diff --git a/aom_dsp/aom_dsp.mk b/aom_dsp/aom_dsp.mk
index 227436923507b81393be6da7f9941243a499838a..f9d675ac033200b902094dcc9df811f619121555 100644
--- a/aom_dsp/aom_dsp.mk
+++ b/aom_dsp/aom_dsp.mk
@@ -290,10 +290,10 @@ DSP_SRCS-yes            += quantize.c
 DSP_SRCS-yes            += quantize.h
 
 DSP_SRCS-$(HAVE_SSE2)   += x86/quantize_sse2.c
-ifeq ($(CONFIG_HIGHBITDEPTH),yes)
+
 DSP_SRCS-$(HAVE_SSE2)   += x86/highbd_quantize_intrin_sse2.c
 DSP_SRCS-$(HAVE_AVX2)   += x86/highbd_quantize_intrin_avx2.c
-endif
+
 ifeq ($(ARCH_X86_64),yes)
 DSP_SRCS-$(HAVE_SSSE3)  += x86/quantize_ssse3_x86_64.asm
 DSP_SRCS-$(HAVE_AVX)    += x86/quantize_avx_x86_64.asm
diff --git a/aom_dsp/aom_dsp_rtcd_defs.pl b/aom_dsp/aom_dsp_rtcd_defs.pl
index 9d8028e43bff6448320aca354fc154287f0d0581..71edf88fdd6d50ed8f8ff34b8c795a18e670ea4f 100755
--- a/aom_dsp/aom_dsp_rtcd_defs.pl
+++ b/aom_dsp/aom_dsp_rtcd_defs.pl
@@ -525,13 +525,12 @@ if (aom_config("CONFIG_AOM_QM") eq "yes") {
 
     add_proto qw/void aom_quantize_b_64x64/, "const tran_low_t *coeff_ptr, intptr_t n_coeffs, int skip_block, const int16_t *zbin_ptr, const int16_t *round_ptr, const int16_t *quant_ptr, const int16_t *quant_shift_ptr, tran_low_t *qcoeff_ptr, tran_low_t *dqcoeff_ptr, const int16_t *dequant_ptr, uint16_t *eob_ptr, const int16_t *scan, const int16_t *iscan, const qm_val_t * qm_ptr, const qm_val_t * iqm_ptr";
 
-    if (aom_config("CONFIG_HIGHBITDEPTH") eq "yes") {
-      add_proto qw/void aom_highbd_quantize_b/, "const tran_low_t *coeff_ptr, intptr_t n_coeffs, int skip_block, const int16_t *zbin_ptr, const int16_t *round_ptr, const int16_t *quant_ptr, const int16_t *quant_shift_ptr, tran_low_t *qcoeff_ptr, tran_low_t *dqcoeff_ptr, const int16_t *dequant_ptr, uint16_t *eob_ptr, const int16_t *scan, const int16_t *iscan, const qm_val_t * qm_ptr, const qm_val_t * iqm_ptr";
+    add_proto qw/void aom_highbd_quantize_b/, "const tran_low_t *coeff_ptr, intptr_t n_coeffs, int skip_block, const int16_t *zbin_ptr, const int16_t *round_ptr, const int16_t *quant_ptr, const int16_t *quant_shift_ptr, tran_low_t *qcoeff_ptr, tran_low_t *dqcoeff_ptr, const int16_t *dequant_ptr, uint16_t *eob_ptr, const int16_t *scan, const int16_t *iscan, const qm_val_t * qm_ptr, const qm_val_t * iqm_ptr";
+
+    add_proto qw/void aom_highbd_quantize_b_32x32/, "const tran_low_t *coeff_ptr, intptr_t n_coeffs, int skip_block, const int16_t *zbin_ptr, const int16_t *round_ptr, const int16_t *quant_ptr, const int16_t *quant_shift_ptr, tran_low_t *qcoeff_ptr, tran_low_t *dqcoeff_ptr, const int16_t *dequant_ptr, uint16_t *eob_ptr, const int16_t *scan, const int16_t *iscan, const qm_val_t * qm_ptr, const qm_val_t * iqm_ptr";
 
-      add_proto qw/void aom_highbd_quantize_b_32x32/, "const tran_low_t *coeff_ptr, intptr_t n_coeffs, int skip_block, const int16_t *zbin_ptr, const int16_t *round_ptr, const int16_t *quant_ptr, const int16_t *quant_shift_ptr, tran_low_t *qcoeff_ptr, tran_low_t *dqcoeff_ptr, const int16_t *dequant_ptr, uint16_t *eob_ptr, const int16_t *scan, const int16_t *iscan, const qm_val_t * qm_ptr, const qm_val_t * iqm_ptr";
+    add_proto qw/void aom_highbd_quantize_b_64x64/, "const tran_low_t *coeff_ptr, intptr_t n_coeffs, int skip_block, const int16_t *zbin_ptr, const int16_t *round_ptr, const int16_t *quant_ptr, const int16_t *quant_shift_ptr, tran_low_t *qcoeff_ptr, tran_low_t *dqcoeff_ptr, const int16_t *dequant_ptr, uint16_t *eob_ptr, const int16_t *scan, const int16_t *iscan, const qm_val_t * qm_ptr, const qm_val_t * iqm_ptr";
 
-      add_proto qw/void aom_highbd_quantize_b_64x64/, "const tran_low_t *coeff_ptr, intptr_t n_coeffs, int skip_block, const int16_t *zbin_ptr, const int16_t *round_ptr, const int16_t *quant_ptr, const int16_t *quant_shift_ptr, tran_low_t *qcoeff_ptr, tran_low_t *dqcoeff_ptr, const int16_t *dequant_ptr, uint16_t *eob_ptr, const int16_t *scan, const int16_t *iscan, const qm_val_t * qm_ptr, const qm_val_t * iqm_ptr";
-    }  # CONFIG_HIGHBITDEPTH
   }  # CONFIG_AV1_ENCODER
 } else {
   if (aom_config("CONFIG_AV1_ENCODER") eq "yes") {
@@ -543,15 +542,14 @@ if (aom_config("CONFIG_AOM_QM") eq "yes") {
 
     add_proto qw/void aom_quantize_b_64x64/, "const tran_low_t *coeff_ptr, intptr_t n_coeffs, int skip_block, const int16_t *zbin_ptr, const int16_t *round_ptr, const int16_t *quant_ptr, const int16_t *quant_shift_ptr, tran_low_t *qcoeff_ptr, tran_low_t *dqcoeff_ptr, const int16_t *dequant_ptr, uint16_t *eob_ptr, const int16_t *scan, const int16_t *iscan";
 
