From: Caleb Etemesi <etemesicaleb@gmail.com>
To: FFmpeg development discussions and patches <ffmpeg-devel@ffmpeg.org>
Subject: Re: [FFmpeg-devel] [PATCH v2] avcodec/jpeg2000: Add support for High-Throughput JPEG 2000 (HTJ2K) decoding.
Date: Wed, 21 Sep 2022 11:13:18 +0300
Message-ID: <CAF7=sG+1seZUgYd9CouGKJECqKfV1DvFBy-155YV3O6K9tknDA@mail.gmail.com> (raw)
In-Reply-To: <e4b88c8f25bb84818f48ab6204f22c649720cebc.camel@acc.umu.se>
>> Please separate struct renaming into a separate patch
Will do in the next patch
>> Corresponds to µ_start, right?
Not sure what µ_start here relates to, but this is the zero bit plane
information in *B.10.5*
of the jpeg2000 spec (ITU Rec. T.800 (06/2019))
>> Could also be called Lcup and Lref. But fine I suppose.
Will change in the next patch
>> No time to review this atm, hopefully later though
No problem.
>> Why are these tables in here when they're only used by jpeg2000htdec.c?
I wasn't aware of any table placing conventions and my personal preference
is usually to place them
inside header files.
On Tue, Sep 20, 2022 at 4:49 PM Tomas Härdin <tjoppen@acc.umu.se> wrote:
> tor 2022-09-08 klockan 23:49 +0300 skrev etemesicaleb@gmail.com:
> >
> > --- a/libavcodec/j2kenc.c
> > +++ b/libavcodec/j2kenc.c
> > @@ -106,7 +106,7 @@ static const int dwt_norms[2][4][10] = { //
> > [dwt_type][band][rlevel] (multiplied
> > typedef struct {
> > Jpeg2000Component *comp;
> > double *layer_rates;
> > -} Jpeg2000Tile;
> > +} Jpeg2000EncTile;
> >
> > typedef struct {
> > AVClass *class;
> > @@ -131,7 +131,7 @@ typedef struct {
> > Jpeg2000CodingStyle codsty;
> > Jpeg2000QuantStyle qntsty;
> >
> > - Jpeg2000Tile *tile;
> > + Jpeg2000EncTile *tile;
>
> Please separate struct renaming into a separate patch
>
> > typedef struct Jpeg2000Cblk {
> > uint8_t npasses;
> > uint8_t ninclpasses; // number coding of passes included in
> > codestream
> > @@ -181,6 +181,7 @@ typedef struct Jpeg2000Cblk {
> > uint16_t *lengthinc;
> > uint8_t nb_lengthinc;
> > uint8_t lblock;
> > + uint8_t zbp; // Zero bit planes
>
> Corresponds to µ_start, right?
>
> > uint8_t *data;
> > size_t data_allocated;
> > int nb_terminations;
> > @@ -189,6 +190,7 @@ typedef struct Jpeg2000Cblk {
> > Jpeg2000Pass *passes;
> > Jpeg2000Layer *layers;
> > int coord[2][2]; // border coordinates {{x0, x1}, {y0, y1}}
> > + int pass_lengths[2];
>
> Could also be called Lcup and Lref. But fine I suppose.
>
> > } Jpeg2000Cblk; // code block
> >
> > typedef struct Jpeg2000Prec {
> > @@ -227,6 +229,103 @@ typedef struct Jpeg2000Component {
> > uint8_t roi_shift; // ROI scaling value for the component
> > } Jpeg2000Component;
> >
> > +#define JP2_SIG_TYPE 0x6A502020
> > +#define JP2_SIG_VALUE 0x0D0A870A
> > +#define JP2_CODESTREAM 0x6A703263
> > +#define JP2_HEADER 0x6A703268
> > +
> > +#define HAD_COC 0x01
> > +#define HAD_QCC 0x02
> > +
> > +#define MAX_POCS 32
>
> Similarly with code movements. I can see this move adds is_htj2k to
> Jpeg2000DecoderContext. Please separate such things. Makes reviewing
> functional changes much easier.
