a bin string onto which the absolute value is binarized, which first bin specifies whether the absolute value is greater than zero or not, using a first adaptive probability model, wherein the probability model is selected among a set of adaptive probability models and the selection depends on the state for the current sample.

28. The apparatus of claim 26 or 27, wherein the apparatus is configured so that the selection of the first adaptive probability model further depends on a parity of the quantization index for an immediately preceding sample of the sequence of samples.

29. The apparatus of one of claims 26 to 28, wherein the apparatus is configured so that the selection of the first adaptive probability model further depends on one or more preceding samples of the sequence of samples having a predetermined locational relationship to the current sample.

30. The apparatus of claim 29, wherein
the sequence of samples are transform coefficients of a transform coefficient block (10), and
the locational relationship to the current sample is determined by a template (122) positioned at the current sample (13’).

31. The apparatus of claim 30, wherein the selection of the first adaptive probability model depends on:
the sum of absolute quantization indices inside the template; and/or
the number of non-zero absolute quantization indices inside the template.

32. The apparatus of one of claims 1 to 31, wherein the apparatus is configured to:
decode an absolute value of the quantization index (56) for the current sample (13’) from the data stream using binary arithmetic decoding by:
entropy decode (85’) a second bin of the a string onto which the absolute value is binarized, which second bin specifies whether the absolute value is greater than one or not, using a second adaptive probability model, wherein the probability model is selected among a set of adaptive probability models and the selection depends on the set (48) of quantization levels selected for the current sample (13’) or the state for the current sample.

33. The apparatus of claim 32, wherein the apparatus is configured so that the selection of the second adaptive probability model further depends on a parity of an immediately preceding sample of the sequence of samples.

34. The apparatus of one of claims 23 to 33, wherein the apparatus is configured so that the selection of the second adaptive probability model depends on one or more immediately preceding samples of the sequence of samples having a predetermined locational relationship to the current sample.

35. The apparatus of claim 34, wherein
the sequence of samples are transform coefficients of a transform coefficient block (10), and
the locational relationship to the current sample is determined by a template (122) positioned at the current sample (13’).

36. The apparatus of one of claims 22 to 35, wherein the apparatus is configured to: entropy decode the bins of the binarization of the absolute values of the quantization indexes in multiple passes over the scanning positions of the block or subblocks of the block.

37. The apparatus of claim 36, wherein the apparatus is configured to: perform a probability model selection for a currently decoded bin of the absolute value of the quantization index among a set of adaptive probability models depending on already decoded bins of the binarization of quantization indices of other samples of the sequence of samples.

38. The apparatus of one of claims 1 to 37, wherein the apparatus is configured to sequentially decode a further sequence of samples (13) which describe the media signal by use of mutually sample independent scalar dequantization.

39. The apparatus of one of claims 1 to 38,
wherein the media signal is a video sequence of pictures and the samples of the sequence of samples are transform coefficients of a transform coefficient block (10) representing a transform of a predetermined transform block of transform bocks (284) the video pictures are partitioned into, and
wherein the apparatus switches between decoding the transform blocks by use of mutually sample independent scalar dequantization and mutually sample dependent scalar dequantization as applied for the predetermined transform block.

40. The apparatus of claim 39, wherein the apparatus is configured to: perform the switching between the decoding the transform blocks by use of mutually sample independent scalar dequantization and mutually sample dependent scalar dequantization depending on one or more coding parameters signaled in the data stream and varying across the picture or between pictures of the video sequence.

41. The apparatus of claim 39 or 40, wherein the apparatus is configured to: derive the switching between the decoding the transform blocks by use of mutually sample independent scalar dequantization and mutually sample dependent scalar dequantization from explicit signaling in the data stream.

42. An apparatus for encoding a media signal into a data stream, wherein the apparatus is configured to sequentially encode a sequence of samples (13) which describe the media signal by:
selecting (54), for a current sample (13’), a set (48) of reconstruction levels out of a plurality (50) of reconstruction level sets (52) depending on quantization indices (58) encoded into the data stream (14) for previous samples of the sequence of samples,
quantizing (64) the current sample (13’) onto one reconstruction level of the set (48) of reconstruction levels, and
encoding a quantization index (56) which indicates the one reconstruction level out of the selected set (48) of reconstruction levels for the current sample (13’) into the data stream (14).

43. The apparatus of claim 42, wherein the media signal comprises a picture (212).

44. The apparatus of claim 42 or 43,
wherein the media signal comprises a picture (212),
wherein the apparatus is configured to transform a picture block (284) of the picture (212) to obtain a transform coefficient block (10), and
wherein predetermined transform coefficients of the transform coefficient block, scanned along a predetermined coefficient scan (14) represent the sequence of samples.

45. The apparatus of claim 42 or 43,
wherein the media signal comprises a picture (212),
wherein the apparatus is configured to predict the picture content of the picture (212) within a picture block (284) of the picture, and
wherein the sequence of samples (13) represents picture samples of a prediction residual of the prediction of the picture content of the picture within the picture block so that a decoding the sequence of samples yields the picture samples of the prediction residual and combining the picture samples of the prediction residual with the prediction of the picture content of the picture within the picture block yields a block of reconstructed samples for the picture.

46. The apparatus of claim 42 or 43,
wherein the media signal comprises a picture (212) and the apparatus is configured to:
predict the picture content of the picture (212) within a picture block (284) of the picture,
wherein the sequence of samples (13) is formed by transform coefficients of a transform coefficient block (10), scanned along a predetermined coefficient scan (14), so that the decoding the sequence of samples yields the transform coefficient block (10), and
wherein the apparatus is configured to determine the transform coefficient block (10) so that:
subjecting (36) the transform coefficient block (10) to an inverse transformation yields a block of residual samples, and
combining the block of residual samples with the prediction of the picture content of the picture within the picture block yields a reconstruction of the picture within the picture block.

47. The apparatus of claim 46, wherein the combining involves adding the block of residual samples and the prediction of the picture content of the picture within the picture block.