-    if (aom_config("CONFIG_HIGHBITDEPTH") eq "yes") {
-      add_proto qw/void aom_highbd_quantize_b/, "const tran_low_t *coeff_ptr, intptr_t n_coeffs, int skip_block, const int16_t *zbin_ptr, const int16_t *round_ptr, const int16_t *quant_ptr, const int16_t *quant_shift_ptr, tran_low_t *qcoeff_ptr, tran_low_t *dqcoeff_ptr, const int16_t *dequant_ptr, uint16_t *eob_ptr, const int16_t *scan, const int16_t *iscan";
-      specialize qw/aom_highbd_quantize_b sse2 avx2/;
+    add_proto qw/void aom_highbd_quantize_b/, "const tran_low_t *coeff_ptr, intptr_t n_coeffs, int skip_block, const int16_t *zbin_ptr, const int16_t *round_ptr, const int16_t *quant_ptr, const int16_t *quant_shift_ptr, tran_low_t *qcoeff_ptr, tran_low_t *dqcoeff_ptr, const int16_t *dequant_ptr, uint16_t *eob_ptr, const int16_t *scan, const int16_t *iscan";
+    specialize qw/aom_highbd_quantize_b sse2 avx2/;
+
+    add_proto qw/void aom_highbd_quantize_b_32x32/, "const tran_low_t *coeff_ptr, intptr_t n_coeffs, int skip_block, const int16_t *zbin_ptr, const int16_t *round_ptr, const int16_t *quant_ptr, const int16_t *quant_shift_ptr, tran_low_t *qcoeff_ptr, tran_low_t *dqcoeff_ptr, const int16_t *dequant_ptr, uint16_t *eob_ptr, const int16_t *scan, const int16_t *iscan";
+    specialize qw/aom_highbd_quantize_b_32x32 sse2/;
 
-      add_proto qw/void aom_highbd_quantize_b_32x32/, "const tran_low_t *coeff_ptr, intptr_t n_coeffs, int skip_block, const int16_t *zbin_ptr, const int16_t *round_ptr, const int16_t *quant_ptr, const int16_t *quant_shift_ptr, tran_low_t *qcoeff_ptr, tran_low_t *dqcoeff_ptr, const int16_t *dequant_ptr, uint16_t *eob_ptr, const int16_t *scan, const int16_t *iscan";
-      specialize qw/aom_highbd_quantize_b_32x32 sse2/;
+    add_proto qw/void aom_highbd_quantize_b_64x64/, "const tran_low_t *coeff_ptr, intptr_t n_coeffs, int skip_block, const int16_t *zbin_ptr, const int16_t *round_ptr, const int16_t *quant_ptr, const int16_t *quant_shift_ptr, tran_low_t *qcoeff_ptr, tran_low_t *dqcoeff_ptr, const int16_t *dequant_ptr, uint16_t *eob_ptr, const int16_t *scan, const int16_t *iscan";
 
-      add_proto qw/void aom_highbd_quantize_b_64x64/, "const tran_low_t *coeff_ptr, intptr_t n_coeffs, int skip_block, const int16_t *zbin_ptr, const int16_t *round_ptr, const int16_t *quant_ptr, const int16_t *quant_shift_ptr, tran_low_t *qcoeff_ptr, tran_low_t *dqcoeff_ptr, const int16_t *dequant_ptr, uint16_t *eob_ptr, const int16_t *scan, const int16_t *iscan";
-    }  # CONFIG_HIGHBITDEPTH
   }  # CONFIG_AV1_ENCODER
 } # CONFIG_AOM_QM
 if (aom_config("CONFIG_AV1") eq "yes") {
diff --git a/aom_dsp/quantize.c b/aom_dsp/quantize.c
index d5c89ebf19bc0f275e58710d853a21ce772ffc5d..fe98b6028584066653b7103a101eee65808557f1 100644
--- a/aom_dsp/quantize.c
+++ b/aom_dsp/quantize.c
@@ -256,7 +256,6 @@ void aom_quantize_dc_64x64(const tran_low_t *coeff_ptr, int skip_block,
 }
 #endif  // CONFIG_TX64X64
 
-#if CONFIG_HIGHBITDEPTH
 void aom_highbd_quantize_dc(const tran_low_t *coeff_ptr, int n_coeffs,
                             int skip_block, const int16_t *round_ptr,
                             const int16_t quant, tran_low_t *qcoeff_ptr,
@@ -523,7 +522,6 @@ void aom_highbd_quantize_b_64x64_c(
   *eob_ptr = eob + 1;
 }
 #endif  // CONFIG_TX64X64
-#endif  // CONFIG_HIGHBITDEPTH
 
 #else  // CONFIG_AOM_QM
 
@@ -602,7 +600,6 @@ void aom_quantize_dc_64x64(const tran_low_t *coeff_ptr, int skip_block,
 }
 #endif  // CONFIG_TX64X64
 
-#if CONFIG_HIGHBITDEPTH
 void aom_highbd_quantize_dc(const tran_low_t *coeff_ptr, int n_coeffs,
                             int skip_block, const int16_t *round_ptr,
                             const int16_t quant, tran_low_t *qcoeff_ptr,
@@ -825,5 +822,4 @@ void aom_highbd_quantize_b_64x64_c(
   *eob_ptr = eob + 1;
 }
 #endif  // CONFIG_TX64X64
-#endif  // CONFIG_HIGHBITDEPTH
 #endif  // CONFIG_AOM_QM
diff --git a/aom_dsp/x86/highbd_quantize_intrin_sse2.c b/aom_dsp/x86/highbd_quantize_intrin_sse2.c
index 3ee24ab161ea1ffafa269ab66b510c00f56bd1db..5570ca5b76ce96774895a8cbc0dd1800716bbaaf 100644
--- a/aom_dsp/x86/highbd_quantize_intrin_sse2.c
+++ b/aom_dsp/x86/highbd_quantize_intrin_sse2.c
@@ -15,7 +15,6 @@
 #include "aom_mem/aom_mem.h"
 #include "aom_ports/mem.h"
 
-#if CONFIG_HIGHBITDEPTH
 void aom_highbd_quantize_b_sse2(const tran_low_t *coeff_ptr, intptr_t count,
                                 int skip_block, const int16_t *zbin_ptr,
                                 const int16_t *round_ptr,
@@ -152,4 +151,3 @@ void aom_highbd_quantize_b_32x32_sse2(
   }
   *eob_ptr = eob + 1;
 }
-#endif
diff --git a/av1/av1_cx.mk b/av1/av1_cx.mk
index 938be38b61b6043ba649c3ba4772ae7d2b05ff1e..5c8c9b8f4a2b1e4bb2db0b1cab030608c5d68946 100644
--- a/av1/av1_cx.mk
+++ b/av1/av1_cx.mk
@@ -125,10 +125,10 @@ endif
 AV1_CX_SRCS-$(HAVE_SSE2) += encoder/x86/av1_quantize_sse2.c
 AV1_CX_SRCS-$(HAVE_AVX2) += encoder/x86/av1_quantize_avx2.c
 AV1_CX_SRCS-$(HAVE_SSE2) += encoder/x86/temporal_filter_apply_sse2.asm
-ifeq ($(CONFIG_HIGHBITDEPTH),yes)
+
 AV1_CX_SRCS-$(HAVE_SSE2) += encoder/x86/highbd_block_error_intrin_sse2.c
 AV1_CX_SRCS-$(HAVE_AVX2) += encoder/x86/av1_highbd_quantize_avx2.c
-endif
+
 