>
> > -static inline void tile_codeblocks(const Jpeg2000DecoderContext *s,
> > Jpeg2000Tile *tile)
> > +static inline void tile_codeblocks(const Jpeg2000DecoderContext *s,
> > Jpeg2000DecTile *tile)
> > {
> > Jpeg2000T1Context t1;
> >
> > int compno, reslevelno, bandno;
> > + int subbandno;
> >
> > /* Loop on tile components */
> > for (compno = 0; compno < s->ncomponents; compno++) {
> > - Jpeg2000Component *comp = tile->comp + compno;
> > - Jpeg2000CodingStyle *codsty = tile->codsty + compno;
> > + Jpeg2000Component *comp = tile->comp + compno;
> > + Jpeg2000CodingStyle *codsty = tile->codsty + compno;
> > + Jpeg2000QuantStyle *quantsty = tile->qntsty + compno;
>
> Cosmetic and functional changes don't mix
>
> > int coded = 0;
> > + subbandno = 0;
> >
> > t1.stride = (1<<codsty->log2_cblk_width) + 2;
> >
> > @@ -1959,7 +1881,7 @@ static inline void tile_codeblocks(const
> > Jpeg2000DecoderContext *s, Jpeg2000Tile
> > for (reslevelno = 0; reslevelno < codsty->nreslevels2decode;
> > reslevelno++) {
> > Jpeg2000ResLevel *rlevel = comp->reslevel + reslevelno;
> > /* Loop on bands */
> > - for (bandno = 0; bandno < rlevel->nbands; bandno++) {
> > + for (bandno = 0; bandno < rlevel->nbands;
> > bandno++,subbandno++) {
> > int nb_precincts, precno;
> > Jpeg2000Band *band = rlevel->band + bandno;
> > int cblkno = 0, bandpos;
> > @@ -1979,12 +1901,21 @@ static inline void tile_codeblocks(const
> > Jpeg2000DecoderContext *s, Jpeg2000Tile
> > for (cblkno = 0;
> > cblkno < prec->nb_codeblocks_width * prec-
> > >nb_codeblocks_height;
> > cblkno++) {
> > - int x, y;
> > + int x, y, ret, magp;
> > Jpeg2000Cblk *cblk = prec->cblk + cblkno;
> > - int ret = decode_cblk(s, codsty, &t1, cblk,
> > - cblk->coord[0][1] - cblk-
> > >coord[0][0],
> > - cblk->coord[1][1] - cblk-
> > >coord[1][0],
> > - bandpos, comp->roi_shift);
> > + // Annex E (Equation E-2) ISO/IEC 15444-
> > 1:2019
> > + magp = quantsty->expn[subbandno] + quantsty-
> > >nguardbits - 1;
>
> Is expn[] allocated for regular j2k? I'd move this inside the is_htj2k
> block below
>
> > +
> > + if (s->is_htj2k)
> > + ret = decode_htj2k(s, codsty, &t1, cblk,
> > + cblk->coord[0][1] -
> > cblk->coord[0][0],
> > + cblk->coord[1][1] -
> > cblk->coord[1][0],
> > + magp, comp-
> > >roi_shift);
>
> > --- /dev/null
> > +++ b/libavcodec/jpeg2000htdec.c
>
> No time to review this atm, hopefully later though
>
> > --- /dev/null
> > +++ b/libavcodec/jpeg2000htdec.h
> >
> > +/**
> > + * @brief CtxVLC tables, borrowed from openhtj2k (
> > https://github.com/osamu620/OpenHTJ2K) (credits to Osamu Watanabe)
> > + */
> > +static const uint16_t dec_CxtVLC_table1[1024] = {
> > +static const uint16_t dec_CxtVLC_table0[1024] = {
>
> Why are these tables in here when they're only used by jpeg2000htdec.c?
>
> /Tomas
>
> _______________________________________________
> 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".
>
_______________________________________________
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".
next prev parent reply other threads:[~2022-09-21 8:13 UTC|newest]
Thread overview: 13+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-09-08 20:49 etemesicaleb
2022-09-09 15:46 ` Michael Niedermayer
2022-09-09 16:52 ` Caleb Etemesi
2022-09-09 17:11 ` Caleb Etemesi
2022-09-09 21:34 ` Andreas Rheinhardt
2022-09-10 9:13 ` Caleb Etemesi
2022-09-20 13:49 ` Tomas Härdin
2022-09-21 8:13 ` Caleb Etemesi [this message]
2022-09-21 10:30 ` Tomas Härdin
2022-09-21 14:08 ` Tomas Härdin
2022-09-22 4:07 ` Caleb Etemesi
2022-09-22 9:15 ` Tomas Härdin
2022-09-22 16:16 ` Caleb Etemesi
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to='CAF7=sG+1seZUgYd9CouGKJECqKfV1DvFBy-155YV3O6K9tknDA@mail.gmail.com' \
--to=etemesicaleb@gmail.com \
--cc=ffmpeg-devel@ffmpeg.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
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