* [FFmpeg-devel] [PATCH] avcodec/nvdec_mpeg2: fix order of quant matrix coefficients
@ 2023-05-15 10:51 Hendrik Leppkes
2023-05-15 11:30 ` Hendrik Leppkes
2023-05-15 11:39 ` [FFmpeg-devel] [PATCH 1/4] " Hendrik Leppkes
0 siblings, 2 replies; 7+ messages in thread
From: Hendrik Leppkes @ 2023-05-15 10:51 UTC (permalink / raw)
To: ffmpeg-devel; +Cc: Hendrik Leppkes
mpeg2dec stores them permutated for the IDCT, nvdec expects them in
plain raster order.
---
libavcodec/nvdec_mpeg12.c | 5 +++--
1 file changed, 3 insertions(+), 2 deletions(-)
diff --git a/libavcodec/nvdec_mpeg12.c b/libavcodec/nvdec_mpeg12.c
index e10735587d..3b9ff60734 100644
--- a/libavcodec/nvdec_mpeg12.c
+++ b/libavcodec/nvdec_mpeg12.c
@@ -83,8 +83,9 @@ static int nvdec_mpeg12_start_frame(AVCodecContext *avctx, const uint8_t *buffer
};
for (i = 0; i < 64; ++i) {
- ppc->QuantMatrixIntra[i] = s->intra_matrix[i];
- ppc->QuantMatrixInter[i] = s->inter_matrix[i];
+ int n = s->idsp.idct_permutation[i];
+ ppc->QuantMatrixIntra[i] = s->intra_matrix[n];
+ ppc->QuantMatrixInter[i] = s->inter_matrix[n];
}
return 0;
--
2.40.1.windows.1
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [FFmpeg-devel] [PATCH] avcodec/nvdec_mpeg2: fix order of quant matrix coefficients
2023-05-15 10:51 [FFmpeg-devel] [PATCH] avcodec/nvdec_mpeg2: fix order of quant matrix coefficients Hendrik Leppkes
@ 2023-05-15 11:30 ` Hendrik Leppkes
2023-05-15 12:01 ` Hendrik Leppkes
2023-05-15 11:39 ` [FFmpeg-devel] [PATCH 1/4] " Hendrik Leppkes
1 sibling, 1 reply; 7+ messages in thread
From: Hendrik Leppkes @ 2023-05-15 11:30 UTC (permalink / raw)
To: ffmpeg-devel
On Mon, May 15, 2023 at 12:51 PM Hendrik Leppkes <h.leppkes@gmail.com> wrote:
>
> mpeg2dec stores them permutated for the IDCT, nvdec expects them in
> plain raster order.
> ---
> libavcodec/nvdec_mpeg12.c | 5 +++--
> 1 file changed, 3 insertions(+), 2 deletions(-)
>
> diff --git a/libavcodec/nvdec_mpeg12.c b/libavcodec/nvdec_mpeg12.c
> index e10735587d..3b9ff60734 100644
> --- a/libavcodec/nvdec_mpeg12.c
> +++ b/libavcodec/nvdec_mpeg12.c
> @@ -83,8 +83,9 @@ static int nvdec_mpeg12_start_frame(AVCodecContext *avctx, const uint8_t *buffer
> };
>
> for (i = 0; i < 64; ++i) {
> - ppc->QuantMatrixIntra[i] = s->intra_matrix[i];
> - ppc->QuantMatrixInter[i] = s->inter_matrix[i];
> + int n = s->idsp.idct_permutation[i];
> + ppc->QuantMatrixIntra[i] = s->intra_matrix[n];
> + ppc->QuantMatrixInter[i] = s->inter_matrix[n];
> }
>
> return 0;
> --
> 2.40.1.windows.1
>
Apparently mpeg4 and VDPAU have the same issue. I can test and fix
mpeg4, but do not have VDPAU setup, so .. untested commits incoming?
- Hendrik
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".
^ permalink raw reply [flat|nested] 7+ messages in thread
* [FFmpeg-devel] [PATCH 1/4] avcodec/nvdec_mpeg2: fix order of quant matrix coefficients
2023-05-15 10:51 [FFmpeg-devel] [PATCH] avcodec/nvdec_mpeg2: fix order of quant matrix coefficients Hendrik Leppkes
2023-05-15 11:30 ` Hendrik Leppkes
@ 2023-05-15 11:39 ` Hendrik Leppkes
2023-05-15 11:39 ` [FFmpeg-devel] [PATCH 2/4] avcodec/nvdec_mpeg4: " Hendrik Leppkes
` (2 more replies)
1 sibling, 3 replies; 7+ messages in thread
From: Hendrik Leppkes @ 2023-05-15 11:39 UTC (permalink / raw)
To: ffmpeg-devel; +Cc: Hendrik Leppkes
The matrix coefficients are stored permutated for the IDCT,
rather then in plain raster order, and need to be un-permutated
for the hardware.