 AV1_CX_SRCS-$(HAVE_SSE2) += encoder/x86/dct_sse2.asm
 AV1_CX_SRCS-$(HAVE_SSE2) += encoder/x86/error_sse2.asm
@@ -140,10 +140,10 @@ endif
 AV1_CX_SRCS-$(HAVE_SSE2) += encoder/x86/dct_intrin_sse2.c
 AV1_CX_SRCS-$(HAVE_SSSE3) += encoder/x86/dct_ssse3.c
 AV1_CX_SRCS-$(HAVE_AVX2) += encoder/x86/hybrid_fwd_txfm_avx2.c
-ifeq ($(CONFIG_HIGHBITDEPTH),yes)
+
 AV1_CX_SRCS-$(HAVE_SSE4_1) += encoder/x86/av1_highbd_quantize_sse4.c
+
 AV1_CX_SRCS-$(HAVE_SSE4_1) += encoder/x86/highbd_fwd_txfm_sse4.c
-endif
 
 ifeq ($(CONFIG_EXT_INTER),yes)
 AV1_CX_SRCS-yes += encoder/wedge_utils.c
diff --git a/av1/common/av1_rtcd_defs.pl b/av1/common/av1_rtcd_defs.pl
index b16f486383059a3ccfc867721da2af95064325e8..1248cd31b7776286a0f7f5d20948a70a0b3f2ef4 100755
--- a/av1/common/av1_rtcd_defs.pl
+++ b/av1/common/av1_rtcd_defs.pl
@@ -417,25 +417,23 @@ if (aom_config("CONFIG_DPCM_INTRA") eq "yes") {
   }
 }
 
-if (aom_config("CONFIG_HIGHBITDEPTH") eq "yes") {
-  #fwd txfm
-  add_proto qw/void av1_fwd_txfm2d_4x8/, "const int16_t *input, int32_t *output, int stride, int tx_type, int bd";
-  add_proto qw/void av1_fwd_txfm2d_8x4/, "const int16_t *input, int32_t *output, int stride, int tx_type, int bd";
-  add_proto qw/void av1_fwd_txfm2d_8x16/, "const int16_t *input, int32_t *output, int stride, int tx_type, int bd";
-  add_proto qw/void av1_fwd_txfm2d_16x8/, "const int16_t *input, int32_t *output, int stride, int tx_type, int bd";
-  add_proto qw/void av1_fwd_txfm2d_16x32/, "const int16_t *input, int32_t *output, int stride, int tx_type, int bd";
-  add_proto qw/void av1_fwd_txfm2d_32x16/, "const int16_t *input, int32_t *output, int stride, int tx_type, int bd";
-  add_proto qw/void av1_fwd_txfm2d_4x4/, "const int16_t *input, int32_t *output, int stride, int tx_type, int bd";
-  specialize qw/av1_fwd_txfm2d_4x4 sse4_1/;
-  add_proto qw/void av1_fwd_txfm2d_8x8/, "const int16_t *input, int32_t *output, int stride, int tx_type, int bd";
-  specialize qw/av1_fwd_txfm2d_8x8 sse4_1/;
-  add_proto qw/void av1_fwd_txfm2d_16x16/, "const int16_t *input, int32_t *output, int stride, int tx_type, int bd";
-  specialize qw/av1_fwd_txfm2d_16x16 sse4_1/;
-  add_proto qw/void av1_fwd_txfm2d_32x32/, "const int16_t *input, int32_t *output, int stride, int tx_type, int bd";
-  specialize qw/av1_fwd_txfm2d_32x32 sse4_1/;
-  add_proto qw/void av1_fwd_txfm2d_64x64/, "const int16_t *input, int32_t *output, int stride, int tx_type, int bd";
-  specialize qw/av1_fwd_txfm2d_64x64 sse4_1/;
-}
+#fwd txfm
+add_proto qw/void av1_fwd_txfm2d_4x8/, "const int16_t *input, int32_t *output, int stride, int tx_type, int bd";
+add_proto qw/void av1_fwd_txfm2d_8x4/, "const int16_t *input, int32_t *output, int stride, int tx_type, int bd";
+add_proto qw/void av1_fwd_txfm2d_8x16/, "const int16_t *input, int32_t *output, int stride, int tx_type, int bd";
+add_proto qw/void av1_fwd_txfm2d_16x8/, "const int16_t *input, int32_t *output, int stride, int tx_type, int bd";
+add_proto qw/void av1_fwd_txfm2d_16x32/, "const int16_t *input, int32_t *output, int stride, int tx_type, int bd";
+add_proto qw/void av1_fwd_txfm2d_32x16/, "const int16_t *input, int32_t *output, int stride, int tx_type, int bd";
+add_proto qw/void av1_fwd_txfm2d_4x4/, "const int16_t *input, int32_t *output, int stride, int tx_type, int bd";
+specialize qw/av1_fwd_txfm2d_4x4 sse4_1/;
+add_proto qw/void av1_fwd_txfm2d_8x8/, "const int16_t *input, int32_t *output, int stride, int tx_type, int bd";
+specialize qw/av1_fwd_txfm2d_8x8 sse4_1/;
+add_proto qw/void av1_fwd_txfm2d_16x16/, "const int16_t *input, int32_t *output, int stride, int tx_type, int bd";
+specialize qw/av1_fwd_txfm2d_16x16 sse4_1/;
+add_proto qw/void av1_fwd_txfm2d_32x32/, "const int16_t *input, int32_t *output, int stride, int tx_type, int bd";
+specialize qw/av1_fwd_txfm2d_32x32 sse4_1/;
+add_proto qw/void av1_fwd_txfm2d_64x64/, "const int16_t *input, int32_t *output, int stride, int tx_type, int bd";
+specialize qw/av1_fwd_txfm2d_64x64 sse4_1/;
 
 #
 # Motion search
@@ -480,33 +478,34 @@ if (aom_config("CONFIG_HIGHBITDEPTH") eq "yes") {
   add_proto qw/int64_t av1_highbd_block_error/, "const tran_low_t *coeff, const tran_low_t *dqcoeff, intptr_t block_size, int64_t *ssz, int bd";
   specialize qw/av1_highbd_block_error sse2/;
 
-  if (aom_config("CONFIG_AOM_QM") eq "yes") {
-    add_proto qw/void av1_highbd_quantize_fp/, "const tran_low_t *coeff_ptr, intptr_t n_coeffs, int skip_block, const int16_t *zbin_ptr, const int16_t *round_ptr, const int16_t *quant_ptr, const int16_t *quant_shift_ptr, tran_low_t *qcoeff_ptr, tran_low_t *dqcoeff_ptr, const int16_t *dequant_ptr, uint16_t *eob_ptr, const int16_t *scan, const int16_t *iscan, const qm_val_t * qm_ptr, const qm_val_t * iqm_ptr, int log_scale";
+  # fdct functions
+  if (aom_config("CONFIG_TX64X64") eq "yes") {
+    add_proto qw/void av1_highbd_fht64x64/, "const int16_t *input, tran_low_t *output, int stride, int tx_type";
+  }
 
-    add_proto qw/void av1_highbd_quantize_fp_32x32/, "const tran_low_t *coeff_ptr, intptr_t n_coeffs, int skip_block, const int16_t *zbin_ptr, const int16_t *round_ptr, const int16_t *quant_ptr, const int16_t *quant_shift_ptr, tran_low_t *qcoeff_ptr, tran_low_t *dqcoeff_ptr, const int16_t *dequant_ptr, uint16_t *eob_ptr, const int16_t *scan, const int16_t *iscan, const qm_val_t * qm_ptr, const qm_val_t * iqm_ptr, int log_scale";
+  add_proto qw/void av1_highbd_temporal_filter_apply/, "uint8_t *frame1, unsigned int stride, uint8_t *frame2, unsigned int block_width, unsigned int block_height, int strength, int filter_weight, unsigned int *accumulator, uint16_t *count";
 
