[Libav-user] How to read in planar data?

[Mark McKay]

Nevermind - it looks like  my problem is coming from running out of 
memory.  I did however learn that I ought to use aFrame->extended_data 
instead of aFrame->data when reading audio data.




On 2019-12-06 17:37, Mark McKay wrote:
> I should have mentioned that the crash was happening at the line
> 
>     _audioBytes.append((const char *)aFrame->data[ch] + i *
> sizeof(float), sizeof(float));
> 
> This happens after 590 packets have been read.
> 
> 
> 
> 
> On 2019-12-06 17:05, Mark McKay wrote:
>> Just reposting with properly formatted code:
>> 
>> void AudioReader::loadAudioFrame(AVPacket& packet)
>> {
>>     int err = avcodec_send_packet(aCodecCtx, &packet);
>>     if (err < 0)
>>     {
>>         qDebug("Error sending packet to decoder");
>>         return;
>>     }
>> 
>>     int numChannels = aCodecCtx->channels;
>>     while (err >= 0)
>>     {
>>         err = avcodec_receive_frame(aCodecCtx, aFrame);
>>         if (err == AVERROR(EAGAIN) || err == AVERROR_EOF)
>>         return;
>> 
>>         if (err < 0)
>>         {
>>             qDebug() << "Error decoding packet: " << err;
>>             return;
>>         }
>> 
>>         int bytesPerSample = 
>> av_get_bytes_per_sample(aCodecCtx->sample_fmt);
>>         if (bytesPerSample < 0) {
>>         /* This should not occur, checking just for paranoia */
>>         fprintf(stderr, "Failed to calculate data size\n");
>>         exit(1);
>>         }
>> 
>> 
>>         int bufferSize = av_samples_get_buffer_size(NULL,
>>         aCodecCtx->channels,
>>         aFrame->nb_samples,
>>         aCodecCtx->sample_fmt,
>>         1);
>> 
>>         switch(aCodecCtx->sample_fmt)
>>         {
>>         ...
>>         case AV_SAMPLE_FMT_FLTP:
>>         {
>>             //Serial format - change to interleaved order
>>             for (int i = 0; i < aFrame->nb_samples; i++)
>>             {
>>                 for (int ch = 0; ch < numChannels; ch++)
>>                 {
>>                     _audioBytes.append((const char *)aFrame->data[ch] 
>> + i * sizeof(float), sizeof(float));
>>                 }
>>             }
>>             break;
>>         }
>>         }
>> }
>> 
>> 
>> 
>> On 2019-12-06 15:55, Mark McKay wrote:
>> 
>>> I'm having trouble trying to figure out how to read in planar data.  
>>> I have a file that has data in the AV_SAMPLE_FMT_FLTP format with two 
>>> channels.  Below is my attempt to read in the data.  This will crash 
>>> with a memory problem if I try to read in the second channel (but it 
>>> will complete if I bypass the second and just read from channel 0).  
>>> Am I doing this right?
>> _______________________________________________
>> Libav-user mailing list
>> Libav-user at ffmpeg.org
>> https://ffmpeg.org/mailman/listinfo/libav-user
>> 
>> To unsubscribe, visit link above, or email
>> libav-user-request at ffmpeg.org with subject "unsubscribe".
> _______________________________________________
> Libav-user mailing list
> Libav-user at ffmpeg.org
> https://ffmpeg.org/mailman/listinfo/libav-user
> 
> To unsubscribe, visit link above, or email
> libav-user-request at ffmpeg.org with subject "unsubscribe".