musicin.c

Motion JPEG编解码器源代码

    }

    *num_samples = audio_bytes/(audio_bits/8);

    if (brt==0)
	{
        if (info->lay==2)
            brt = (info->mode == MPG_MD_MONO) ? 112 : 224;
        else
            brt = (info->mode == MPG_MD_MONO) ? 192 : 384;
	}

    for(j=0;j<15;j++) if (bitrate[info->lay-1][j] == brt) break;

    if (j==15)
        mjpeg_error_exit1("Bitrate of %d KBit/s not allowed!",brt);

    info->bitrate_index = j;

    if(info->lay==2 && brt>192 && info->mode==MPG_MD_MONO)
        mjpeg_error_exit1("Bitrate of %d KBit/s not allowed for MONO",brt);
    open_bit_stream_w(&bs, *encoded_file_name, BUFFER_SIZE);
}

/*
 * print_config
 *
 * PURPOSE:  Prints the encoding parameters used
*/

void
print_config(fr_ps, psy, num_samples, outPath)
frame_params *fr_ps;
int     *psy;
unsigned long *num_samples;
char    *outPath;
{
	layer *info = fr_ps->header;

	mjpeg_debug("Encoding configuration:");
	if(info->mode != MPG_MD_JOINT_STEREO)
		mjpeg_debug("Layer=%s   mode=%s   extn=%d   psy model=%d",
					layer_names[info->lay-1], mode_names[info->mode],
					info->mode_ext, *psy);
	else mjpeg_debug("Layer=%s   mode=%s   extn=data dependant   psy model=%d",
					 layer_names[info->lay-1], mode_names[info->mode], *psy);
	mjpeg_debug("samp frq=%.1f kHz   total bitrate=%d kbps",
				s_freq[info->sampling_frequency],
				bitrate[info->lay-1][info->bitrate_index]);
	mjpeg_debug("de-emph=%d   c/right=%d   orig=%d   errprot=%d",
				info->emphasis, info->copyright, info->original,
				info->error_protection);
	mjpeg_debug("output file: '%s'", outPath);
}
 
/*
 * main
 *
 * PURPOSE:  MPEG I Encoder supporting layers 1 and 2, and
 * psychoacoustic models 1 (MUSICAM) and 2 (AT&T)
 *
 * SEMANTICS:  One overlapping frame of audio of up to 2 channels are
 * processed at a time in the following order:
 * (associated routines are in parentheses)
 *
 * 1.  Filter sliding window of data to get 32 subband
 * samples per channel.
 * (window_subband,filter_subband)
 *
 * 2.  If joint stereo mode, combine left and right channels
 * for subbands above #jsbound#.
 * (*_combine_LR)
 *
 * 3.  Calculate scalefactors for the frame, and if layer 2,
 * also calculate scalefactor select information.
 * (*_scale_factor_calc)
 *
 * 4.  Calculate psychoacoustic masking levels using selected
 * psychoacoustic model.
 * (*_Psycho_One, psycho_anal)
 *
 * 5.  Perform iterative bit allocation for subbands with low
 * mask_to_noise ratios using masking levels from step 4.
 * (*_main_bit_allocation)
 *
 * 6.  If error protection flag is active, add redundancy for
 * error protection.
 * (*_CRC_calc)
 *
 * 7.  Pack bit allocation, scalefactors, and scalefactor select
 * information (layer 2) onto bitstream.
 * (*_encode_bit_alloc,*_encode_scale,II_transmission_pattern)
 *
 * 8.  Quantize subbands and pack them into bitstream
 * (*_subband_quantization, *_sample_encoding)
*/ 

int main(argc, argv)
int     argc;
char    **argv;
{
typedef double SBS[2][3][SCALE_BLOCK][SBLIMIT];
    SBS  *sb_sample;
typedef double JSBS[3][SCALE_BLOCK][SBLIMIT];
    JSBS *j_sample;
typedef double IN[2][HAN_SIZE];
    IN   *win_que;
typedef unsigned int SUB[2][3][SCALE_BLOCK][SBLIMIT];
    SUB  *subband;
 