-    if (aom_config("CONFIG_TX64X64") eq "yes") {
-      add_proto qw/void av1_highbd_quantize_fp_64x64/, "const tran_low_t *coeff_ptr, intptr_t n_coeffs, int skip_block, const int16_t *zbin_ptr, const int16_t *round_ptr, const int16_t *quant_ptr, const int16_t *quant_shift_ptr, tran_low_t *qcoeff_ptr, tran_low_t *dqcoeff_ptr, const int16_t *dequant_ptr, uint16_t *eob_ptr, const int16_t *scan, const int16_t *iscan, const qm_val_t * qm_ptr, const qm_val_t * iqm_ptr, int log_scale";
-    }
+}
 
-    add_proto qw/void av1_highbd_quantize_b/, "const tran_low_t *coeff_ptr, intptr_t n_coeffs, int skip_block, const int16_t *zbin_ptr, const int16_t *round_ptr, const int16_t *quant_ptr, const int16_t *quant_shift_ptr, tran_low_t *qcoeff_ptr, tran_low_t *dqcoeff_ptr, const int16_t *dequant_ptr, uint16_t *eob_ptr, const int16_t *scan, const int16_t *iscan, const qm_val_t * qm_ptr, const qm_val_t * iqm_ptr, int log_scale";
-  } else {
-    add_proto qw/void av1_highbd_quantize_fp/, "const tran_low_t *coeff_ptr, intptr_t n_coeffs, int skip_block, const int16_t *zbin_ptr, const int16_t *round_ptr, const int16_t *quant_ptr, const int16_t *quant_shift_ptr, tran_low_t *qcoeff_ptr, tran_low_t *dqcoeff_ptr, const int16_t *dequant_ptr, uint16_t *eob_ptr, const int16_t *scan, const int16_t *iscan, int log_scale";
-    specialize qw/av1_highbd_quantize_fp sse4_1 avx2/;
+if (aom_config("CONFIG_AOM_QM") eq "yes") {
+  add_proto qw/void av1_highbd_quantize_fp/, "const tran_low_t *coeff_ptr, intptr_t n_coeffs, int skip_block, const int16_t *zbin_ptr, const int16_t *round_ptr, const int16_t *quant_ptr, const int16_t *quant_shift_ptr, tran_low_t *qcoeff_ptr, tran_low_t *dqcoeff_ptr, const int16_t *dequant_ptr, uint16_t *eob_ptr, const int16_t *scan, const int16_t *iscan, const qm_val_t * qm_ptr, const qm_val_t * iqm_ptr, int log_scale";
 
-    add_proto qw/void av1_highbd_quantize_b/, "const tran_low_t *coeff_ptr, intptr_t n_coeffs, int skip_block, const int16_t *zbin_ptr, const int16_t *round_ptr, const int16_t *quant_ptr, const int16_t *quant_shift_ptr, tran_low_t *qcoeff_ptr, tran_low_t *dqcoeff_ptr, const int16_t *dequant_ptr, uint16_t *eob_ptr, const int16_t *scan, const int16_t *iscan, int log_scale";
-  }
+  add_proto qw/void av1_highbd_quantize_fp_32x32/, "const tran_low_t *coeff_ptr, intptr_t n_coeffs, int skip_block, const int16_t *zbin_ptr, const int16_t *round_ptr, const int16_t *quant_ptr, const int16_t *quant_shift_ptr, tran_low_t *qcoeff_ptr, tran_low_t *dqcoeff_ptr, const int16_t *dequant_ptr, uint16_t *eob_ptr, const int16_t *scan, const int16_t *iscan, const qm_val_t * qm_ptr, const qm_val_t * iqm_ptr, int log_scale";
 
-  # fdct functions
   if (aom_config("CONFIG_TX64X64") eq "yes") {
-    add_proto qw/void av1_highbd_fht64x64/, "const int16_t *input, tran_low_t *output, int stride, int tx_type";
+    add_proto qw/void av1_highbd_quantize_fp_64x64/, "const tran_low_t *coeff_ptr, intptr_t n_coeffs, int skip_block, const int16_t *zbin_ptr, const int16_t *round_ptr, const int16_t *quant_ptr, const int16_t *quant_shift_ptr, tran_low_t *qcoeff_ptr, tran_low_t *dqcoeff_ptr, const int16_t *dequant_ptr, uint16_t *eob_ptr, const int16_t *scan, const int16_t *iscan, const qm_val_t * qm_ptr, const qm_val_t * iqm_ptr, int log_scale";
   }
 
-  add_proto qw/void av1_highbd_fwht4x4/, "const int16_t *input, tran_low_t *output, int stride";
-
-  add_proto qw/void av1_highbd_temporal_filter_apply/, "uint8_t *frame1, unsigned int stride, uint8_t *frame2, unsigned int block_width, unsigned int block_height, int strength, int filter_weight, unsigned int *accumulator, uint16_t *count";
+  add_proto qw/void av1_highbd_quantize_b/, "const tran_low_t *coeff_ptr, intptr_t n_coeffs, int skip_block, const int16_t *zbin_ptr, const int16_t *round_ptr, const int16_t *quant_ptr, const int16_t *quant_shift_ptr, tran_low_t *qcoeff_ptr, tran_low_t *dqcoeff_ptr, const int16_t *dequant_ptr, uint16_t *eob_ptr, const int16_t *scan, const int16_t *iscan, const qm_val_t * qm_ptr, const qm_val_t * iqm_ptr, int log_scale";
+} else {
+  add_proto qw/void av1_highbd_quantize_fp/, "const tran_low_t *coeff_ptr, intptr_t n_coeffs, int skip_block, const int16_t *zbin_ptr, const int16_t *round_ptr, const int16_t *quant_ptr, const int16_t *quant_shift_ptr, tran_low_t *qcoeff_ptr, tran_low_t *dqcoeff_ptr, const int16_t *dequant_ptr, uint16_t *eob_ptr, const int16_t *scan, const int16_t *iscan, int log_scale";
+  specialize qw/av1_highbd_quantize_fp sse4_1 avx2/;
 
+  add_proto qw/void av1_highbd_quantize_b/, "const tran_low_t *coeff_ptr, intptr_t n_coeffs, int skip_block, const int16_t *zbin_ptr, const int16_t *round_ptr, const int16_t *quant_ptr, const int16_t *quant_shift_ptr, tran_low_t *qcoeff_ptr, tran_low_t *dqcoeff_ptr, const int16_t *dequant_ptr, uint16_t *eob_ptr, const int16_t *scan, const int16_t *iscan, int log_scale";
 }
+
+add_proto qw/void av1_highbd_fwht4x4/, "const int16_t *input, tran_low_t *output, int stride";
+
 # End av1_high encoder functions
 
 if (aom_config("CONFIG_EXT_INTER") eq "yes") {
diff --git a/av1/encoder/av1_quantize.c b/av1/encoder/av1_quantize.c
index 8c56d2133de11f1ae3bf592a49bc1166ff74e652..dd53d42232ede0bbfa5b4fe81ee31626f7481772 100644
--- a/av1/encoder/av1_quantize.c
+++ b/av1/encoder/av1_quantize.c
@@ -845,7 +845,6 @@ void av1_quantize_dc_nuq_facade(const tran_low_t *coeff_ptr, intptr_t n_coeffs,
 }
 #endif  // CONFIG_NEW_QUANT
 
