shithub: libvpx

Info  •  Files  •  Log  •  Branches

change HBD pixel value from uint8_t to uint16_t

This fixes a regression in 10/12 bit encoding results.

Change-Id: I438877352a41aae0a864a8d9979afe4aa2061d81

--- a/vp9/encoder/vp9_firstpass.c

+++ b/vp9/encoder/vp9_firstpass.c

@@ -643,7 +643,7 @@

   uint8_t *tmp_ptr;

   uint16_t *tmp_ptr16;

   uint8_t *kernal_ptr;

-  uint8_t dn_val;

+  uint16_t dn_val;

   uint16_t centre_val = *CONVERT_TO_SHORTPTR(src_ptr);

   kernal_ptr = fp_dn_kernal_3;