| Message ID | 20201013091017.1899676-6-andreas.rheinhardt@gmail.com |
|---|---|
| State | Accepted |
| Commit | 746ab8327c0c8ae1b7e894f3c21d7b7a44c0a0b3 |
| Headers | show |
| Series | [FFmpeg-devel,1/9] avcodec/asvdec: Reduce the size of some VLCs | expand |
| Context | Check | Description |
|---|---|---|
| andriy/x86_make | success | Make finished |
| andriy/x86_make_fate | success | Make fate finished |
On Tue, Oct 13, 2020 at 11:10:14AM +0200, Andreas Rheinhardt wrote: > The ASUS V2 format is designed for a little-endian bitstream reader, yet > our encoder used an ordinary big-endian bitstream writer to write it; > the bits of every byte were swapped at the end and some data (namely the > numbers not in static tables) had to be bitreversed before writing it at > all, so that it would be reversed twice. > > This commit stops doing so; instead, a little-endian bitstream writer is > used. This also necessitated to switch certain static tables, which > required trivial modifications to the decoder (that uses the same > tables). [...] > -const uint8_t ff_asv2_level_tab[63][2] = { > - { 0x3F, 10 }, { 0x2F, 10 }, { 0x37, 10 }, { 0x27, 10 }, { 0x3B, 10 }, { 0x2B, 10 }, { 0x33, 10 }, { 0x23, 10 }, > - { 0x3D, 10 }, { 0x2D, 10 }, { 0x35, 10 }, { 0x25, 10 }, { 0x39, 10 }, { 0x29, 10 }, { 0x31, 10 }, { 0x21, 10 }, > - { 0x1F, 8 }, { 0x17, 8 }, { 0x1B, 8 }, { 0x13, 8 }, { 0x1D, 8 }, { 0x15, 8 }, { 0x19, 8 }, { 0x11, 8 }, > - { 0x0F, 6 }, { 0x0B, 6 }, { 0x0D, 6 }, { 0x09, 6 }, > - { 0x07, 4 }, { 0x05, 4 }, > - { 0x03, 2 }, > - { 0x00, 5 }, > - { 0x02, 2 }, > - { 0x04, 4 }, { 0x06, 4 }, > - { 0x08, 6 }, { 0x0C, 6 }, { 0x0A, 6 }, { 0x0E, 6 }, > - { 0x10, 8 }, { 0x18, 8 }, { 0x14, 8 }, { 0x1C, 8 }, { 0x12, 8 }, { 0x1A, 8 }, { 0x16, 8 }, { 0x1E, 8 }, > - { 0x20, 10 }, { 0x30, 10 }, { 0x28, 10 }, { 0x38, 10 }, { 0x24, 10 }, { 0x34, 10 }, { 0x2C, 10 }, { 0x3C, 10 }, > - { 0x22, 10 }, { 0x32, 10 }, { 0x2A, 10 }, { 0x3A, 10 }, { 0x26, 10 }, { 0x36, 10 }, { 0x2E, 10 }, { 0x3E, 10 }, > +const uint16_t ff_asv2_level_tab[63][2] = { > + { 0x3F0, 10 }, { 0x3D0, 10 }, { 0x3B0, 10 }, { 0x390, 10 }, { 0x370, 10 }, > + { 0x350, 10 }, { 0x330, 10 }, { 0x310, 10 }, { 0x2F0, 10 }, { 0x2D0, 10 }, > + { 0x2B0, 10 }, { 0x290, 10 }, { 0x270, 10 }, { 0x250, 10 }, { 0x230, 10 }, > + { 0x210, 10 }, > + { 0x0F8, 8 }, { 0x0E8, 8 }, { 0x0D8, 8 }, { 0x0C8, 8 }, { 0x0B8, 8 }, > + { 0x0A8, 8 }, { 0x098, 8 }, { 0x088, 8 }, > + { 0x03C, 6 }, { 0x034, 6 }, { 0x02C, 6 }, { 0x024, 6 }, > + { 0x00E, 4 }, { 0x00A, 4 }, > + { 0x003, 2 }, > + { 0x000, 5 }, > + { 0x001, 2 }, > + { 0x002, 4 }, { 0x006, 4 }, > + { 0x004, 6 }, { 0x00C, 6 }, { 0x014, 6 }, { 0x01C, 6 }, > + { 0x008, 8 }, { 0x018, 8 }, { 0x028, 8 }, { 0x038, 8 }, { 0x048, 8 }, > + { 0x058, 8 }, { 0x068, 8 }, { 0x078, 8 }, > + { 0x010, 10 }, { 0x030, 10 }, { 0x050, 10 }, { 0x070, 10 }, { 0x090, 10 }, > + { 0x0B0, 10 }, { 0x0D0, 10 }, { 0x0F0, 10 }, { 0x110, 10 }, { 0x130, 10 }, > + { 0x150, 10 }, { 0x170, 10 }, { 0x190, 10 }, { 0x1B0, 10 }, { 0x1D0, 10 }, > + { 0x1F0, 10 } > }; would it be worth it to handle codes with length > the 8*sizeof in such a way that this fits still in uint8 ? [...]