48. The apparatus of claims 45 to 47, wherein the apparatus is configured to:
perform the prediction of the picture content of the picture within the picture block by
intra- prediction, or
inter prediction.

49. The apparatus of claim 48, wherein the apparatus is configured to: provide the data stream with an indication indicating whether intra- or inter prediction is used.

50. The apparatus of one of claims 42 to 49, wherein the number of reconstruction level sets (52) of the plurality (50) of reconstruction level sets (52) is two.

51. The apparatus of one of claims 42 to 50, wherein the apparatus is configured to:
parametrize the plurality (50) of reconstruction level sets (52) by way of a predetermined quantization step size, and
insert information on the predetermined quantization step size into the data stream (14).

52. The apparatus of one of claims 42 to 51,
wherein the media signal comprises a picture (212), and
wherein the apparatus is configured to:
partition the picture (212) into picture blocks (284),
insert information on a predetermined quantization step size into the data stream (14) in a manner varying among the picture blocks, and
parametrize the plurality (50) of quantization level sets (52) by way of the predetermined quantization step size.

53. The apparatus of one of claims 42 to 52,
wherein the media signal comprises a picture (212), and
wherein the apparatus is configured to:
partition the picture (212) into picture blocks (284) and
determine for each of a subset of the picture blocks (284) of the picture (212) a transform coefficient block (10), wherein the transform coefficients of a predetermined transform coefficient block, scanned along a predetermined coefficient scan, form the sequence of samples, and determine for each of a subset of the picture blocks (284) of the picture (212) the transform coefficient block (10) such that same is reconstructable by way of an inverse transformation of the transform coefficient block (10),
insert information on a predetermined quantization step size into the data stream (14) in a manner varying among subset of transform blocks, and
parametrize the plurality (50) of quantization level sets (52) by way of the predetermined quantization step size.

54. The apparatus of claim 52 or 53,
wherein the predetermined quantization step size is defined by a quantization parameter that applies to a single picture block or a group of picture blocks, and
wherein the quantization parameter is encoded by:
predicting a quantization parameter for a predetermined picture block based on quantization parameters of neighboring picture blocks;
entropy encoding a quantization parameter difference for the predetermined picture block or group of picture blocks into the data stream;
wherein adding the quantization parameter difference to the prediction of the quantization parameter yields the quantization parameter for the predetermined picture block or group of picture blocks.

55. The apparatus of claims 53 or 54,
wherein the sequence of samples are transform coefficients of a transform coefficient block, scanned along a predetermined coefficient scan (14), and
wherein the apparatus is configured to:
insert a base quantization step size for the transform coefficient block into the data stream,
insert scale information into the data stream which defines as to how the base quantization step size is scaled in order to obtain quantization step sizes for the transform coefficients of the transform coefficient block (10) so that the quantization step sizes vary across the transform coefficient locations inside the transform coefficient block (10),
wherein the plurality (50) of quantization level sets (52) are parametrized by a predetermined quantization step size obtained by scaling the base quantization step size according to the scale information.

56. The apparatus of one of claims 50 to 54,
wherein each of the plurality (50) of reconstruction level sets (52) for the current sample consists of integer multiples of a predetermined quantization step size, and
wherein the quantization step size is the same for all reconstruction level sets of the plurality (50) of reconstruction level sets (52) for the current sample.

57. The apparatus of one of claims 42 to 56,
wherein the number of reconstruction level sets (52) of the plurality (50) of reconstruction level sets (52) is two, and
wherein the plurality of quantization level sets comprises:
a first reconstruction level set that comprises zero and even multiples of a predetermined quantization step size, and
a second reconstruction level set that comprises zero and odd multiples of the predetermined quantization step size.

58. The apparatus of one of claims 42 to 57,
wherein all reconstruction levels of all reconstruction level sets represent integer multiples of a predetermined quantization step size, and
wherein the apparatus is configured to quantize each sample to:
a product of an intermediate integer value derivable depending on the selected reconstruction level set for the respective sample and the entropy encoded quantization index for the respective sample, and
the predetermined quantization step size for the respective sample.

59. The apparatus of claim 58,
wherein the number of reconstruction level sets (52) of the plurality (50) of reconstruction level sets (52) is two, and
wherein the apparatus is configured so that the intermediate value for each sample is derivable by:
if the selected reconstruction level set for the respective sample is a first set, multiply the quantization index for the respective sample by two to obtain the intermediate value for the respective sample; and
if the selected reconstruction level set for a respective sample is a second set and the quantization index for the respective sample is equal to zero, set the intermediate value for the respective sample equal to zero; and
if the selected reconstruction level set for a respective sample is a second set and the quantization index for the respective sample is greater than zero, multiply the quantization index for the respective sample by two and subtract one from the result of the multiplication to obtain the intermediate value for the respective sample; and
if the selected reconstruction level set for a current sample is a second set and the quantization index for the respective sample is less than zero, multiply the quantization index for the respective sample by two and add one to the result of the multiplication to obtain the intermediate value for the respective sample.

60. The apparatus of one of claims 42 to 59, wherein the apparatus is configured to: select (54), for the current sample (13’), the set (48) of reconstruction levels out of the plurality (50) of reconstruction level sets (52) depending on a LSB portion of the quantization indices (58) or previously coded bins of a binarization of the quantization indices (58) encoded into the data stream (14) for previous samples of the sequence of samples.

61. The apparatus of one of claims 42 to 60, wherein the apparatus is configured to: select (54), for the current sample (13’), the set (48) of reconstruction levels out of the plurality (50) of reconstruction level sets (52) depending on the results of a binary function of the quantization indices (58) encoded into the data stream (14) for previous samples of the sequence of samples.

62. The apparatus of one of claims 42 to 61, wherein the apparatus is configured to: select (54), for the current sample (13’), the set (48) of quantization levels out of the plurality (50) of quantization level sets (52) depending on a parity of the quantization indices (56) encoded into the data stream (14) for previous samples of the sequence of samples.

