Cool! I didn't even think to add vorbis support when I landed the id3v2
stuff this week. Nice idea, Ralph!
The "tag_spec" is supposed to be a private data structure so I don't
want us to be reading it directly for the long term, but this will
method work fine for now. :-)
I'll try the patch out tonight when I get home and see if I can land it
in the tree. Stay tuned.
On Thu, Jul 06, 2000 at 04:41:59PM -0700, Ralph Giles wrote:
> Here's a rough patch to make the --t* options work with the vorbis
> encoder. I just read the id3tag_spec structure directly, but the vorbis
> comment header is a lot more flexible than the id3 stuff, so I'd really
> want to write a general token-value interface to the metadata if I were
> "doing it right." Maybe some other time.
This is probably a stupid question, but are there docs online about the
comment header (he asks without even looking at the vorbis site first)?
> bugs:
>
> Doesn't support track number of genre. The second is easy to add; the
> first requires that the vorbis people settle on a standard form for it.
Let me see if I can contact them about their plans for this. If you
find out anything, let us all know. Maybe I can hack in the genre. :-)
> Ogg encoding doesn't actually work for me; it prints a lot of:
>
> *********
> decoded bytes=0 16384
>
> And generates a 128 byte file. This happens without my patch as well.
>
> Oops, on further investigation those 128 bytes are an id3 tag, so we need
> an interlock of some sort. I'll take a look at this later.
Cool. Just send another patch.
> Cheers,
> -ralph
>
> --
> [EMAIL PROTECTED]
> Ahh, the joys of tracking cvs!
>
> Index: parse.c
> ===================================================================
> RCS file: /cvsroot/lame/lame/parse.c,v
> retrieving revision 1.68
> diff -u -r1.68 parse.c
> --- parse.c 2000/07/04 03:28:43 1.68
> +++ parse.c 2000/07/06 23:34:44
> @@ -180,6 +180,9 @@
> PRINTF1(" characters), or the `--add-id3v2' or `--id3v2-only' options are
>used,\n");
> PRINTF1(" or output is redirected to stdout.\n");
> PRINTF1("\n");
> + PRINTF1(" The --t* options will work with the Ogg encoder, but the
>id3-specific\n");
> + PRINTF1(" features will be ignored. Genre and tracknumber are currently
>ignored.\n");
> + PRINTF1("\n");
> #ifdef HAVEGTK
> PRINTF1(" -g run graphical analysis on <infile>\n");
> #endif
> @@ -832,6 +835,3 @@
> }
>
> }
> -
> -
> -
> Index: vorbis_interface.c
> ===================================================================
> RCS file: /cvsroot/lame/lame/vorbis_interface.c,v
> retrieving revision 1.10
> diff -u -r1.10 vorbis_interface.c
> --- vorbis_interface.c 2000/06/30 14:25:53 1.10
> +++ vorbis_interface.c 2000/07/06 23:34:50
> @@ -93,7 +93,7 @@
> set up the Vorbis decoder */
>
> /* The next two packets in order are the comment and codebook headers.
> - They're likely large and may span multiple pages. Thus we reead
> + They're likely large and may span multiple pages. Thus we read
> and submit data until we get our two pacakets, watching that no
> pages are missing. If a page is missing, error out; losing a
> header page is the only place where missing data is fatal. */
> @@ -225,11 +225,6 @@
> break;
> }
>
> -
> -
> -
> -
> -
> result=ogg_sync_pageout(&oy,&og);
>
> if(result==0) {
> @@ -300,15 +295,6 @@
>
>
>
> -
> -
> -
> -
> -
> -
> -
> -
> -
> ogg_stream_state os2; /* take physical pages, weld into a logical
> stream of packets */
> ogg_page og2; /* one Ogg bitstream page. Vorbis packets are inside */
> @@ -320,13 +306,13 @@
> vorbis_dsp_state vd2; /* central working state for the packet->PCM decoder */
> vorbis_block vb2; /* local working space for packet->PCM decode */
>
> -
>
> +#define vorbis_comment_size 256
>
> int lame_encode_ogg_init(lame_global_flags *gfp)
> {
> lame_internal_flags *gfc=gfp->internal_flags;
> -
> + char comment[vorbis_comment_size];
>
> /********** Encode setup ************/
>
> @@ -337,10 +323,43 @@
> vi2.rate = gfp->out_samplerate;
>
>
> - /* add a comment */
> + /* fill out the comments header */
> vorbis_comment_init(&vc2);
> vorbis_comment_add(&vc2,"Track encoded using L.A.M.E. libvorbis interface.");
> -
> +
> + /* these fields are marked "private data members" in the header, so
> + 'twould be better to add id3tag_get_*() functions instead. However,
> + the vorbis comment fields are much more flexible than the id3 stuff
> + so I'll just do this in anticipation of (another) metatag rewrite */
> +
> + if(gfp->tag_spec.title) {
> + snprintf(comment,vorbis_comment_size,
> + "TITLE=%s",gfp->tag_spec.title);
> + vorbis_comment_add(&vc2,comment);
> + }
> + if(gfp->tag_spec.artist) {
> + snprintf(comment,vorbis_comment_size,
> + "ARTIST=%s",gfp->tag_spec.artist);
> + vorbis_comment_add(&vc2,comment);
> + }
> + if(gfp->tag_spec.album) {
> + snprintf(comment,vorbis_comment_size,
> + "ALBUM=%s",gfp->tag_spec.album);
> + vorbis_comment_add(&vc2,comment);
> + }
> + /* pretend these are equivalent */
> + if(gfp->tag_spec.comment) {
> + snprintf(comment,vorbis_comment_size,
> + "DESCRIPTION=%s",gfp->tag_spec.comment);
> + vorbis_comment_add(&vc2,comment);
> + }
> + if(gfp->tag_spec.year) {
> + snprintf(comment,vorbis_comment_size,
> + "DATE=%d",gfp->tag_spec.year);
> + vorbis_comment_add(&vc2,comment);
> + }
> + /* FIXME: we drop the track number and genre for now */
> +
> /* set up the analysis state and auxiliary encoding storage */
> vorbis_analysis_init(&vd2,&vi2);
> vorbis_block_init(&vd2,&vb2);
> @@ -375,6 +394,7 @@
> return 0;
> }
>
> +#undef vorbis_comment_size
>
>
> int lame_encode_ogg_finish(lame_global_flags *gfp,
--
Don Melton
Mangler, Navigator
mailto:[EMAIL PROTECTED]
--
MP3 ENCODER mailing list ( http://geek.rcc.se/mp3encoder/ )