-#if CONFIG_HIGHBITDEPTH
 void av1_highbd_quantize_fp_facade(const tran_low_t *coeff_ptr,
                                    intptr_t n_coeffs, const MACROBLOCK_PLANE *p,
                                    tran_low_t *qcoeff_ptr,
@@ -951,7 +950,6 @@ void av1_highbd_quantize_b_facade(const tran_low_t *coeff_ptr,
   }
 }
 
-#if CONFIG_HIGHBITDEPTH
 static INLINE void highbd_quantize_dc(
     const tran_low_t *coeff_ptr, int n_coeffs, int skip_block,
     const int16_t *round_ptr, const int16_t quant, tran_low_t *qcoeff_ptr,
@@ -980,7 +978,6 @@ static INLINE void highbd_quantize_dc(
   }
   *eob_ptr = eob + 1;
 }
-#endif  // CONFIG_HIGHBITDEPTH
 
 void av1_highbd_quantize_dc_facade(const tran_low_t *coeff_ptr,
                                    intptr_t n_coeffs, const MACROBLOCK_PLANE *p,
@@ -1519,9 +1516,7 @@ void av1_highbd_quantize_dc_nuq_facade(
   }
 }
 #endif  // CONFIG_NEW_QUANT
-#endif  // CONFIG_HIGHBITDEPTH
 
-#if CONFIG_HIGHBITDEPTH
 void av1_highbd_quantize_fp_c(const tran_low_t *coeff_ptr, intptr_t count,
                               int skip_block, const int16_t *zbin_ptr,
                               const int16_t *round_ptr,
@@ -1579,8 +1574,6 @@ void av1_highbd_quantize_fp_c(const tran_low_t *coeff_ptr, intptr_t count,
   *eob_ptr = eob + 1;
 }
 
-#endif  // CONFIG_HIGHBITDEPTH
-
 static void invert_quant(int16_t *quant, int16_t *shift, int d) {
   uint32_t t;
   int l, m;
diff --git a/av1/encoder/av1_quantize.h b/av1/encoder/av1_quantize.h
index 4bc9cccc21382dd02da3d218c88ea6302b3a174c..e5fc8b528eec40d9a82590728986da5b6e1eadec 100644
--- a/av1/encoder/av1_quantize.h
+++ b/av1/encoder/av1_quantize.h
@@ -146,7 +146,6 @@ void av1_quantize_dc_nuq_facade(const tran_low_t *coeff_ptr, intptr_t n_coeffs,
                                 const QUANT_PARAM *qparam);
 #endif  // CONFIG_NEW_QUANT
 
-#if CONFIG_HIGHBITDEPTH
 void av1_highbd_quantize_fp_facade(const tran_low_t *coeff_ptr,
                                    intptr_t n_coeffs, const MACROBLOCK_PLANE *p,
                                    tran_low_t *qcoeff_ptr,
@@ -190,7 +189,6 @@ void av1_highbd_quantize_dc_nuq_facade(
     tran_low_t *dqcoeff_ptr, uint16_t *eob_ptr, const SCAN_ORDER *sc,
     const QUANT_PARAM *qparam);
 #endif  // CONFIG_NEW_QUANT
-#endif  // CONFIG_HIGHBITDEPTH
 
 #ifdef __cplusplus
 }  // extern "C"
diff --git a/av1/encoder/dct.c b/av1/encoder/dct.c
index 8ed48dd85b3291962964c0917b9ab97c0d28a4f7..60c641f6db3f82ad74cb453f435d418bfa3a4f6c 100644
--- a/av1/encoder/dct.c
+++ b/av1/encoder/dct.c
@@ -2020,12 +2020,10 @@ void av1_fht16x16_c(const int16_t *input, tran_low_t *output, int stride,
   }
 }
 
-#if CONFIG_HIGHBITDEPTH
 void av1_highbd_fwht4x4_c(const int16_t *input, tran_low_t *output,
                           int stride) {
   av1_fwht4x4_c(input, output, stride);
 }
-#endif  // CONFIG_HIGHBITDEPTH
 
 void av1_fht32x32_c(const int16_t *input, tran_low_t *output, int stride,
                     int tx_type) {
diff --git a/av1/encoder/encodemb.c b/av1/encoder/encodemb.c
index c441421376b780bf5df3ac1c23afe65eedaf644e..4f8e0cd620fc048c398c71a986984d2a0cb8fc96 100644
--- a/av1/encoder/encodemb.c
+++ b/av1/encoder/encodemb.c
@@ -494,7 +494,6 @@ int av1_optimize_b(const AV1_COMMON *cm, MACROBLOCK *mb, int plane, int block,
 }
 
 #if !CONFIG_PVQ
-#if CONFIG_HIGHBITDEPTH
 typedef enum QUANT_FUNC {
   QUANT_FUNC_LOWBD = 0,
   QUANT_FUNC_HIGHBD = 1,
@@ -514,29 +513,12 @@ static AV1_QUANT_FACADE
 #endif  // !CONFIG_NEW_QUANT
       { NULL, NULL }
     };
+#endif  // !CONFIG_PVQ
 
-#else
-
-typedef enum QUANT_FUNC {
-  QUANT_FUNC_LOWBD = 0,
-  QUANT_FUNC_TYPES = 1
-} QUANT_FUNC;
-
-static AV1_QUANT_FACADE quant_func_list[AV1_XFORM_QUANT_TYPES]
-                                       [QUANT_FUNC_TYPES] = {
-#if !CONFIG_NEW_QUANT
-                                         { av1_quantize_fp_facade },
-                                         { av1_quantize_b_facade },
-                                         { av1_quantize_dc_facade },
-#else   // !CONFIG_NEW_QUANT
-                                         { av1_quantize_fp_nuq_facade },
-                                         { av1_quantize_b_nuq_facade },
-                                         { av1_quantize_dc_nuq_facade },
-#endif  // !CONFIG_NEW_QUANT
-                                         { NULL }
-                                       };
-#endif  // CONFIG_HIGHBITDEPTH
-#endif  // CONFIG_PVQ
+typedef void (*fwdTxfmFunc)(const int16_t *diff, tran_low_t *coeff, int stride,
+                            FWD_TXFM_PARAM *param);
+static const fwdTxfmFunc fwd_txfm_func[2] = { av1_fwd_txfm,
+                                              av1_highbd_fwd_txfm };
 
 void av1_xform_quant(const AV1_COMMON *cm, MACROBLOCK *x, int plane, int block,
                      int blk_row, int blk_col, BLOCK_SIZE plane_bsize,
@@ -668,29 +650,13 @@ void av1_xform_quant(const AV1_COMMON *cm, MACROBLOCK *x, int plane, int block,
   fwd_txfm_param.lossless = xd->lossless[mbmi->segment_id];
 