    frame_params fr_ps;
    layer info;
    char *encoded_file_name;
    short **win_buf;
static short buffer[2][1152];
static unsigned int bit_alloc[2][SBLIMIT], scfsi[2][SBLIMIT];
static unsigned int scalar[2][3][SBLIMIT], j_scale[3][SBLIMIT];
static double ltmin[2][SBLIMIT], lgmin[2][SBLIMIT], max_sc[2][SBLIMIT];
    FLOAT snr32[32];
    short sam[2][1056];
    int whole_SpF, extra_slot = 0;
    double avg_slots_per_frame, frac_SpF, slot_lag;
    int model, stereo, error_protection;
static unsigned int crc;
    int i, j, k, adb;
    unsigned long bitsPerSlot, samplesPerFrame, frameNum = 0;
    unsigned long frameBits, sentBits = 0;
    unsigned long num_samples;

    /* Most large variables are declared dynamically to ensure
       compatibility with smaller machines */

    sb_sample = (SBS *) mem_alloc(sizeof(SBS), "sb_sample");
    j_sample = (JSBS *) mem_alloc(sizeof(JSBS), "j_sample");
    win_que = (IN *) mem_alloc(sizeof(IN), "Win_que");
    subband = (SUB *) mem_alloc(sizeof(SUB),"subband");
    win_buf = (short **) mem_alloc(sizeof(short *)*2, "win_buf");
 
    /* clear buffers */
    memset((char *) buffer, 0, sizeof(buffer));
    memset((char *) bit_alloc, 0, sizeof(bit_alloc));
    memset((char *) scalar, 0, sizeof(scalar));
    memset((char *) j_scale, 0, sizeof(j_scale));
    memset((char *) scfsi, 0, sizeof(scfsi));
    memset((char *) ltmin, 0, sizeof(ltmin));
    memset((char *) lgmin, 0, sizeof(lgmin));
    memset((char *) max_sc, 0, sizeof(max_sc));
    memset((char *) snr32, 0, sizeof(snr32));
    memset((char *) sam, 0, sizeof(sam));
 
    fr_ps.header = &info;
    fr_ps.tab_num = -1;             /* no table loaded */
    fr_ps.alloc = NULL;
    info.version = MPEG_AUDIO_ID;

    programName = argv[0];
    get_params(argc, argv, &fr_ps, &model, &num_samples, &encoded_file_name);
    print_config(&fr_ps, &model, &num_samples, encoded_file_name);

    hdr_to_frps(&fr_ps);
    stereo = fr_ps.stereo;
    error_protection = info.error_protection;
 
    if (info.lay == 1) { bitsPerSlot = 32; samplesPerFrame = 384;  }
    else               { bitsPerSlot = 8;  samplesPerFrame = 1152; }
    /* Figure average number of 'slots' per frame. */
    /* Bitrate means TOTAL for both channels, not per side. */
    avg_slots_per_frame = ((double)samplesPerFrame /
                           s_freq[info.sampling_frequency]) *
                          ((double)bitrate[info.lay-1][info.bitrate_index] /
                           (double)bitsPerSlot);
    whole_SpF = (int) avg_slots_per_frame;
    frac_SpF  = avg_slots_per_frame - (double)whole_SpF;
    slot_lag  = -frac_SpF;
	mjpeg_info("SpF=%d, frac SpF=%.3f, bitrate=%d kbps, sfreq=%.1f kHz",
	   whole_SpF, frac_SpF,
           bitrate[info.lay-1][info.bitrate_index],
           s_freq[info.sampling_frequency]);
 
    if (frac_SpF != 0)
       mjpeg_info("Fractional number of slots, padding required");
    else info.padding = 0;
 