Michael Niedermayer: > On Tue, Oct 13, 2020 at 11:10:14AM +0200, Andreas Rheinhardt wrote: >> The ASUS V2 format is designed for a little-endian bitstream reader, yet >> our encoder used an ordinary big-endian bitstream writer to write it; >> the bits of every byte were swapped at the end and some data (namely the >> numbers not in static tables) had to be bitreversed before writing it at >> all, so that it would be reversed twice. >> >> This commit stops doing so; instead, a little-endian bitstream writer is >> used. This also necessitated to switch certain static tables, which >> required trivial modifications to the decoder (that uses the same >> tables). > [...] > >> -const uint8_t ff_asv2_level_tab[63][2] = { >> - { 0x3F, 10 }, { 0x2F, 10 }, { 0x37, 10 }, { 0x27, 10 }, { 0x3B, 10 }, { 0x2B, 10 }, { 0x33, 10 }, { 0x23, 10 }, >> - { 0x3D, 10 }, { 0x2D, 10 }, { 0x35, 10 }, { 0x25, 10 }, { 0x39, 10 }, { 0x29, 10 }, { 0x31, 10 }, { 0x21, 10 }, >> - { 0x1F, 8 }, { 0x17, 8 }, { 0x1B, 8 }, { 0x13, 8 }, { 0x1D, 8 }, { 0x15, 8 }, { 0x19, 8 }, { 0x11, 8 }, >> - { 0x0F, 6 }, { 0x0B, 6 }, { 0x0D, 6 }, { 0x09, 6 }, >> - { 0x07, 4 }, { 0x05, 4 }, >> - { 0x03, 2 }, >> - { 0x00, 5 }, >> - { 0x02, 2 }, >> - { 0x04, 4 }, { 0x06, 4 }, >> - { 0x08, 6 }, { 0x0C, 6 }, { 0x0A, 6 }, { 0x0E, 6 }, >> - { 0x10, 8 }, { 0x18, 8 }, { 0x14, 8 }, { 0x1C, 8 }, { 0x12, 8 }, { 0x1A, 8 }, { 0x16, 8 }, { 0x1E, 8 }, >> - { 0x20, 10 }, { 0x30, 10 }, { 0x28, 10 }, { 0x38, 10 }, { 0x24, 10 }, { 0x34, 10 }, { 0x2C, 10 }, { 0x3C, 10 }, >> - { 0x22, 10 }, { 0x32, 10 }, { 0x2A, 10 }, { 0x3A, 10 }, { 0x26, 10 }, { 0x36, 10 }, { 0x2E, 10 }, { 0x3E, 10 }, >> +const uint16_t ff_asv2_level_tab[63][2] = { >> + { 0x3F0, 10 }, { 0x3D0, 10 }, { 0x3B0, 10 }, { 0x390, 10 }, { 0x370, 10 }, >> + { 0x350, 10 }, { 0x330, 10 }, { 0x310, 10 }, { 0x2F0, 10 }, { 0x2D0, 10 }, >> + { 0x2B0, 10 }, { 0x290, 10 }, { 0x270, 10 }, { 0x250, 10 }, { 0x230, 10 }, >> + { 0x210, 10 }, >> + { 0x0F8, 8 }, { 0x0E8, 8 }, { 0x0D8, 8 }, { 0x0C8, 8 }, { 0x0B8, 8 }, >> + { 0x0A8, 8 }, { 0x098, 8 }, { 0x088, 8 }, >> + { 0x03C, 6 }, { 0x034, 6 }, { 0x02C, 6 }, { 0x024, 6 }, >> + { 0x00E, 4 }, { 0x00A, 4 }, >> + { 0x003, 2 }, >> + { 0x000, 5 }, >> + { 0x001, 2 }, >> + { 0x002, 4 }, { 0x006, 4 }, >> + { 0x004, 6 }, { 0x00C, 6 }, { 0x014, 6 }, { 0x01C, 6 }, >> + { 0x008, 8 }, { 0x018, 8 }, { 0x028, 8 }, { 0x038, 8 }, { 0x048, 8 }, >> + { 0x058, 8 }, { 0x068, 8 }, { 0x078, 8 }, >> + { 0x010, 10 }, { 0x030, 10 }, { 0x050, 10 }, { 0x070, 10 }, { 0x090, 10 }, >> + { 0x0B0, 10 }, { 0x0D0, 10 }, { 0x0F0, 10 }, { 0x110, 10 }, { 0x130, 10 }, >> + { 0x150, 10 }, { 0x170, 10 }, { 0x190, 10 }, { 0x1B0, 10 }, { 0x1D0, 10 }, >> + { 0x1F0, 10 } >> }; > > would it be worth it to handle codes with length > the 8*sizeof in such a way > that this fits still in uint8 ? > You seem to believe that the reason I switch to little-endian codes (which entails switching to uint16_t) in this array lies in the VLC API. This is not so. The reason for the switch is that the little endian values are now directly written by put_bits_le(). If the values were not reversed, one would have to reverse them at runtime in the encoder. - Andreas
On Thu, Oct 15, 2020 at 04:12:01PM +0200, Andreas Rheinhardt wrote: > Michael Niedermayer: > > On Tue, Oct 13, 2020 at 11:10:14AM +0200, Andreas Rheinhardt wrote: > >> The ASUS V2 format is designed for a little-endian bitstream reader, yet > >> our encoder used an ordinary big-endian bitstream writer to write it; > >> the bits of every byte were swapped at the end and some data (namely the > >> numbers not in static tables) had to be bitreversed before writing it at > >> all, so that it would be reversed twice. > >> > >> This commit stops doing so; instead, a little-endian bitstream writer is > >> used. This also necessitated to switch certain static tables, which > >> required trivial modifications to the decoder (that uses the same > >> tables). > > [...] > > > >> -const uint8_t ff_asv2_level_tab[63][2] = { > >> - { 0x3F, 10 }, { 0x2F, 10 }, { 0x37, 10 }, { 0x27, 10 }, { 0x3B, 10 }, { 0x2B, 10 }, { 0x33, 10 }, { 0x23, 10 }, > >> - { 0x3D, 10 }, { 0x2D, 10 }, { 0x35, 10 }, { 0x25, 10 }, { 0x39, 10 }, { 0x29, 10 }, { 0x31, 10 }, { 0x21, 10 }, > >> - { 0x1F, 8 }, { 0x17, 8 }, { 0x1B, 8 }, { 0x13, 8 }, { 0x1D, 8 }, { 0x15, 8 }, { 0x19, 8 }, { 0x11, 8 }, > >> - { 0x0F, 6 }, { 0x0B, 6 }, { 0x0D, 6 }, { 0x09, 