 #if !CONFIG_PVQ
-#if CONFIG_HIGHBITDEPTH
   fwd_txfm_param.bd = xd->bd;
-  if (xd->cur_buf->flags & YV12_FLAG_HIGHBITDEPTH) {
-    av1_highbd_fwd_txfm(src_diff, coeff, diff_stride, &fwd_txfm_param);
-    if (xform_quant_idx != AV1_XFORM_QUANT_SKIP_QUANT) {
-      if (LIKELY(!x->skip_block)) {
-        quant_func_list[xform_quant_idx][QUANT_FUNC_HIGHBD](
-            coeff, tx2d_size, p, qcoeff, pd, dqcoeff, eob, scan_order, &qparam);
-      } else {
-        av1_quantize_skip(tx2d_size, qcoeff, dqcoeff, eob);
-      }
-    }
-#if CONFIG_LV_MAP
-    p->txb_entropy_ctx[block] =
-        (uint8_t)av1_get_txb_entropy_context(qcoeff, scan_order, *eob);
-#endif  // CONFIG_LV_MAP
-    return;
-  }
-#endif  // CONFIG_HIGHBITDEPTH
-  av1_fwd_txfm(src_diff, coeff, diff_stride, &fwd_txfm_param);
+  const int is_hbd = get_bitdepth_data_path_index(xd);
+  fwd_txfm_func[is_hbd](src_diff, coeff, diff_stride, &fwd_txfm_param);
+
   if (xform_quant_idx != AV1_XFORM_QUANT_SKIP_QUANT) {
     if (LIKELY(!x->skip_block)) {
-      quant_func_list[xform_quant_idx][QUANT_FUNC_LOWBD](
+      quant_func_list[xform_quant_idx][is_hbd](
           coeff, tx2d_size, p, qcoeff, pd, dqcoeff, eob, scan_order, &qparam);
     } else {
       av1_quantize_skip(tx2d_size, qcoeff, dqcoeff, eob);
@@ -700,7 +666,8 @@ void av1_xform_quant(const AV1_COMMON *cm, MACROBLOCK *x, int plane, int block,
   p->txb_entropy_ctx[block] =
       (uint8_t)av1_get_txb_entropy_context(qcoeff, scan_order, *eob);
 #endif  // CONFIG_LV_MAP
-#else   // #if !CONFIG_PVQ
+  return;
+#else  // CONFIG_PVQ
   (void)xform_quant_idx;
 #if CONFIG_HIGHBITDEPTH
   fwd_txfm_param.bd = xd->bd;
diff --git a/av1/encoder/hybrid_fwd_txfm.c b/av1/encoder/hybrid_fwd_txfm.c
index a5409c2247f221ec256e1b03dbfa77ab0e368348..0ff7524d6da3594e0f7aad7f599eb2674836f2f6 100644
--- a/av1/encoder/hybrid_fwd_txfm.c
+++ b/av1/encoder/hybrid_fwd_txfm.c
@@ -203,7 +203,6 @@ static void fwd_txfm_8x32(const int16_t *src_diff, tran_low_t *coeff,
 }
 #endif  // CONFIG_EXT_TX && CONFIG_RECT_TX && CONFIG_RECT_TX_EXT
 
-#if CONFIG_HIGHBITDEPTH
 #if CONFIG_CHROMA_2X2
 static void highbd_fwd_txfm_2x2(const int16_t *src_diff, tran_low_t *coeff,
                                 int diff_stride, TX_TYPE tx_type, int lossless,
@@ -237,6 +236,7 @@ static void highbd_fwd_txfm_2x2(const int16_t *src_diff, tran_low_t *coeff,
 static void highbd_fwd_txfm_4x4(const int16_t *src_diff, tran_low_t *coeff,
                                 int diff_stride, TX_TYPE tx_type, int lossless,
                                 const int bd) {
+  int32_t *dst_coeff = (int32_t *)coeff;
   if (lossless) {
     assert(tx_type == DCT_DCT);
     av1_highbd_fwht4x4(src_diff, coeff, diff_stride);
@@ -248,7 +248,7 @@ static void highbd_fwd_txfm_4x4(const int16_t *src_diff, tran_low_t *coeff,
     case DCT_ADST:
     case ADST_ADST:
       // fallthrough intended
-      av1_fwd_txfm2d_4x4(src_diff, coeff, diff_stride, tx_type, bd);
+      av1_fwd_txfm2d_4x4(src_diff, dst_coeff, diff_stride, tx_type, bd);
       break;
 #if CONFIG_EXT_TX
     case FLIPADST_DCT:
@@ -257,7 +257,7 @@ static void highbd_fwd_txfm_4x4(const int16_t *src_diff, tran_low_t *coeff,
     case ADST_FLIPADST:
     case FLIPADST_ADST:
       // fallthrough intended
-      av1_fwd_txfm2d_4x4(src_diff, coeff, diff_stride, tx_type, bd);
+      av1_fwd_txfm2d_4x4(src_diff, dst_coeff, diff_stride, tx_type, bd);
       break;
     // use the c version for anything including identity for now
     case V_DCT:
@@ -268,7 +268,7 @@ static void highbd_fwd_txfm_4x4(const int16_t *src_diff, tran_low_t *coeff,
     case H_FLIPADST:
     case IDTX:
       // fallthrough intended
-      av1_fwd_txfm2d_4x4_c(src_diff, coeff, diff_stride, tx_type, bd);
+      av1_fwd_txfm2d_4x4_c(src_diff, dst_coeff, diff_stride, tx_type, bd);
       break;
 #endif  // CONFIG_EXT_TX
     default: assert(0);
@@ -279,47 +279,54 @@ static void highbd_fwd_txfm_4x8(const int16_t *src_diff, tran_low_t *coeff,
                                 int diff_stride, TX_TYPE tx_type,
                                 FWD_TXFM_OPT fwd_txfm_opt, const int bd) {
   (void)fwd_txfm_opt;
-  av1_fwd_txfm2d_4x8_c(src_diff, coeff, diff_stride, tx_type, bd);
+  int32_t *dst_coeff = (int32_t *)coeff;
+  av1_fwd_txfm2d_4x8_c(src_diff, dst_coeff, diff_stride, tx_type, bd);
 }
 
 static void highbd_fwd_txfm_8x4(const int16_t *src_diff, tran_low_t *coeff,
                                 int diff_stride, TX_TYPE tx_type,
                                 FWD_TXFM_OPT fwd_txfm_opt, const int bd) {
   (void)fwd_txfm_opt;
-  av1_fwd_txfm2d_8x4_c(src_diff, coeff, diff_stride, tx_type, bd);
+  int32_t *dst_coeff = (int32_t *)coeff;
+  av1_fwd_txfm2d_8x4_c(src_diff, dst_coeff, diff_stride, tx_type, bd);
 }
 
 static void highbd_fwd_txfm_8x16(const int16_t *src_diff, tran_low_t *coeff,
                                  int diff_stride, TX_TYPE tx_type,
                                  FWD_TXFM_OPT fwd_txfm_opt, const int bd) {
   (void)fwd_txfm_opt;
-  av1_fwd_txfm2d_8x16_c(src_diff, coeff, diff_stride, tx_type, bd);
+  int32_t *dst_coeff = (int32_t *)coeff;
+  av1_fwd_txfm2d_8x16_c(src_diff, dst_coeff, diff_stride, tx_type, bd);
 }
 