63. The apparatus of any of claims 42 to 62,
wherein the number of reconstruction level sets (52) of the plurality (50) of reconstruction level sets (52) is two, and
wherein the apparatus is configured to:
derive a subset index for each sample based on the selected set of reconstruction levels for the respective sample and a binary function of the quantization index for the respective sample, resulting in four possible values for the subset index; and
select (54), for the current sample (13’), the set (48) of reconstruction levels out of the plurality (50) of reconstruction level sets (52) depending on the subset indices for previous samples of the sequence of samples.

64. The apparatus of claim 63, wherein the apparatus is configured to: select (54), for the current sample (13’), the set (48) of reconstruction levels out of the plurality (50) of reconstruction level sets (52) using a selection rule which depends on the subset indices for a number of immediately preceding samples of the sequence of samples and to use the selection rule for all, or a portion, of the sequence of samples.

65. The apparatus of claim 64, wherein the number of immediately preceding samples of the sequence of samples on which the selection rule depends is two.

66. The apparatus of one of claims 63 to 65, wherein the subset index for each sample is derived based on the selected set of reconstruction levels for the sample and a parity of the quantization index for the sample.

67. The apparatus of one of claims 60 to 66,
wherein the selection rule for selecting a reconstruction level set out of a plurality of reconstruction level sets is realized via a state transition process, in such a way that:
a state associated with the current sample (13’) uniquely determines the set (48) of reconstruction levels used for the current sample (13’), and
the state for the current sample depends on the state for an immediately preceding sample of the sequence of samples and a quantization index (58) encoded into the data stream for the immediately preceding sample of the sequence of samples.

68. The apparatus of claim 67, wherein the state for the current sample depends on the state for an immediately preceding sample of the sequence of samples and a binary function of the quantization index (58) encoded into the data stream for the immediately preceding sample of the sequence of samples.

69. The apparatus of claim 68, wherein the state for the current sample depends on the state for an immediately preceding sample of the sequence of samples and a parity of the quantization index (58) encoded into the data stream for the immediately preceding sample of the sequence of samples.

70. The apparatus of one of claims 67 to 69, wherein the number of possible states is four.

71. The apparatus of claim 70,
wherein the number of reconstruction level sets (52) of the plurality (50) of reconstruction level sets (52) is two and the number of possible states is four (numbered from 0 to 3, inclusive), and
wherein the apparatus is configured to:
select a first reconstruction level set for a current sample if the state for the current sample is 0 or 1; and
select a second reconstruction level set for a current sample if the state for the current sample is 2 or 3.

72. The apparatus of claim 71, wherein the apparatus is configured to perform a state transition in the state transition process by setting:
the state for the current sample equal to 0, if the state for the preceding sample is equal to 0 and the parity of the preceding quantization index is equal to 0, or if the state for the preceding sample is equal to 1 and the parity of the preceding quantization index is equal to 1; and
the state for the current sample equal to 1, if the state for the preceding sample is equal to 2 and the parity of the preceding quantization index is equal to 0, or if the state for the preceding sample is equal to 3 and the parity of the preceding quantization index is equal to 1; and
the state for the current sample equal to 2, if the state for the preceding sample is equal to 1 and the parity of the preceding quantization index is equal to 0, or if the state for the preceding sample is equal to 0 and the parity of the preceding quantization index is equal to 1; and
the state for the current sample equal to 3, if the state for the preceding sample is equal to 3 and the parity of the preceding quantization index is equal to 0, or if the state for the preceding sample is equal to 2 and the parity of the preceding quantization index is equal to 1.

73. The apparatus of one of claims 67 to 72, wherein a state for the first - in an order of the sequence of samples - sample of the sequence of samples is set equal to a predefined value.

74. The apparatus of claim 73, wherein the state for the first sample of the sequence of samples is set equal to zero.

75. The apparatus of one of claims 42 to 74, wherein the apparatus is configured to:
encode the quantization index (56) for the current sample (13') in form of
an absolute value which is indicative of the absolute of:
a rank distance between a rank of zero and a rank of the one reconstruction level when ordering the selected set of quantization levels according to their values and, if zero is not included in the selected set of quantization levels, a rank distance between the rank of the one level and a rank of a smallest level of equal sign, when ordering the set of quantization levels according to their values, plus one, or
a rank distance between a rank of a predetermined level in the set of quantization levels which is of minimum absolute value, and a rank of the one reconstruction level when ordering the set of quantization levels according to their values, and if the absolute value is greater than zero, a sign value which is indicative of the sign of the one reconstruction level.

76. The apparatus of one of claims 42 to 74, wherein:
the number of reconstruction level sets (52) of the plurality (50) of reconstruction level sets (52) is two, and a first reconstruction level set of the plurality (50) of reconstruction level sets (52) comprises zero and even multiples of a predetermined quantization step size, and a second reconstruction level set of the plurality (50) of reconstruction level sets (52) comprises zero and odd multiples of the predetermined quantization step size, and
the apparatus configured to code the quantization index (56) for the current sample (13’) in form of an absolute value and, if the absolute value is greater than zero, a sign value which is indicative of the sign of the one reconstruction level, so that the reconstruction level is derivable from the absolute value and the sign value via a first function applied to the absolute value and the sign if the selected reconstruction level set is the first reconstruction level set and via a second function applied to the absolute value and the sign, if the selected reconstruction level set is the second reconstruction level set, with the first and second functions being symmetric with respect to the absolute value and reconstruction level.

77. The apparatus of one of claims 42 to 76, wherein the apparatus is configured to encode an absolute value of the quantization index (56) for the current sample (13’) from the data stream using a binarization of the absolute value which comprises a first bin which specifies whether the absolute value is greater than zero or not.

78. The apparatus of claim 77,
wherein the binarization of the absolute value further comprises a second bin (100) which specifies whether the absolute value is greater than one or not, and
wherein the apparatus is configured to encode the second bin into the data stream only if the first bin indicates that the absolute value is greater than zero.