6 }, > >> - { 0x07, 4 }, { 0x05, 4 }, > >> - { 0x03, 2 }, > >> - { 0x00, 5 }, > >> - { 0x02, 2 }, > >> - { 0x04, 4 }, { 0x06, 4 }, > >> - { 0x08, 6 }, { 0x0C, 6 }, { 0x0A, 6 }, { 0x0E, 6 }, > >> - { 0x10, 8 }, { 0x18, 8 }, { 0x14, 8 }, { 0x1C, 8 }, { 0x12, 8 }, { 0x1A, 8 }, { 0x16, 8 }, { 0x1E, 8 }, > >> - { 0x20, 10 }, { 0x30, 10 }, { 0x28, 10 }, { 0x38, 10 }, { 0x24, 10 }, { 0x34, 10 }, { 0x2C, 10 }, { 0x3C, 10 }, > >> - { 0x22, 10 }, { 0x32, 10 }, { 0x2A, 10 }, { 0x3A, 10 }, { 0x26, 10 }, { 0x36, 10 }, { 0x2E, 10 }, { 0x3E, 10 }, > >> +const uint16_t ff_asv2_level_tab[63][2] = { > >> + { 0x3F0, 10 }, { 0x3D0, 10 }, { 0x3B0, 10 }, { 0x390, 10 }, { 0x370, 10 }, > >> + { 0x350, 10 }, { 0x330, 10 }, { 0x310, 10 }, { 0x2F0, 10 }, { 0x2D0, 10 }, > >> + { 0x2B0, 10 }, { 0x290, 10 }, { 0x270, 10 }, { 0x250, 10 }, { 0x230, 10 }, > >> + { 0x210, 10 }, > >> + { 0x0F8, 8 }, { 0x0E8, 8 }, { 0x0D8, 8 }, { 0x0C8, 8 }, { 0x0B8, 8 }, > >> + { 0x0A8, 8 }, { 0x098, 8 }, { 0x088, 8 }, > >> + { 0x03C, 6 }, { 0x034, 6 }, { 0x02C, 6 }, { 0x024, 6 }, > >> + { 0x00E, 4 }, { 0x00A, 4 }, > >> + { 0x003, 2 }, > >> + { 0x000, 5 }, > >> + { 0x001, 2 }, > >> + { 0x002, 4 }, { 0x006, 4 }, > >> + { 0x004, 6 }, { 0x00C, 6 }, { 0x014, 6 }, { 0x01C, 6 }, > >> + { 0x008, 8 }, { 0x018, 8 }, { 0x028, 8 }, { 0x038, 8 }, { 0x048, 8 }, > >> + { 0x058, 8 }, { 0x068, 8 }, { 0x078, 8 }, > >> + { 0x010, 10 }, { 0x030, 10 }, { 0x050, 10 }, { 0x070, 10 }, { 0x090, 10 }, > >> + { 0x0B0, 10 }, { 0x0D0, 10 }, { 0x0F0, 10 }, { 0x110, 10 }, { 0x130, 10 }, > >> + { 0x150, 10 }, { 0x170, 10 }, { 0x190, 10 }, { 0x1B0, 10 }, { 0x1D0, 10 }, > >> + { 0x1F0, 10 } > >> }; > > > > would it be worth it to handle codes with length > the 8*sizeof in such a way > > that this fits still in uint8 ? > > > You seem to believe that the reason I switch to little-endian codes > (which entails switching to uint16_t) in this array lies in the VLC API. > This is not so. The reason for the switch is that the little endian > values are now directly written by put_bits_le(). If the values were not > reversed, one would have to reverse them at runtime in the encoder. right, i missed that this was of course also used in the encoder thx [...]
