SVC enhancement

[leslie-wang]

several enhancement: 1) add SVC syntax parser 2) implement SEI syntax generator 3) add one SVC split utility program.

[aizvorski]

@leslie-wang I wanted to cut a bugfix/maintenance release before adding this major feature in. The bugfix release would be 0.2.0 (released just now), and I think with the svc support we can call it 0.3.0.

Quick question: is this expected to work normally for non-svc files?

[leslie-wang]

My pull request can also work with non-svc stream. SVC is another profile of h.264. so for non-svc stream, it will go to normal parsing flow.

[aizvorski]

Looks good!

I'll add a few tests and call it a release. Sample files: http://concert.itec.aau.at/SVCDataset/svcseqs/

[leslie-wang]

sure. I can also help you adding some tests too.

Best Regards Leslie Qi Wang

On Thu, May 25, 2017 at 10:55 PM, Alex Izvorski notifications@github.com wrote:

Looks good!

I'll add a few tests and call it a release. Sample files: http://concert.itec.aau.at/SVCDataset/svcseqs/

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/aizvorski/h264bitstream/pull/17#issuecomment-304198639, or mute the thread https://github.com/notifications/unsubscribe-auth/ACa6Z44izm3bbafmEV_62ErV6pKAfsTGks5r9mlngaJpZM4MpuW7 .

[aizvorski]

Hi Leslie! I'm trying to make it so that running perl process.pl > h264_stream.c < h264_stream.in.c produces the exact same h264_stream.c as is currently checked in, but I'm running into some trouble you may be able to help with.

E.g. in this block of code:

               memcpy(h->sps_subset_table[h->sps_subset->sps->seq_parameter_set_id], h->sps_subset, sizeof(sps_subset_t));
               //memcpy(h->sps_subset_table[h->sps_subset->sps->seq_parameter_set_id]->sps, h->sps_subset->sps, sizeof(sps_t));
               //memcpy(h->sps_subset_table[h->sps_subset->sps->seq_parameter_set_id]->sps_svc_ext, h->sps_subset->sps_svc_ext, sizeof(sps_svc_ext_t));
               //h->sps_subset_table[h->sps_subset->sps->seq_parameter_set_id]->additional_extension2_flag = h->sps_subset->additional_extension2_flag;

it looks like sometimes when expanding that the first line should be commented out, sometimes the other 3 lines. Could you tell me more about how this is supposed to work? (There is one other piece of code like that)

There is a macro if( is_reading ) which expands to if ( 1 ) for reading and if ( 0 ) for writing, could this be used to accomplish what is needed here?

[leslie-wang]

can you please explain more on the cases? Seems like they should not be commented out.

Best Regards Leslie Qi Wang

On Fri, May 26, 2017 at 12:35 AM, Alex Izvorski notifications@github.com wrote:

Hi Leslie! I'm trying to make it so that running perl process.pl > h264_stream.c < h264_stream.in.c produces the exact same h264_stream.c as is currently checked in, but I'm running into some trouble you may be able to help with.

E.g. in this block of code:

           memcpy(h->sps_subset_table[h->sps_subset->sps->seq_parameter_set_id], h->sps_subset, sizeof(sps_subset_t));
           //memcpy(h->sps_subset_table[h->sps_subset->sps->seq_parameter_set_id]->sps, h->sps_subset->sps, sizeof(sps_t));
           //memcpy(h->sps_subset_table[h->sps_subset->sps->seq_parameter_set_id]->sps_svc_ext, h->sps_subset->sps_svc_ext, sizeof(sps_svc_ext_t));
           //h->sps_subset_table[h->sps_subset->sps->seq_parameter_set_id]->additional_extension2_flag = h->sps_subset->additional_extension2_flag;

it looks like sometimes when expanding that the first line should be commented out, sometimes the other 3 lines. Could you tell me more about how this is supposed to work? (There is one other piece of code like that)

There is a macro if( is_reading ) which expands to if ( 1 ) for reading and if ( 0 ) for writing, could this be used to accomplish what is needed here?

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/aizvorski/h264bitstream/pull/17#issuecomment-304213945, or mute the thread https://github.com/notifications/unsubscribe-auth/ACa6ZxAGHOl083jX-lnQdi5lwnDYG7vvks5r9oDNgaJpZM4MpuW7 .