79. The apparatus of one of claims 77 or 78,
wherein the binarization of the absolute value further comprises a further bin (100) which specifies a parity of the absolute value, and
wherein the apparatus is configured to encode the further bin into the data stream only if the first bin indicates that the absolute value is greater than zero

80. The apparatus of claim 78,
wherein the binarization of the absolute value further comprises a further bin (100) which specifies a parity of the absolute value, and
wherein the apparatus is configured to encode the further bin into the data stream only if the second bin is included in the data stream and the second bin indicates that the absolute value is greater than one.

81. The apparatus of one of claims 42 to 80, wherein the apparatus is configured to encode an absolute value of the quantization index (56) for the current sample (13’) into the data stream using binary arithmetic encoding by:
entropy encoding (85’) a first bin of a bin string onto which the absolute value is binarized, which first bin specifies whether the absolute value is greater than zero or not, using a first adaptive probability model, wherein the probability model is selected among a set of adaptive probability models and the selection depends on the set (48) of reconstruction levels selected for the current sample (13’).

82. The apparatus of one of claims 67 to 80, wherein the apparatus is configured to encode an absolute value of the quantization index (56) for the current sample (13’) into the data stream using binary arithmetic encoding by:
entropy encoding a first bin of a bin string onto which the absolute value is binarized, which first bin specifies whether the absolute value is greater than zero or not, using a first adaptive probability model, wherein the probability model is selected among a set of adaptive probability models and the selection depends on the state for the current sample.

83. The apparatus of claim 81 or 82, wherein the apparatus is configured so that the selection of the first adaptive probability model further depends on a parity of the quantization index for an immediately preceding sample of the sequence of samples.

84. The apparatus of one of claims 81 to 83, wherein the apparatus is configured so that the selection of the first adaptive probability model further depends on one or more preceding samples of the sequence of samples having a predetermined locational relationship to the current sample.

85. The apparatus of claim 84, wherein:
the sequence of samples are transform coefficients of a transform coefficient block (10), and
the locational relationship to the current sample is determined by a template (122) positioned at the current sample (13’).

86. The apparatus of claim 85, wherein the selection of the first adaptive probability model depends on:
the sum of absolute quantization indices inside the template; and/or
the number of non-zero absolute quantization indices inside the template.

87. The apparatus of one of claims 42 to 86, wherein the apparatus is configured to encode an absolute value of the quantization index (56) for the current sample (13’) into the data stream using binary arithmetic encoding by:
entropy encoding (85') a second bin of a string onto which the absolute value is binarized, which second bin specifies whether the absolute value is greater than one or not, using a second adaptive probability model,
wherein the probability model is selected among a set of adaptive probability models and the selection depends on the set (48) of quantization levels selected for the current sample (13’) or the state for the current sample.

88. The apparatus of claim 87, wherein the apparatus is configured so that the selection of the second adaptive probability model further depends on a parity of an immediately preceding sample of the sequence of samples.

89. The apparatus of one of claims 78 to 88, wherein the apparatus is configured so that the selection of the second adaptive probability model depends on one or more immediately preceding samples of the sequence of samples having a predetermined locational relationship to the current sample.

90. The apparatus of claim 89, wherein:
the sequence of samples are transform coefficients of a transform coefficient block (10), and
the locational relationship to the current sample is determined by a template (122) positioned at the current sample (13’).

91. The apparatus of one of claims 77 to 90, wherein the apparatus is configured to entropy encode the bins of the binarization of the absolute values of the quantization indexes in multiple passes over the scanning positions of the block or subblocks of the block.

92. The apparatus of claim 91, wherein the apparatus is configured to perform a probability model selection for a currently encoded bin of the absolute value of the quantization index among a set of adaptive probability models depending on already encoded bins of the binarization of quantization indices of other samples of the sequence of samples.

93. The apparatus of one of claims 42 to 92, wherein the apparatus is configured to sequentially encode a further sequence of samples (13) which describe the media signal by use of mutually sample independent scalar dequantization.

94. The apparatus of one of claims 42 to 93,
wherein the media signal is a video sequence of pictures and the samples of the sequence of samples are transform coefficients of a transform coefficient block (10) representing a transform of a predetermined transform block of transform bocks (284) the video pictures are partitioned into, and
wherein the apparatus switches between encoding the transform blocks by use of mutually sample independent scalar quantization and mutually sample dependent scalar quantization as applied for the predetermined transform block.

95. The apparatus of claim 94, wherein the apparatus is configured to perform the switching between the encoding the transform blocks by use of mutually sample independent scalar quantization and mutually sample dependent scalar quantization depending on one or more coding parameters signaled in the data stream and varying across the picture or between pictures of the video sequence.

96. The apparatus of claim 94 or 95, wherein the apparatus is configured to signal the switching between the encoding the transform blocks by use of mutually sample independent scalar quantization and mutually sample dependent scalar quantization by way of explicit signaling in the data stream.

97. A method for encoding a media signal into a data stream, comprising:
sequentially encoding a sequence of samples (13) which describe the media signal by:
selecting (54), for a current sample (13’), a set (48) of reconstruction levels out of a plurality (50) of reconstruction level sets (52) depending on quantization indices (58) encoded into the data stream (14) for previous samples of the sequence of samples,
quantizing (64) the current sample (13') onto one reconstruction level of the set (48) of reconstruction levels, and
encoding a quantization index (56) which indicates the one reconstruction level out of the selected set (48) of reconstruction levels for the current sample (13’) into the data stream (14).

98. An apparatus for decoding a media signal from a data stream,
wherein the apparatus is configured to sequentially decode a sequence of samples (13) which describe the media signal by:
selecting (54), for a current sample (13’), a set (48) of quantization levels out of a plurality (50) of quantization level sets (52) depending on quantization indices (58) decoded from the data stream (14) for previous samples of the sequence of samples,
decoding a quantization index (56) for the current sample (13’) from the data stream (14),
dequantizing (62) the current sample (13’) onto one level of the set (48) of quantization levels.