diff --git a/libavcodec/asv.c b/libavcodec/asv.c index 14fdf73333..dcae90982a 100644 --- a/libavcodec/asv.c +++ b/libavcodec/asv.c @@ -53,31 +53,37 @@ const uint8_t ff_asv_level_tab[7][2] = { }; const uint8_t ff_asv_dc_ccp_tab[8][2] = { - { 0x1, 2 }, { 0xD, 4 }, { 0xF, 4 }, { 0xC, 4 }, - { 0x5, 3 }, { 0xE, 4 }, { 0x4, 3 }, { 0x0, 2 }, + { 0x2, 2 }, { 0xB, 4 }, { 0xF, 4 }, { 0x3, 4 }, + { 0x5, 3 }, { 0x7, 4 }, { 0x1, 3 }, { 0x0, 2 }, }; const uint8_t ff_asv_ac_ccp_tab[16][2] = { - { 0x00, 2 }, { 0x3B, 6 }, { 0x0A, 4 }, { 0x3A, 6 }, - { 0x02, 3 }, { 0x39, 6 }, { 0x3C, 6 }, { 0x38, 6 }, - { 0x03, 3 }, { 0x3D, 6 }, { 0x08, 4 }, { 0x1F, 5 }, - { 0x09, 4 }, { 0x0B, 4 }, { 0x0D, 4 }, { 0x0C, 4 }, + { 0x00, 2 }, { 0x37, 6 }, { 0x05, 4 }, { 0x17, 6 }, + { 0x02, 3 }, { 0x27, 6 }, { 0x0F, 6 }, { 0x07, 6 }, + { 0x06, 3 }, { 0x2F, 6 }, { 0x01, 4 }, { 0x1F, 5 }, + { 0x09, 4 }, { 0x0D, 4 }, { 0x0B, 4 }, { 0x03, 4 }, }; -const uint8_t ff_asv2_level_tab[63][2] = { - { 0x3F, 10 }, { 0x2F, 10 }, { 0x37, 10 }, { 0x27, 10 }, { 0x3B, 10 }, { 0x2B, 10 }, { 0x33, 10 }, { 0x23, 10 }, - { 0x3D, 10 }, { 0x2D, 10 }, { 0x35, 10 }, { 0x25, 10 }, { 0x39, 10 }, { 0x29, 10 }, { 0x31, 10 }, { 0x21, 10 }, - { 0x1F, 8 }, { 0x17, 8 }, { 0x1B, 8 }, { 0x13, 8 }, { 0x1D, 8 }, { 0x15, 8 }, { 0x19, 8 }, { 0x11, 8 }, - { 0x0F, 6 }, { 0x0B, 6 }, { 0x0D, 6 }, { 0x09, 6 }, - { 0x07, 4 }, { 0x05, 4 }, - { 0x03, 2 }, - { 0x00, 5 }, - { 0x02, 2 }, - { 0x04, 4 }, { 0x06, 4 }, - { 0x08, 6 }, { 0x0C, 6 }, { 0x0A, 6 }, { 0x0E, 6 }, - { 0x10, 8 }, { 0x18, 8 }, { 0x14, 8 }, { 0x1C, 8 }, { 0x12, 8 }, { 0x1A, 8 }, { 0x16, 8 }, { 0x1E, 8 }, - { 0x20, 10 }, { 0x30, 10 }, { 0x28, 10 }, { 0x38, 10 }, { 0x24, 10 }, { 0x34, 10 }, { 0x2C, 10 }, { 0x3C, 10 }, - { 0x22, 10 }, { 0x32, 10 }, { 0x2A, 10 }, { 0x3A, 10 }, { 0x26, 10 }, { 0x36, 10 }, { 0x2E, 10 }, { 0x3E, 10 }, +const uint16_t ff_asv2_level_tab[63][2] = { + { 0x3F0, 10 }, { 0x3D0, 10 }, { 0x3B0, 10 }, { 0x390, 10 }, { 0x370, 10 }, + { 0x350, 10 }, { 0x330, 10 }, { 0x310, 10 }, { 0x2F0, 10 }, { 0x2D0, 10 }, + { 0x2B0, 10 }, { 0x290, 10 }, { 0x270, 10 }, { 0x250, 10 }, { 0x230, 10 }, + { 0x210, 10 }, + { 0x0F8, 8 }, { 0x0E8, 8 }, { 0x0D8, 8 }, { 0x0C8, 8 }, { 0x0B8, 8 }, + { 0x0A8, 8 }, { 0x098, 8 }, { 0x088, 8 }, + { 0x03C, 6 }, { 0x034, 6 }, { 0x02C, 6 }, { 0x024, 6 }, + { 0x00E, 4 }, { 0x00A, 4 }, + { 0x003, 2 }, + { 0x000, 5 }, + { 0x001, 2 }, + { 0x002, 4 }, { 0x006, 4 }, + { 0x004, 6 }, { 0x00C, 6 }, { 0x014, 6 }, { 0x01C, 6 }, + { 0x008, 8 }, { 0x018, 8 }, { 0x028, 8 }, { 0x038, 8 }, { 0x048, 8 }, + { 0x058, 8 }, { 0x068, 8 }, { 0x078, 8 }, + { 0x010, 10 }, { 0x030, 10 }, { 0x050, 10 }, { 0x070, 10 }, { 0x090, 10 }, + { 0x0B0, 10 }, { 0x0D0, 10 }, { 0x0F0, 10 }, { 0x110, 10 }, { 0x130, 10 }, + { 0x150, 10 }, { 0x170, 10 }, { 0x190, 10 }, { 0x1B0, 10 }, { 0x1D0, 10 }, + { 0x1F0, 10 } }; av_cold void ff_asv_common_init(AVCodecContext *avctx) diff --git a/libavcodec/asv.h b/libavcodec/asv.h index a1366b6fe4..e778c72d34 100644 --- a/libavcodec/asv.h +++ b/libavcodec/asv.h @@ -66,7 +66,7 @@ extern const uint8_t ff_asv_ccp_tab[17][2]; extern const uint8_t ff_asv_level_tab[7][2]; extern const uint8_t ff_asv_dc_ccp_tab[8][2]; extern const uint8_t ff_asv_ac_ccp_tab[16][2]; -extern const uint8_t ff_asv2_level_tab[63][2]; +extern const uint16_t ff_asv2_level_tab[63][2]; void ff_asv_common_init(AVCodecContext *avctx); diff --git a/libavcodec/asvdec.c b/libavcodec/asvdec.c index 78e53b6402..6b6d7b3fa7 100644 --- a/libavcodec/asvdec.c +++ b/libavcodec/asvdec.c @@ -55,21 +55,18 @@ static av_cold void init_vlcs(ASV1Context *a) INIT_VLC_STATIC(&ccp_vlc, CCP_VLC_BITS, 17, &ff_asv_ccp_tab[0][1], 2, 1, &ff_asv_ccp_tab[0][0], 2, 1, 32); - INIT_CUSTOM_VLC_STATIC(&dc_ccp_vlc, DC_CCP_VLC_BITS, 8, - &ff_asv_dc_ccp_tab[0][1], 2, 1, - &ff_asv_dc_ccp_tab[0][0], 2, 1, - INIT_VLC_OUTPUT_LE, 16); - INIT_CUSTOM_VLC_STATIC(&ac_ccp_vlc, AC_CCP_VLC_BITS, 16, - &ff_asv_ac_ccp_tab[0][1], 2, 1, - &ff_asv_ac_ccp_tab[0][0], 2, 1, - INIT_VLC_OUTPUT_LE, 64); + INIT_LE_VLC_STATIC(&dc_ccp_vlc, DC_CCP_VLC_BITS, 8, + &ff_asv_dc_ccp_tab[0][1], 2, 1, + &ff_asv_dc_ccp_tab[0][0], 2, 1, 16); + INIT_LE_VLC_STATIC(&ac_ccp_vlc, AC_CCP_VLC_BITS, 16, + &ff_asv_ac_ccp_tab[0][1], 2, 1, + &ff_asv_ac_ccp_tab[0][0], 2, 1, 64); INIT_VLC_STATIC(&level_vlc, ASV1_LEVEL_VLC_BITS, 7, &ff_asv_level_tab[0][1], 2, 1, &ff_asv_level_tab[0][0], 2, 1, 16); - INIT_CUSTOM_VLC_STATIC(&asv2_level_vlc, ASV2_LEVEL_VLC_BITS, 63, - &ff_asv2_level_tab[0][1], 2, 1, - &ff_asv2_level_tab[0][0], 2, 1, - INIT_VLC_OUTPUT_LE, 1024); + INIT_LE_VLC_STATIC(&asv2_level_vlc, ASV2_LEVEL_VLC_BITS, 63, + &ff_asv2_level_tab[0][1], 4, 2, + &ff_asv2_level_tab[0][0], 4, 2, 1024); } } diff --git a/libavcodec/asvenc.c b/libavcodec/asvenc.c index 49df3ffbe1..718ffa1836 100644 --- a/libavcodec/asvenc.c +++ b/libavcodec/asvenc.c @@ -32,14 +32,8 @@ #include "dct.h" #include "fdctdsp.h" #include "internal.h" -#include "mathops.h" #include "mpeg12data.h" -static inline void asv2_put_bits(PutBitContext *pb, int n, int v) -{ - put_bits(pb, n, ff_reverse[v << (8 - n)]); -} - static inline void asv1_put_level(PutBitContext *pb, int level) { unsigned int index = level + 3; @@ -57,14 +51,14 @@ static inline void asv2_put_level(ASV1Context *a, PutBitContext *pb, int level) unsigned int index = level + 31; if (index <= 62) { - put_bits(pb, ff_asv2_level_tab[index][1], ff_asv2_level_tab[index][0]); + put_bits_le(pb, ff_asv2_level_tab[index][1], ff_asv2_level_tab[index][0]); } else { - put_bits(pb, 5, 0); /* Escape code */ + put_bits_le(pb, 5, 0); /* Escape code */ if (level < -128 || level > 127) { av_log(a->avctx, AV_LOG_WARNING, "Clipping level %d, increase qscale\n", level); level = av_clip_int8(level); } - asv2_put_bits(pb, 8, level & 0xFF); + put_bits_le(pb, 8, level & 0xFF); } } @@ -127,8 +121,8 @@ static inline void asv2_encode_block(ASV1Context *a, int16_t block[64]) count >>= 2; - asv2_put_bits(&a->pb, 4, count); - asv2_put_bits(&a->pb, 8, (block[0] + 32) >> 6); + put_bits_le(&a->pb, 4, count); + put_bits_le(&a->pb, 8, (block[0] + 32) >> 6); block[0] = 0; for (i = 0; i <= count; i++) { @@ -150,9 +144,9 @@ static inline void asv2_encode_block(ASV1Context *a, int16_t block[64]) av_assert2(i || ccp < 8); if (i) - put_bits(&a->pb, ff_asv_ac_ccp_tab[ccp][1], ff_asv_ac_ccp_tab[ccp][0]); + put_bits_le(&a->pb, ff_asv_ac_ccp_tab[ccp][1], ff_asv_ac_ccp_tab[ccp][0]); else - put_bits(&a->pb, ff_asv_dc_ccp_tab[ccp][1], ff_asv_dc_ccp_tab[ccp][0]); + put_bits_le(&a->pb, ff_asv_dc_ccp_tab[ccp][1], ff_asv_dc_ccp_tab[ccp][0]); if (ccp) { if (ccp & 8) @@ -291,17 +285,16 @@ static int encode_frame(AVCodecContext *avctx, AVPacket *pkt, } emms_c(); - flush_put_bits(&a->pb); + if (avctx->codec_id == AV_CODEC_ID_ASV1) + flush_put_bits(&a->pb); + else + flush_put_bits_le(&a->pb); AV_WN32(put_bits_ptr(&a->pb), 0); size = (put_bits_count(&a->pb) + 31) / 32; if (avctx->codec_id == AV_CODEC_ID_ASV1) { a->bbdsp.bswap_buf((uint32_t *) pkt->data, (uint32_t *) pkt->data, size); - } else { - int i; - for (i = 0; i < 4 * size; i++) - pkt->data[i] = ff_reverse[pkt->data[i]]; } pkt->size = size * 4;
The ASUS V2 format is designed for a little-endian bitstream reader, yet our encoder used an ordinary big-endian bitstream writer to write it; the bits of every byte were swapped at the end and some data (namely the numbers not in static tables) had to be bitreversed before writing it at all, so that it would be reversed twice. This commit stops doing so; instead, a little-endian bitstream writer is used. This also necessitated to switch certain static tables, which required trivial modifications to the decoder (that uses the same tables). Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com> --- libavcodec/asv.c | 46 +++++++++++++++++++++++++-------------------- libavcodec/asv.h | 2 +- libavcodec/asvdec.c | 21 +++++++++------------ libavcodec/asvenc.c | 29 +++++++++++----------------- 4 files changed, 47 insertions(+), 51 deletions(-)