---
libavcodec/nvdec_mpeg12.c | 5 +++--
1 file changed, 3 insertions(+), 2 deletions(-)
diff --git a/libavcodec/nvdec_mpeg12.c b/libavcodec/nvdec_mpeg12.c
index e10735587d..3b9ff60734 100644
--- a/libavcodec/nvdec_mpeg12.c
+++ b/libavcodec/nvdec_mpeg12.c
@@ -83,8 +83,9 @@ static int nvdec_mpeg12_start_frame(AVCodecContext *avctx, const uint8_t *buffer
};
for (i = 0; i < 64; ++i) {
- ppc->QuantMatrixIntra[i] = s->intra_matrix[i];
- ppc->QuantMatrixInter[i] = s->inter_matrix[i];
+ int n = s->idsp.idct_permutation[i];
+ ppc->QuantMatrixIntra[i] = s->intra_matrix[n];
+ ppc->QuantMatrixInter[i] = s->inter_matrix[n];
}
return 0;
--
2.40.1.windows.1
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".
^ permalink raw reply [flat|nested] 7+ messages in thread
* [FFmpeg-devel] [PATCH 2/4] avcodec/nvdec_mpeg4: fix order of quant matrix coefficients
2023-05-15 11:39 ` [FFmpeg-devel] [PATCH 1/4] " Hendrik Leppkes
@ 2023-05-15 11:39 ` Hendrik Leppkes
2023-05-15 11:39 ` [FFmpeg-devel] [PATCH 3/4] avcodec/vdpau_mpeg12: " Hendrik Leppkes
2023-05-15 11:39 ` [FFmpeg-devel] [PATCH 4/4] avcodec/vdpau_mpeg4: " Hendrik Leppkes
2 siblings, 0 replies; 7+ messages in thread
From: Hendrik Leppkes @ 2023-05-15 11:39 UTC (permalink / raw)
To: ffmpeg-devel; +Cc: Hendrik Leppkes
The matrix coefficients are stored permutated for the IDCT,
rather then in plain raster order, and need to be un-permutated
for the hardware.
---
libavcodec/nvdec_mpeg4.c | 5 +++--
1 file changed, 3 insertions(+), 2 deletions(-)
diff --git a/libavcodec/nvdec_mpeg4.c b/libavcodec/nvdec_mpeg4.c
index eac138cc38..c193f6b6e4 100644
--- a/libavcodec/nvdec_mpeg4.c
+++ b/libavcodec/nvdec_mpeg4.c
@@ -88,8 +88,9 @@ static int nvdec_mpeg4_start_frame(AVCodecContext *avctx, const uint8_t *buffer,
};
for (i = 0; i < 64; ++i) {
- ppc->QuantMatrixIntra[i] = s->intra_matrix[i];
- ppc->QuantMatrixInter[i] = s->inter_matrix[i];
+ int n = s->idsp.idct_permutation[i];
+ ppc->QuantMatrixIntra[i] = s->intra_matrix[n];
+ ppc->QuantMatrixInter[i] = s->inter_matrix[n];
}
// We need to pass the full frame buffer and not just the slice
--
2.40.1.windows.1
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".
^ permalink raw reply [flat|nested] 7+ messages in thread
* [FFmpeg-devel] [PATCH 3/4] avcodec/vdpau_mpeg12: fix order of quant matrix coefficients
2023-05-15 11:39 ` [FFmpeg-devel] [PATCH 1/4] " Hendrik Leppkes
2023-05-15 11:39 ` [FFmpeg-devel] [PATCH 2/4] avcodec/nvdec_mpeg4: " Hendrik Leppkes
@ 2023-05-15 11:39 ` Hendrik Leppkes
2023-05-15 11:39 ` [FFmpeg-devel] [PATCH 4/4] avcodec/vdpau_mpeg4: " Hendrik Leppkes
2 siblings, 0 replies; 7+ messages in thread
From: Hendrik Leppkes @ 2023-05-15 11:39 UTC (permalink / raw)
To: ffmpeg-devel; +Cc: Hendrik Leppkes
The matrix coefficients are stored permutated for the IDCT,
rather then in plain raster order, and need to be un-permutated
for the hardware.
---
libavcodec/vdpau_mpeg12.c | 5 +++--
1 file changed, 3 insertions(+), 2 deletions(-)
diff --git a/libavcodec/vdpau_mpeg12.c b/libavcodec/vdpau_mpeg12.c
index 354239cad5..79007aa1a8 100644
--- a/libavcodec/vdpau_mpeg12.c
+++ b/libavcodec/vdpau_mpeg12.c
@@ -75,8 +75,9 @@ static int vdpau_mpeg_start_frame(AVCodecContext *avctx,
info->f_code[1][0] = s->mpeg_f_code[1][0];
info->f_code[1][1] = s->mpeg_f_code[1][1];
for (i = 0; i < 64; ++i) {
- info->intra_quantizer_matrix[i] = s->intra_matrix[i];
- info->non_intra_quantizer_matrix[i] = s->inter_matrix[i];
+ int n = s->idsp.idct_permutation[i];
+ info->intra_quantizer_matrix[i] = s->intra_matrix[n];
+ info->non_intra_quantizer_matrix[i] = s->inter_matrix[n];
}
return ff_vdpau_common_start_frame(pic_ctx, buffer, size);
--
2.40.1.windows.1
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".