99. The apparatus of claim 98, wherein the media signal comprises a picture (212).

100. The apparatus of claim 98 or 99,
wherein the media signal comprises a picture (212) and the sequence of samples is formed by transform coefficients of a transform coefficient block (10), scanned along a predetermined coefficient scan (14), and
wherein the apparatus is configured to subject (36) the transform coefficient block (10) to an inverse transformation to obtain a transform block (284) of the picture (212).

101. The apparatus of claim 100,
wherein the apparatus is configured to predict the a picture content of the picture (212) within the transform block (284), and
wherein the transform block (284) represents a prediction residual of the prediction of the picture content of the picture within the transform block.

102. The apparatus of one of claims 98 to 101, wherein the apparatus is configured to select (54), for the current sample (13’), the set (48) of quantization levels out of the plurality (50) of quantization level sets (52) depending on a LSB portion of the quantization indices (58) decoded from the data stream (14) for previous samples of the sequence of samples.

103. The apparatus of claim 98 or 99, wherein the number of quantization level sets (52) of the plurality (50) of quantization level sets (52) is two.

104. The apparatus of one of claims 98 to 103, wherein the apparatus is configured to select (54), for the current sample (13’), the set (48) of quantization levels out of the plurality (50) of quantization level sets (52) depending on a parity of the quantization indices (56) decoded from the data stream (14) for previous samples of the sequence of samples.

105. The apparatus of one of claims 98 to 104, wherein the apparatus is configured to select (54), for the current sample (13’), the set (48) of quantization levels out of the plurality (50) of quantization level sets (52) using a selection rule which depends on the quantization indices (58) decoded from the data stream for a number of immediately preceding samples of the sequence of samples and to use the selection rule for all, or a portion, of the sequence of samples.

106. The apparatus of claim 105, wherein the number of immediately preceding samples of the sequence of samples on which the selection rule depends is two.

107. The apparatus of claim 105 or 106, wherein the selection rule is realized via a state transition process, in such a way that a state associated with the current sample (13’) uniquely determines the set (48) of quantization levels used for the current sample (13') and depends on the state for an immediately preceding sample of the sequence of samples and a quantization index (58) decoded from the data stream for the immediately preceding sample of the sequence of samples.

108. The apparatus of claim 107, wherein a state for the first - in an order of the sequence of samples - sample of the sequence of samples is set equal to a predefined value.

109. The apparatus of claim 105 or 107, wherein the selection rule depends on the quantization index (58) decoded from the data stream for an immediately preceding sample of the sequence of samples and a set of quantization levels selected for the immediately preceding sample of the sequence of samples.

110. The apparatus of one of claims 98 to 109, wherein the apparatus is configured to:
parametrize the plurality (50) of quantization level sets (52) by way of a predetermined quantization step size, and
derive information on the predetermined quantization step size from the data stream (14).

111. The apparatus of one of claims 98 to 110,
wherein the media signal comprises a picture (212), and
wherein the apparatus is configured to:
partition the picture (212) into transform blocks (284) and derive each of a subset of the transform blocks (284) of the picture (212) by way of an inverse transformation of a transform coefficient block (10), wherein the transform coefficients of a predetermined transform coefficient block, scanned along a predetermined coefficient scan, form the sequence of samples, and
derive information on a predetermined quantization step size from the data stream (14) in a manner varying among subset of transform blocks, and
parametrize the plurality (50) of quantization level sets (52) by way of the predetermined quantization step size.

112. The apparatus of one of claims 98 to 101, wherein each of the plurality (50) of quantization level sets (52) consists of multiples of a predetermined quantization step size which is constant for the plurality (50) of quantization level sets (52).

113. The apparatus of one of claims 98 to 102,
wherein the number of quantization level sets (52) of the plurality (50) of quantization level sets (52) is two, and
wherein the plurality of quantization level sets comprises:
a first quantization level set comprising zero and even multiples of a predetermined quantization step size, and
a second quantization level set comprising zero and odd multiples of the predetermined quantization step size.

114. The apparatus of one of claims 98 to 113, wherein the apparatus is configured to decode the quantization index (56) to the one level for the current sample (13’) in form of an absolute value which is indicative of the absolute of:
a rank distance between a rank of zero and a rank of the one level when ordering the set of quantization levels according to their values and, if zero is not included in the set of quantization levels, a rank distance between the rank of the one level and a rank of a smallest level of equal sign, when ordering the set of quantization levels according to their values, plus one, or
a rank distance between a rank of a predetermined level in the set of quantization levels which is of minimum absolute value, and a rank of the one level when ordering the set of quantization levels according to their values, and if the absolute value is greater than zero, a sign value which is indicative of the sign of the one level.

115. The apparatus of one of claims 98 to 114, wherein the apparatus is configured to decode an absolute value of the quantization index (56) to the one level for the current sample (13’) from the data stream using a binarization of the absolute value which comprises:
a first bin which specifies whether the absolute value is greater than zero or not, and

116. The apparatus of claims 115, wherein the binarization of the absolute value further comprises a second bin (100) which specifies whether the absolute value is greater than one or not, where the second bin is only included in the data stream if the first bin indicates that the absolute values is greater than zero.

117. The apparatus of claims 115 or 116, wherein the binarization of the absolute value further comprises a further bin (100) which specifies a parity of the absolute value, where the further bin is only included in the data stream if the first bin indicates that the absolute values is greater than zero.

118. The apparatus of one of claims 98 to 117, wherein the apparatus is configured to decode an absolute value of the quantization index (56) to the one level for the current sample (13’) from the data stream using binary arithmetic decoding by:
entropy decoding (85’) a first bin of a bin string onto which the absolute value is binarized, which first bin specifies whether the absolute value is greater than zero or not, using a first probability distribution estimation which depends on the set (48) of quantization values selected for the current sample (13’).

