Skip clipping in the inverse wht transform C implementation
The spec requires that the final residual values of a compliant
bitstream must be representable in bitdepth+1 bits, which is not
possible if intermediate values are large enough to be clipped.
diff --git a/src/itx_tmpl.c b/src/itx_tmpl.c
index fd463cc..01bed56 100644
--- a/src/itx_tmpl.c
+++ b/src/itx_tmpl.c
@@ -161,9 +161,6 @@
coef *const coeff, const int eob
HIGHBD_DECL_SUFFIX)
{
- const int bitdepth = bitdepth_from_max(bitdepth_max);
- const int col_clip_max = (1 << (imax(bitdepth + 6, 16) - 1)) -1;
- const int col_clip_min = -col_clip_max - 1;
int32_t tmp[4 * 4], out[4], in_mem[4];
for (int i = 0; i < 4; i++) {
@@ -171,8 +168,6 @@
in_mem[j] = coeff[i + j * 4];
dav1d_inv_wht4_1d_c(in_mem, 1, &tmp[i * 4], 1, 0);
}
- for (int k = 0; k < 4 * 4; k++)
- tmp[k] = iclip(tmp[k], col_clip_min, col_clip_max);
for (int i = 0; i < 4; i++) {
dav1d_inv_wht4_1d_c(&tmp[i], 4, out, 1, 1);
