[FFmpeg-devel] [PATCH] avcodec/mediacodecenc: set quality in cq mode

Zhao Zhili quinkblack at foxmail.com
Tue Dec 19 16:16:02 EET 2023


On 2023/12/19 18:42, Matthieu Bouron wrote:

> On Sun, Dec 17, 2023 at 1:51 PM Zhao Zhili <quinkblack at foxmail.com> wrote:
>> From: Zhao Zhili <zhilizhao at tencent.com>
>>
>>  From AOSP doc, these values are device and codec specific, but lower
>> values generally result in more efficient (smaller-sized) encoding.
>>
>> For example, global_quality 50 on Pixel 6 results a 1080P 30 FPS
>> HEVC with 3744 kb/s, while global_quality 80 results 28178 kb/s.
>>
>> Fix #10689
>>
>> Signed-off-by: Zhao Zhili <zhilizhao at tencent.com>
>> ---
>>   libavcodec/mediacodecenc.c | 5 ++++-
>>   1 file changed, 4 insertions(+), 1 deletion(-)
>>
>> diff --git a/libavcodec/mediacodecenc.c b/libavcodec/mediacodecenc.c
>> index a1ae5d5ec5..7eae542248 100644
>> --- a/libavcodec/mediacodecenc.c
>> +++ b/libavcodec/mediacodecenc.c
>> @@ -268,8 +268,11 @@ static av_cold int mediacodec_init(AVCodecContext *avctx)
>>
>>       if (avctx->bit_rate)
>>           ff_AMediaFormat_setInt32(format, "bitrate", avctx->bit_rate);
>> -    if (s->bitrate_mode >= 0)
>> +    if (s->bitrate_mode >= 0) {
>>           ff_AMediaFormat_setInt32(format, "bitrate-mode", s->bitrate_mode);
>> +        if (s->bitrate_mode == BITRATE_MODE_CQ && avctx->global_quality > 0)
>> +            ff_AMediaFormat_setInt32(format, "quality", avctx->global_quality);
>> +    }
>>       // frame-rate and i-frame-interval are required to configure codec
>>       if (avctx->framerate.num >= avctx->framerate.den && avctx->framerate.den > 0) {
>>           s->fps = avctx->framerate.num / avctx->framerate.den;
>> --
>> 2.34.1
>>
> Does the `global_quality` value need to be checked/clamped against
> `MediaCodecInfo.EncoderCapabilities.getQualityRange()` before use ?

Thanks for the review.

Firstly, since 'quality' is device and codec specific, not only valid 
range but also the

relationship between 'quality' value and the real encoding quality (and 
bitrate), some

prerequisite knowledge(or trail and error) is required. With 
prerequisite knowledge,

check valid range is unnecessary.

Secondly, from my test the device can do clamp automatically. For 
example, on pixel 6,

set H.265 quality to 100 or 1000 get the same results (I think it's 
lossless with very high

bitrate, which should never be used except in rare case).

Last but not the least, there is no NDK API for MediaCodecInfo. I want 
to avoid complex

JNI calls unless there are a lot of gains.

> Otherwise, LGTM.
> _______________________________________________
> ffmpeg-devel mailing list
> ffmpeg-devel at ffmpeg.org
> https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
>
> To unsubscribe, visit link above, or email
> ffmpeg-devel-request at ffmpeg.org with subject "unsubscribe".



More information about the ffmpeg-devel mailing list