119. The apparatus of one of claims 98 to 117,
wherein the apparatus is configured to select (54), for the current sample (13’), the set (48) of quantization levels out of the plurality (50) of quantization level sets (52) using a selection rule which depends on the quantization indices (58) decoded from the data stream for a number of immediately preceding samples of the sequence of samples and to use the selection rule for all, or a portion, of the sequence of samples,
wherein the number of immediately preceding samples of the sequence of samples on which the selection rule depends is two,
wherein the selection rule is realized via a state transition process, in such a way that a state associated with the current sample (13’) uniquely determines the set (48) of quantization levels used for the current sample (13’) and depends on the state for an immediately preceding sample of the sequence of samples and a quantization index (58) decoded from the data stream for the immediately preceding sample of the sequence of samples, and
wherein the apparatus is configured to decode an absolute value of the quantization index (56) to the one level for the current sample (13’) from the data stream using binary arithmetic decoding by:
entropy decoding (85’) a first bin of a bin string onto which the absolute value is binarized, which first bin specifies whether the absolute value is greater than zero or not, using a first probability distribution estimation which depends on the state associated with the current sample (13’).

120. The apparatus of claim 119, wherein the apparatus is configured so that the first probability distribution estimation further depends on a parity of an immediately preceding sample of the sequence of samples.

121. The apparatus of claim 119 or 120, wherein the apparatus is configured so that the first probability distribution estimation further depends on one or more immediately preceding samples of the sequence of samples having a predetermined locational relationship to the current sample.

122. The apparatus of claim 121, wherein:
the sequence of samples are transform coefficients of a transform coefficient block (10), and
the locational relationship to the current sample is determined by a template (122) positioned at the current sample (13’).

123. The apparatus of one of claims 98 to 122, wherein the apparatus is configured to decode an absolute value of the quantization index (56) to the one level for the current sample (13’) from the data stream using binary arithmetic decoding by:
entropy decode (85’) a second bin of the a string onto which the absolute value is binarized, which second bin specifies whether the absolute value is greater than one or not, using a second probability distribution estimation which depends on the set (48) of quantization values selected for the current sample (13’).

124. The apparatus of claim 123, wherein the apparatus is configured so that the second probability distribution estimation further depends on a parity of an immediately preceding sample of the sequence of samples.

125. The apparatus of claim 123 or 124, wherein the apparatus is configured so that the second probability distribution estimation further depends on one or more immediately preceding samples of the sequence of samples having a predetermined locational relationship to the current sample.

126. The apparatus of claim 125, wherein the sequence of samples are transform coefficients of a transform coefficient block (10), and the locational relationship to the current sample is determined by a template (122) positioned at the current sample (13’).

127. The apparatus of one of claims 98 to 126, wherein the sequence of samples are predetermined transform coefficients of a transform coefficient block, scanned along a predetermined coefficient scan (14),
wherein the apparatus is configured to:
derive a base quantization step size for the transform coefficient block from the data stream,
derive scale information from the data stream which defines as to how the base quantization step size is scaled in order to obtain quantization step sizes for the transform coefficients of the transform coefficient block (10) so that the quantization step sizes vary across the transform coefficient block (10), and
parametrize the plurality (50) of quantization level sets (52) by a predetermined quantization step size obtained by scaling the base quantization step size according to the scale information.

128. The apparatus of one of claims 98 to 127, wherein the apparatus is configured to sequentially decode a further sequence of samples (13) which describe the media signal by use of mutually sample independent scalar dequantization.

129. The apparatus of one of claims 98 to 128,
wherein the media signal is a picture (212) and the samples of the sequence of samples are transform coefficients of a transform coefficient block (10) representing a transform of a predetermined transform block of transform bocks (284) the picture is partitioned into, and
wherein the apparatus switches between decoding the transform blocks by use of mutually sample independent scalar dequantization and mutually sample dependent scalar dequantizationa as applied for the predetermined transform block.

130. The apparatus of claim 129, wherein the apparatus is configured to perform the switching between the decoding the transform blocks by use of mutually sample independent scalar dequantization and mutually sample dependent scalar dequantization depending on one or more coding parameters signaled in the data stream and varying across the picture.

131. The apparatus of claim 127 or 130, wherein the apparatus is configured to derive the switching between the decoding the transform blocks by use of mutually sample independent scalar dequantization and mutually sample dependent scalar dequantization from explicit signaling in the data stream.

132. A method for decoding a media signal from a data stream, comprising:
sequentially decoding a sequence of samples (13) which describe the media signal by:
selecting (54), for a current sample (13’), a set (48) of quantization levels out of a plurality (50) of quantization level sets (52) depending on indices (58) decoded from the data stream (14) for previous samples of the sequence of samples,
decoding a quantization index (56) for the current sample (13’) from the data stream (14),
dequantizing (62) the current sample (13’) onto the one level of the set (48) of quantization levels.

133. An apparatus for encoding a media signal into a data stream, configured to:
sequentially encode a sequence of samples (13) which describe the media signal by:
selecting (54), for a current sample (13’), a set (48) of quantization levels out of a plurality (50) of quantization level sets (52) depending on quantization indices (58) encoded into the data stream (14) for previous samples of the sequence of samples,
quantizing (64) the current sample (13’) onto one level of the set (48) of quantization levels, and
encoding a quantization index (56) to the one level for the current sample (13’) into the data stream (14).

134. The apparatus of claim 133, wherein the media signal comprises a picture (212).

135. The apparatus of claim 133 or 134,
wherein the media signal comprises a picture (212),
wherein the apparatus is configured to transform (46) a transform block (284) of the picture (212) to obtain a transform coefficient block (10), and
wherein predetermined transform coefficients of the transform coefficient block, scanned along a predetermined coefficient scan, form the sequence of samples.

136. The apparatus of claim 135,
wherein the apparatus is configured to predict the a picture content of the picture (212) within the transform block (284), and
wherein the transform block (284) represents a prediction residual of the prediction of the picture content of the picture within the transform block.

137. The apparatus of one of claims 133 to 136, wherein the apparatus is configured to select (54), for the current sample (13’), the set (48) of quantization levels out of the plurality (50) of quantization level sets (52) depending on a LSB portion of the quantization indices (58) encoded into the data stream (14) for previous samples of the sequence of samples.