^ permalink raw reply [flat|nested] 7+ messages in thread
* [FFmpeg-devel] [PATCH 4/4] avcodec/vdpau_mpeg4: fix order of quant matrix coefficients
2023-05-15 11:39 ` [FFmpeg-devel] [PATCH 1/4] " Hendrik Leppkes
2023-05-15 11:39 ` [FFmpeg-devel] [PATCH 2/4] avcodec/nvdec_mpeg4: " Hendrik Leppkes
2023-05-15 11:39 ` [FFmpeg-devel] [PATCH 3/4] avcodec/vdpau_mpeg12: " Hendrik Leppkes
@ 2023-05-15 11:39 ` Hendrik Leppkes
2 siblings, 0 replies; 7+ messages in thread
From: Hendrik Leppkes @ 2023-05-15 11:39 UTC (permalink / raw)
To: ffmpeg-devel; +Cc: Hendrik Leppkes
The matrix coefficients are stored permutated for the IDCT,
rather then in plain raster order, and need to be un-permutated
for the hardware.
---
libavcodec/vdpau_mpeg4.c | 5 +++--
1 file changed, 3 insertions(+), 2 deletions(-)
diff --git a/libavcodec/vdpau_mpeg4.c b/libavcodec/vdpau_mpeg4.c
index 6e082eefc6..1211b1df2c 100644
--- a/libavcodec/vdpau_mpeg4.c
+++ b/libavcodec/vdpau_mpeg4.c
@@ -74,8 +74,9 @@ static int vdpau_mpeg4_start_frame(AVCodecContext *avctx,
info->alternate_vertical_scan_flag = s->alternate_scan;
info->top_field_first = s->top_field_first;
for (i = 0; i < 64; ++i) {
- info->intra_quantizer_matrix[i] = s->intra_matrix[i];
- info->non_intra_quantizer_matrix[i] = s->inter_matrix[i];
+ int n = s->idsp.idct_permutation[i];
+ info->intra_quantizer_matrix[i] = s->intra_matrix[n];
+ info->non_intra_quantizer_matrix[i] = s->inter_matrix[n];
}
ff_vdpau_common_start_frame(pic_ctx, buffer, size);
--
2.40.1.windows.1
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [FFmpeg-devel] [PATCH] avcodec/nvdec_mpeg2: fix order of quant matrix coefficients
2023-05-15 11:30 ` Hendrik Leppkes
@ 2023-05-15 12:01 ` Hendrik Leppkes
0 siblings, 0 replies; 7+ messages in thread
From: Hendrik Leppkes @ 2023-05-15 12:01 UTC (permalink / raw)
To: ffmpeg-devel
On Mon, May 15, 2023 at 1:30 PM Hendrik Leppkes <h.leppkes@gmail.com> wrote:
>
> Apparently mpeg4 and VDPAU have the same issue. I can test and fix
> mpeg4, but do not have VDPAU setup, so .. untested commits incoming?
>
Tested and confirmed mpeg2 and mpeg4 are fixed for NVDEC, and elenril
tested VDPAU with the same result.
OK'ed by BtbN on IRC.
Going to push later today.
- Hendrik
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".
^ permalink raw reply [flat|nested] 7+ messages in thread
end of thread, other threads:[~2023-05-15 12:02 UTC | newest]
Thread overview: 7+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2023-05-15 10:51 [FFmpeg-devel] [PATCH] avcodec/nvdec_mpeg2: fix order of quant matrix coefficients Hendrik Leppkes
2023-05-15 11:30 ` Hendrik Leppkes
2023-05-15 12:01 ` Hendrik Leppkes
2023-05-15 11:39 ` [FFmpeg-devel] [PATCH 1/4] " Hendrik Leppkes
2023-05-15 11:39 ` [FFmpeg-devel] [PATCH 2/4] avcodec/nvdec_mpeg4: " Hendrik Leppkes
2023-05-15 11:39 ` [FFmpeg-devel] [PATCH 3/4] avcodec/vdpau_mpeg12: " Hendrik Leppkes
2023-05-15 11:39 ` [FFmpeg-devel] [PATCH 4/4] avcodec/vdpau_mpeg4: " Hendrik Leppkes
Git Inbox Mirror of the ffmpeg-devel mailing list - see https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
This inbox may be cloned and mirrored by anyone:
git clone --mirror https://master.gitmailbox.com/ffmpegdev/0 ffmpegdev/git/0.git
# If you have public-inbox 1.1+ installed, you may
# initialize and index your mirror using the following commands:
public-inbox-init -V2 ffmpegdev ffmpegdev/ https://master.gitmailbox.com/ffmpegdev \
ffmpegdev@gitmailbox.com
public-inbox-index ffmpegdev
Example config snippet for mirrors.
AGPL code for this site: git clone https://public-inbox.org/public-inbox.git