Data Flow

The data flows of the proposed algorithm as depicted in Fig. 3.3 (a)-(f) are described in the following, where the coarse-solid lines in Fig. 3.3 (a)-(f) indicate the flows according to different cases. Fig. 3.3 (a) shows one-plane type; Fig. 3.3 (b) shows two-plane type including rising, vertical, and horizontal cases; Fig. 3.3 (c) shows two-plane type, including falling cases; Fig. 3.3 (d)-(f) show the data flow in uncompression mode.

In details, Fig. 3.3 (d) illustrates the two sets of break points according to the upper-left and lower-left pixels are unsupported. Fig. 3.3 (e) and (f) show the set of differentials according to the 2^nd reference point in two-plane type including rising, vertical, horizontal, and falling cases does not pass break-point-match. Fig. 3.4 shows an example of uncomoression mode for case 2. Furthermore, assume that only 2-bit DDPCM scheme is applied in Fig. 3.4. In Fig. 3.4, the break points of differentials according to the upper-left pixel are determined as a rising case. However, the break points of differentials according to the lower-right pixel are determined as an uncompressed case. Because the two sets of break points are determined as different cases, this kind of tile finally is classified into the uncompression mode.

Compute 2^nd order Break-point map check and

two-plane mode check Break-point map check and

two-plane mode check

(a) (b)

Fig. 3.3. (a) Data flow illustration of the proposed reconfigurable depth buffer compression in one-plane type. (b) Data flow illustration of the proposed reconfigurable depth buffer compression in two-plane type for rising/vertical/horizontal cases.

Compute 2^nd order Break-point map check and

two-plane mode check Break-point map check and

two-plane mode check

(c) (d)

Fig. 3.3. (c) Data flow illustration of the proposed reconfigurable depth buffer compression in two-plane type for falling cases. (d) Data flow illustration of the proposed reconfigurable depth buffer compression in uncompression mode for case 1.

Compute 2^nd order differentials according to

upper-left pixel

Break-point generation and one-plane mode check

Break-point map check and two-plane mode check Break-point map check and

two-plane mode check

(e) (f)

Fig. 3.3. (e) Data flow illustration of the proposed reconfigurable depth buffer compression in uncompression mode for case 2. (f) Data flow illustration of the proposed reconfigurable depth buffer compression in uncompression mode for case 3.

0 1 2 3

(c) Compute 2nd order differentials based on lower-right pixel

(a) Original tile (b) Compute 2^nd order differentials based on upper-left pixel

Fig. 3.4. Example of uncompression mode for case 2.

In the first step, we compute the 1^st and 2^nd order differentials according to the 1^st reference point. In the proposed algorithm, the 1^st reference point is the upper-left pixel.

In the second step, we check the range of these 2^nd order differentials. If all differentials are inside the restricted range that the 7-bit DDPCM scheme can serve, this tile will be determined as one-plane type.

If any differential is larger than the maximum number or less than the minimum number that the 7-bit DDPCM scheme can serve, we will check the break points to determine which combination case is satisfied. If the tile does not pass the break-point-match step, this tile will be categorized into uncompression-mode case.

However, if the tile passes the break-point-check, we have to compute another set of differentials according to the 2^nd reference point. In this proposed algorithm, the 2^nd reference point in rising, vertical and horizontal case is the lower-right pixel, and in falling case the 2^nd reference point is the upper-right pixel. Additionally, if a tile is categorized into the falling case, the differentials according to the 1^st reference point at upper-left pixel have to be updated, and the updated differentials will depend on the lower-left pixel referred as the 1^st reference point in the falling case.

When we combine two sets of differentials in the two-plane type, we do the following operations for each row of the tile. In each row, we scan from the first column to the eighth column. If the break point is 0 and the combination case is not a falling case, the differentials according to the upper-left pixel will be chosen. If the combination case is a falling case, the differentials according to the lower-left pixel will be chosen. On the other hand, if the break point is 1 and the combination case is not a falling case, the differentials according to the lower-right pixel will chosen. If the combination mode is a falling case, the differentials according to the upper-right pixel will be selected. There is an exception that in each row scanning when we have scanned a break point with 1 and then all the break points of the remainder columns will be viewed as 1, no matter what the original values of these break points are. Compression scheme selection determines which compression scheme will be adopted according the range of differentials. Notice that a tile is divided into two parts and these two parts can apply different compression schemes, independently.

Finally, we will pack reference points, differentials and control-code, including compression flag, compression schemes, break points, and etc., together.