138. The apparatus of one of claims 133 to 137, wherein the number of quantization level sets (52) of the plurality (50) of quantization level sets (52) is two.

139. The apparatus of one of claims 133 to 138, wherein the apparatus is configured to select (54), for the current sample (13’), the set (48) of quantization levels out of the plurality (50) of quantization level sets (52) depending on a parity of the quantization indices (56) encoded into the data stream (14) for previous samples of the sequence of samples.

140. The apparatus of one of claims 133 to 139, wherein the apparatus is configured to select (54), for the current sample (13’), the set (48) of quantization levels out of the plurality (50) of quantization level sets (52) using a selection rule which depends on the quantization indices (58) encoded into the data stream for a number of immediately preceding samples of the sequence of samples and to use the selection rule for all, or a portion, of the sequence of samples.

141. The apparatus of claim 140, wherein the number of immediately preceding samples of the sequence of samples on which the selection rule depends is two.

142. The apparatus of claim 140 or 141, wherein the selection rule is realized via a state transition process, in such a way that a state associated with the current sample (13’) uniquely determines the set (48) of quantization levels to be used for the current sample and depends on a state for an immediately preceding sample of the sequence of samples and the quantization index (58) encoded into the data stream for the immediately preceding sample of the sequence of samples.

143. The apparatus of claim 142, wherein a state for a first - in an order of the sequence - sample of the sequence of samples is set equal to a predefined value.

144. The apparatus of one of claims 140 to 143, wherein the selection rule depends on the quantization index (58) encoded into the data stream for an immediately preceding sample of the sequence of samples and a set of quantization levels selected for the immediately preceding sample of the sequence of samples.

145. The apparatus of one of claims 133 to 144, wherein the apparatus is configured to parametrize the plurality (50) of quantization level sets (52) by way of a predetermined quantization step size and signal information on the predetermined quantization step size in the data stream (14).

146. The apparatus of one of claims 133 to 145,
wherein the media signal comprises a picture (212), and
wherein the apparatus is configured to:
partition the picture (212) into transform blocks (284) and
transform (46) each of a subset of the transform blocks (284) of the picture (212) to obtain a transform coefficient block (10), wherein predetermined transform coefficients of a predetermined transform coefficient block, scanned along a predetermined coefficient scan, form the sequence of samples, and
parametrize the plurality (50) of quantization level sets (52) by way of a predetermined quantization step size and signal information on the predetermined quantization step size in the data stream (14) in a manner varying among subset of transform blocks.

147. The apparatus of one of claims 133 to 145, wherein each of the plurality (50) of quantization level sets (52) consists of multiples of a predetermined quantization step size which is constant for the plurality (50) of quantization level sets (52).

148. The apparatus of one of claims 133 to 147,
wherein the number of quantization level sets (52) of the plurality (50) of quantization level sets (52) is two, and
the plurality of quantization level sets comprises:
a first quantization level set comprising zero and even multiples of a predetermined quantization step size, and
a second quantization level set comprising zero and odd multiples of the predetermined quantization step size.

149. The apparatus of one of claims 133 to 148, wherein the apparatus is configured to encode the quantization index (56) to the one level for the current sample (13’) as an absolute value which is indicative of the absolute of:
a rank distance between a rank of zero and a rank of the one level when ordering the set of quantization levels according to their values and, if zero is not included in the set of quantization levels, a rank distance between the rank of the one level and a rank of a smallest level of equal sign, when ordering the set of quantization levels according to their values, plus one, or
a rank distance between a rank of a predetermined level in the set of quantization levels which is of minimum absolute value, and a rank of the one level when ordering the set of quantization levels according to their values, and
if the absolute value is greater than zero, a sign value which is indicative of the sign of the one level.

150. The apparatus of one of claims 133 to 149, wherein the apparatus is configured to encode an absolute value of the quantization index (56) to the one level for the current sample (13’) using a binarization of the absolute value which comprises a first bin which specifies whether the absolute value is greater than zero or not, and

151. The apparatus of claim 150, wherein the binarization of the absolute value further comprises a second bin (100) which specifies whether the absolute values is greater than one or not, where the second bin is only included in the data stream if the first bin indicates that the absolute value is greater than zero.

152. The apparatus of claim 148 or 151, wherein the binarization of the absolute value further comprises a further bin (100) which specifies a parity of the absolute value, where the further bin is only included in the data stream if the first bin indicates that the absolute values is greater than zero.

153. The apparatus of one of claims 133 to 152, wherein the apparatus is configured to encode an absolute value of the quantization index (56) to the one level for the current sample (13’) using binary arithmetic coding by:
binarizing (80) the absolute value to obtain a bin string (82), and
entropy code (85) a first bin of the bin string, which specifies whether the absolute value is greater than zero or not, using a first probability distribution estimation which depends on the set (48) of quantization values selected for the current sample (13').

154. The apparatus of claim 153,
wherein the apparatus is configured to select (54), for the current sample (13’), the set (48) of quantization levels out of the plurality (50) of quantization level sets (52) using a selection rule which depends on the quantization indices (58) encoded into the data stream for a number of immediately preceding samples of the sequence of samples and to use the selection rule for all, or a portion, of the sequence of samples,
wherein the number of immediately preceding samples of the sequence of samples on which the selection rule depends is two,
wherein the selection rule is realized via a state transition process, in such a way that a state associated with the current sample (13’) uniquely determines the set (48) of quantization levels used for the current sample (13’) and depends on the state for an immediately preceding sample of the sequence of samples and a quantization index (58) encoded into the data stream for the immediately preceding sample of the sequence of samples,
wherein the apparatus is configured to encode an absolute value of the quantization index (56) to the one level for the current sample (13’) into the data stream using binary arithmetic encoding by
entropy encoding (85’) a first bin of a bin string onto which the absolute value is binarized, which first bin specifies whether the absolute value is greater than zero or not, using a first probability distribution estimation which depends on the state associated with the current sample (13’).