 static void highbd_fwd_txfm_16x8(const int16_t *src_diff, tran_low_t *coeff,
                                  int diff_stride, TX_TYPE tx_type,
                                  FWD_TXFM_OPT fwd_txfm_opt, const int bd) {
   (void)fwd_txfm_opt;
-  av1_fwd_txfm2d_16x8_c(src_diff, coeff, diff_stride, tx_type, bd);
+  int32_t *dst_coeff = (int32_t *)coeff;
+  av1_fwd_txfm2d_16x8_c(src_diff, dst_coeff, diff_stride, tx_type, bd);
 }
 
 static void highbd_fwd_txfm_16x32(const int16_t *src_diff, tran_low_t *coeff,
                                   int diff_stride, TX_TYPE tx_type,
                                   FWD_TXFM_OPT fwd_txfm_opt, const int bd) {
   (void)fwd_txfm_opt;
-  av1_fwd_txfm2d_16x32_c(src_diff, coeff, diff_stride, tx_type, bd);
+  int32_t *dst_coeff = (int32_t *)coeff;
+  av1_fwd_txfm2d_16x32_c(src_diff, dst_coeff, diff_stride, tx_type, bd);
 }
 
 static void highbd_fwd_txfm_32x16(const int16_t *src_diff, tran_low_t *coeff,
                                   int diff_stride, TX_TYPE tx_type,
                                   FWD_TXFM_OPT fwd_txfm_opt, const int bd) {
   (void)fwd_txfm_opt;
-  av1_fwd_txfm2d_32x16_c(src_diff, coeff, diff_stride, tx_type, bd);
+  int32_t *dst_coeff = (int32_t *)coeff;
+  av1_fwd_txfm2d_32x16_c(src_diff, dst_coeff, diff_stride, tx_type, bd);
 }
 
 static void highbd_fwd_txfm_8x8(const int16_t *src_diff, tran_low_t *coeff,
                                 int diff_stride, TX_TYPE tx_type,
                                 FWD_TXFM_OPT fwd_txfm_opt, const int bd) {
+  int32_t *dst_coeff = (int32_t *)coeff;
   (void)fwd_txfm_opt;
   switch (tx_type) {
     case DCT_DCT:
@@ -327,7 +334,7 @@ static void highbd_fwd_txfm_8x8(const int16_t *src_diff, tran_low_t *coeff,
     case DCT_ADST:
     case ADST_ADST:
       // fallthrough intended
-      av1_fwd_txfm2d_8x8(src_diff, coeff, diff_stride, tx_type, bd);
+      av1_fwd_txfm2d_8x8(src_diff, dst_coeff, diff_stride, tx_type, bd);
       break;
 #if CONFIG_EXT_TX
     case FLIPADST_DCT:
@@ -336,7 +343,7 @@ static void highbd_fwd_txfm_8x8(const int16_t *src_diff, tran_low_t *coeff,
     case ADST_FLIPADST:
     case FLIPADST_ADST:
       // fallthrough intended
-      av1_fwd_txfm2d_8x8(src_diff, coeff, diff_stride, tx_type, bd);
+      av1_fwd_txfm2d_8x8(src_diff, dst_coeff, diff_stride, tx_type, bd);
       break;
     // use the c version for anything including identity for now
     case V_DCT:
@@ -347,7 +354,7 @@ static void highbd_fwd_txfm_8x8(const int16_t *src_diff, tran_low_t *coeff,
     case H_FLIPADST:
     case IDTX:
       // fallthrough intended
-      av1_fwd_txfm2d_8x8_c(src_diff, coeff, diff_stride, tx_type, bd);
+      av1_fwd_txfm2d_8x8_c(src_diff, dst_coeff, diff_stride, tx_type, bd);
       break;
 #endif  // CONFIG_EXT_TX
     default: assert(0);
@@ -357,6 +364,7 @@ static void highbd_fwd_txfm_8x8(const int16_t *src_diff, tran_low_t *coeff,
 static void highbd_fwd_txfm_16x16(const int16_t *src_diff, tran_low_t *coeff,
                                   int diff_stride, TX_TYPE tx_type,
                                   FWD_TXFM_OPT fwd_txfm_opt, const int bd) {
+  int32_t *dst_coeff = (int32_t *)coeff;
   (void)fwd_txfm_opt;
   switch (tx_type) {
     case DCT_DCT:
@@ -364,7 +372,7 @@ static void highbd_fwd_txfm_16x16(const int16_t *src_diff, tran_low_t *coeff,
     case DCT_ADST:
     case ADST_ADST:
       // fallthrough intended
-      av1_fwd_txfm2d_16x16(src_diff, coeff, diff_stride, tx_type, bd);
+      av1_fwd_txfm2d_16x16(src_diff, dst_coeff, diff_stride, tx_type, bd);
       break;
 #if CONFIG_EXT_TX
     case FLIPADST_DCT:
@@ -373,7 +381,7 @@ static void highbd_fwd_txfm_16x16(const int16_t *src_diff, tran_low_t *coeff,
     case ADST_FLIPADST:
     case FLIPADST_ADST:
       // fallthrough intended
-      av1_fwd_txfm2d_16x16(src_diff, coeff, diff_stride, tx_type, bd);
+      av1_fwd_txfm2d_16x16(src_diff, dst_coeff, diff_stride, tx_type, bd);
       break;
     // use the c version for anything including identity for now
     case V_DCT:
@@ -384,7 +392,7 @@ static void highbd_fwd_txfm_16x16(const int16_t *src_diff, tran_low_t *coeff,
     case H_FLIPADST:
     case IDTX:
       // fallthrough intended
-      av1_fwd_txfm2d_16x16_c(src_diff, coeff, diff_stride, tx_type, bd);
+      av1_fwd_txfm2d_16x16_c(src_diff, dst_coeff, diff_stride, tx_type, bd);
       break;
 #endif  // CONFIG_EXT_TX
     default: assert(0);
@@ -394,6 +402,7 @@ static void highbd_fwd_txfm_16x16(const int16_t *src_diff, tran_low_t *coeff,
 static void highbd_fwd_txfm_32x32(const int16_t *src_diff, tran_low_t *coeff,
                                   int diff_stride, TX_TYPE tx_type,
                                   FWD_TXFM_OPT fwd_txfm_opt, const int bd) {
+  int32_t *dst_coeff = (int32_t *)coeff;
   (void)fwd_txfm_opt;
   switch (tx_type) {
     case DCT_DCT:
@@ -401,7 +410,7 @@ static void highbd_fwd_txfm_32x32(const int16_t *src_diff, tran_low_t *coeff,
     case DCT_ADST:
     case ADST_ADST:
       // fallthrough intended
-      av1_fwd_txfm2d_32x32(src_diff, coeff, diff_stride, tx_type, bd);
+      av1_fwd_txfm2d_32x32(src_diff, dst_coeff, diff_stride, tx_type, bd);
       break;
 #if CONFIG_EXT_TX
     case FLIPADST_DCT:
@@ -410,7 +419,7 @@ static void highbd_fwd_txfm_32x32(const int16_t *src_diff, tran_low_t *coeff,
     case ADST_FLIPADST:
     case FLIPADST_ADST:
       // fallthrough intended
-      av1_fwd_txfm2d_32x32(src_diff, coeff, diff_stride, tx_type, bd);
+      av1_fwd_txfm2d_32x32(src_diff, dst_coeff, diff_stride, tx_type, bd);
       break;
     // use the c version for anything including identity for now
     case V_DCT:
@@ -421,7 +430,7 @@ static void highbd_fwd_txfm_32x32(const int16_t *src_diff, tran_low_t *coeff,
     case H_FLIPADST:
     case IDTX:
       // fallthrough intended
-      av1_fwd_txfm2d_32x32_c(src_diff, coeff, diff_stride, tx_type, bd);
+      av1_fwd_txfm2d_32x32_c(src_diff, dst_coeff, diff_stride, tx_type, bd);
       break;
 #endif  // CONFIG_EXT_TX
     default: assert(0);
@@ -432,11 +441,12 @@ static void highbd_fwd_txfm_32x32(const int16_t *src_diff, tran_low_t *coeff,
 static void highbd_fwd_txfm_64x64(const int16_t *src_diff, tran_low_t *coeff,
                                   int diff_stride, TX_TYPE tx_type,
                                   FWD_TXFM_OPT fwd_txfm_opt, const int bd) {
+  int32_t *dst_coeff = (int32_t *)coeff;
   (void)fwd_txfm_opt;
   (void)bd;
   switch (tx_type) {
     case DCT_DCT:
-      av1_fwd_txfm2d_64x64(src_diff, coeff, diff_stride, tx_type, bd);
+      av1_fwd_txfm2d_64x64(src_diff, dst_coeff, diff_stride, tx_type, bd);
       break;
 #if CONFIG_EXT_TX
     case ADST_DCT:
@@ -459,7 +469,7 @@ static void highbd_fwd_txfm_64x64(const int16_t *src_diff, tran_low_t *coeff,
       // in a later change. This shouldn't impact performance since
       // DCT_DCT is the only extended type currently allowed for 64x64,
       // as dictated by get_ext_tx_set_type in blockd.h.
-      av1_fwd_txfm2d_64x64_c(src_diff, coeff, diff_stride, DCT_DCT, bd);
+      av1_fwd_txfm2d_64x64_c(src_diff, dst_coeff, diff_stride, DCT_DCT, bd);
       break;
     case IDTX: av1_fwd_idtx_c(src_diff, coeff, diff_stride, 64, tx_type); break;
 #endif  // CONFIG_EXT_TX
@@ -467,7 +477,6 @@ static void highbd_fwd_txfm_64x64(const int16_t *src_diff, tran_low_t *coeff,
   }
 }
 #endif  // CONFIG_TX64X64
-#endif  // CONFIG_HIGHBITDEPTH
 