155. The apparatus of claim 154, wherein the apparatus is configured so that the first probability distribution estimation further depends on a parity of an immediately preceding sample of the sequence of samples.

156. The apparatus of claim 153 or 155, wherein the apparatus is configured so that the first probability distribution estimation further depends on one or more immediately preceding samples of the sequence of samples having a predetermined locational relationship to the current sample.

157. The apparatus of claim 156, wherein
the sequence of samples is transform coefficients of a transform coefficient block (10), and
the locational relationship to the current sample is determined by a template (122) positioned at the current sample (13’).

158. The apparatus of one of claims 133 to 157, wherein the apparatus is configured to encode an absolute value of the quantization index (56) to the one level for the current sample (13’) using binary arithmetic coding by:
binarizing (80) the absolute value to obtain a bin string (82), and
entropy code (85) a second bin of the bin string, which specifies whether the absolute value is greater than one or not, using a second probability distribution estimation which depends on the set (48) of quantization values selected for the current sample (13’).

159. The apparatus of claim 158, wherein the apparatus is configured so that the second probability distribution estimation further depends on a parity of an immediately preceding sample of the sequence of samples.

160. The apparatus of claim 158 or 159, wherein the apparatus is configured so that the second probability distribution estimation further depends on one or more immediately preceding samples of the sequence of samples having a predetermined locational relationship to the current sample.

161. The apparatus of claim 160, wherein:
the sequence of samples are transform coefficients of a transform coefficient block (10), and
the locational relationship to the current sample is determined by a template (122) positioned at the current sample (13’).

162. The apparatus of one of claims 133 to 161,
wherein the sequence of samples is predetermined transform coefficients of a transform coefficient block, scanned along a predetermined coefficient scan (14),
wherein the apparatus is configured to:
signal a base quantization step size for the transform coefficient block in the data stream,
signal scale information in the data stream which defines as to how the base quantization step size is scaled in order to obtain quantization step sizes for the transform coefficients of the transform coefficient block (10) so that the quantization step sizes vary across the transform coefficient block (10),
parametrize the plurality (50) of quantization level sets (52) by a predetermined quantization step size obtained by scaling the base quantization step size according to the scale information.

163. The apparatus of one of claims 133 to 162, wherein the apparatus is configured to sequentially encode a further sequence of samples (13) which describe the media signal by use of mutually sample independent scalar quantization.

164. The apparatus of one of claims 133 to 163,
wherein the media signal is a picture (212) and the samples of the sequence of samples are transform coefficients of a transform coefficient block (10) representing a transform of a predetermined transform block of transform bocks (284) the picture is partitioned into, and
wherein the apparatus switches between coding the transform blocks by use of mutually sample independent scalar quantization and mutually sample dependent scalar quantization as applied for the predetermined transform block.

165. The apparatus of claim 164, wherein the apparatus is configured to perform the switching between the coding the transform blocks by use of mutually sample independent scalar quantization and mutually sample dependent scalar quantization depending on one or more coding parameters signaled in the data stream and varying across the picture.

166. The apparatus of claim 164 or 165, wherein the apparatus is configured to signal the switching between the coding the transform blocks by use of mutually sample independent scalar quantization and mutually sample dependent scalar quantization by explicit signaling.

167. A method for encoding a media signal into a data stream, the method comprising:
sequentially encoding a sequence of samples (13) which describe the media signal by:
selecting (54), for a current sample (13'), a set (48) of quantization levels out of a plurality (50) of quantization level sets (52) depending on quantization indices (58) encoded into the data stream (14) for previous samples of the sequence of samples,
quantizing (64) the current sample (13’) onto one level of the set (48) of quantization levels, and
encoding a quantization index (56) to the one level for the current sample
(13’) into the data stream (14).

168. A non-transitory storage medium storing a computer program having a program code for performing, when running on a computer, the method according to claim 97, 132 or 167.

169. A data stream stored on a non-transitory storage medium generated using the method according to claim 97 or 167.

170. An apparatus for decoding a media signal from a data stream, the apparatus configured to:
sequentially decode a sequence of samples which describe the media signal by selecting, for a current sample,
a set of reconstruction levels out of a plurality of reconstruction level sets, the selecting being based on quantization indices decoded from the data stream for previous samples of the sequence of samples,
entropy decoding a quantization index for the current sample from the data stream,
wherein the quantization index indicates one reconstruction level out of the selected set of reconstruction levels for the current sample,
dequantizing the current sample onto the one reconstruction level of the selected set of reconstruction levels that is indicated by the quantization index for the current sample,
wherein the number of reconstruction level sets of the plurality of reconstruction level sets is two and the plurality of reconstruction level sets comprises,
a first reconstruction level set that comprises zero and even multiples of a predetermined quantization step size, and
a second reconstruction level set that comprises zero and odd multiples of the predetermined quantization step size.

171. A method for decoding a media signal from a data stream, the method comprising:
sequentially decoding a sequence of samples which describe the media signal by selecting, for a current sample,
a set of reconstruction levels out of a plurality of reconstruction level sets, the selecting being based on quantization indices decoded from the data stream for previous samples of the sequence of samples,
entropy decoding a quantization index for the current sample from the data stream,
wherein the quantization index indicates one reconstruction level out of the selected set of reconstruction levels for the current sample,
dequantizing the current sample onto the one reconstruction level of the selected set of reconstruction levels that is indicated by the quantization index for the current sample,
wherein the number of reconstruction level sets of the plurality of reconstruction level sets is two and the plurality of reconstruction level sets comprises,
a first reconstruction level set that comprises zero and even multiples of a predetermined quantization step size, and
a second reconstruction level set that comprises zero and odd multiples of the predetermined quantization step size.

172. A non-transitory storage medium storing a computer program having a program code for performing, when running on a computer, the method according to claim 171.

173. A data stream, stored on a non-transitory storage medium, generated using the method according to claim 171.