 void av1_fwd_txfm(const int16_t *src_diff, tran_low_t *coeff, int diff_stride,
                   FWD_TXFM_PARAM *fwd_txfm_param) {
@@ -534,7 +543,6 @@ void av1_fwd_txfm(const int16_t *src_diff, tran_low_t *coeff, int diff_stride,
   }
 }
 
-#if CONFIG_HIGHBITDEPTH
 void av1_highbd_fwd_txfm(const int16_t *src_diff, tran_low_t *coeff,
                          int diff_stride, FWD_TXFM_PARAM *fwd_txfm_param) {
   const int fwd_txfm_opt = FWD_TXFM_OPT_NORMAL;
@@ -596,4 +604,3 @@ void av1_highbd_fwd_txfm(const int16_t *src_diff, tran_low_t *coeff,
     default: assert(0); break;
   }
 }
-#endif  // CONFIG_HIGHBITDEPTH
diff --git a/av1/encoder/hybrid_fwd_txfm.h b/av1/encoder/hybrid_fwd_txfm.h
index e6fd17275e0166d9ca6b438a87d5ec46ebff7b74..88a72b542a049c47ca6efee90a2426af4b094b82 100644
--- a/av1/encoder/hybrid_fwd_txfm.h
+++ b/av1/encoder/hybrid_fwd_txfm.h
@@ -20,9 +20,7 @@ typedef struct FWD_TXFM_PARAM {
   TX_TYPE tx_type;
   TX_SIZE tx_size;
   int lossless;
-#if CONFIG_HIGHBITDEPTH
   int bd;
-#endif  // CONFIG_HIGHBITDEPTH
 } FWD_TXFM_PARAM;
 
 #ifdef __cplusplus
@@ -32,10 +30,8 @@ extern "C" {
 void av1_fwd_txfm(const int16_t *src_diff, tran_low_t *coeff, int diff_stride,
                   FWD_TXFM_PARAM *fwd_txfm_param);
 
-#if CONFIG_HIGHBITDEPTH
 void av1_highbd_fwd_txfm(const int16_t *src_diff, tran_low_t *coeff,
                          int diff_stride, FWD_TXFM_PARAM *fwd_txfm_param);
-#endif  // CONFIG_HIGHBITDEPTH
 
 #ifdef __cplusplus
 }  // extern "C"
diff --git a/av1/encoder/x86/highbd_fwd_txfm_sse4.c b/av1/encoder/x86/highbd_fwd_txfm_sse4.c
index fb74068fe15702e5a8fa989036ecc099f7eadfac..cab36f2bdee24edeaaa1584162966e53c878f95e 100644
--- a/av1/encoder/x86/highbd_fwd_txfm_sse4.c
+++ b/av1/encoder/x86/highbd_fwd_txfm_sse4.c
@@ -113,7 +113,7 @@ static void fdct4x4_sse4_1(__m128i *in, int bit) {
   in[3] = _mm_unpackhi_epi64(v1, v3);
 }
 
-static INLINE void write_buffer_4x4(__m128i *res, tran_low_t *output) {
+static INLINE void write_buffer_4x4(__m128i *res, int32_t *output) {
   _mm_store_si128((__m128i *)(output + 0 * 4), res[0]);
   _mm_store_si128((__m128i *)(output + 1 * 4), res[1]);
   _mm_store_si128((__m128i *)(output + 2 * 4), res[2]);
@@ -404,7 +404,7 @@ static INLINE void col_txfm_8x8_rounding(__m128i *in, int shift) {
   in[15] = _mm_srai_epi32(in[15], shift);
 }
 
-static INLINE void write_buffer_8x8(const __m128i *res, tran_low_t *output) {
+static INLINE void write_buffer_8x8(const __m128i *res, int32_t *output) {
   _mm_store_si128((__m128i *)(output + 0 * 4), res[0]);
   _mm_store_si128((__m128i *)(output + 1 * 4), res[1]);
   _mm_store_si128((__m128i *)(output + 2 * 4), res[2]);
@@ -1788,7 +1788,7 @@ static void col_txfm_16x16_rounding(__m128i *in, int shift) {
   col_txfm_8x8_rounding(&in[48], shift);
 }
 
-static void write_buffer_16x16(const __m128i *in, tran_low_t *output) {
+static void write_buffer_16x16(const __m128i *in, int32_t *output) {
   const int size_8x8 = 16 * 4;
   write_buffer_8x8(&in[0], output);
   output += size_8x8;