    while (get_audio(musicin, buffer, num_samples, stereo, info.lay) != 0) {
       frameNum++;
       win_buf[0] = &buffer[0][0];
       win_buf[1] = &buffer[1][0];
       if (frac_SpF != 0) {
          if (slot_lag > (frac_SpF-1.0) ) {
             slot_lag -= frac_SpF;
             extra_slot = 0;
             info.padding = 0;
          }
          else {
             extra_slot = 1;
             info.padding = 1;
             slot_lag += (1-frac_SpF);
          }
       }
       adb = (whole_SpF+extra_slot) * bitsPerSlot;

       switch (info.lay) {
 
/***************************** Layer I **********************************/
 
          case 1 :
             for (j=0;j<SCALE_BLOCK;j++)
             for (k=0;k<stereo;k++) {
                window_subband(&win_buf[k], &(*win_que)[k][0], k);
                filter_subband(&(*win_que)[k][0], &(*sb_sample)[k][0][j][0]);
             }

             I_scale_factor_calc(*sb_sample, scalar, stereo);
             if(fr_ps.actual_mode == MPG_MD_JOINT_STEREO) {
                I_combine_LR(*sb_sample, *j_sample);
                I_scale_factor_calc(j_sample, &j_scale, 1);
             }
 
             put_scale(scalar, &fr_ps, max_sc);
 
             if (model == 1) I_Psycho_One(buffer, max_sc, ltmin, &fr_ps);
             else {
                for (k=0;k<stereo;k++) {
                   psycho_anal(&buffer[k][0],&sam[k][0], k, info.lay, snr32,
                               (FLOAT)s_freq[info.sampling_frequency]*1000);
                   for (i=0;i<SBLIMIT;i++) ltmin[k][i] = (double) snr32[i];
                }
             }
 
             I_main_bit_allocation(ltmin, bit_alloc, &adb, &fr_ps);
 
             if (error_protection) I_CRC_calc(&fr_ps, bit_alloc, &crc);
 
             encode_info(&fr_ps, &bs);
 
             if (error_protection) encode_CRC(crc, &bs);
 
             I_encode_bit_alloc(bit_alloc, &fr_ps, &bs);
             I_encode_scale(scalar, bit_alloc, &fr_ps, &bs);
             I_subband_quantization(scalar, *sb_sample, j_scale, *j_sample,
                                    bit_alloc, *subband, &fr_ps);
             I_sample_encoding(*subband, bit_alloc, &fr_ps, &bs);
             for (i=0;i<adb;i++) put1bit(&bs, 0);
          break;
 
/***************************** Layer 2 **********************************/
 
          case 2 :
             for (i=0;i<3;i++) for (j=0;j<SCALE_BLOCK;j++)
                for (k=0;k<stereo;k++) {
                   window_subband(&win_buf[k], &(*win_que)[k][0], k);
                   filter_subband(&(*win_que)[k][0], &(*sb_sample)[k][i][j][0]);
                }
 
                II_scale_factor_calc(*sb_sample, scalar, stereo, fr_ps.sblimit);
                pick_scale(scalar, &fr_ps, max_sc);
                if(fr_ps.actual_mode == MPG_MD_JOINT_STEREO) {
                   II_combine_LR(*sb_sample, *j_sample, fr_ps.sblimit);
                   II_scale_factor_calc(j_sample, &j_scale, 1, fr_ps.sblimit);
                }       /* this way we calculate more mono than we need */
                        /* but it is cheap */
 
                if (model == 1) II_Psycho_One(buffer, max_sc, ltmin, &fr_ps);
                else {
                   for (k=0;k<stereo;k++) {
                      psycho_anal(&buffer[k][0],&sam[k][0], k, 
                                 info.lay, snr32,
                                 (FLOAT)s_freq[info.sampling_frequency]*1000);
                      for (i=0;i<SBLIMIT;i++) ltmin[k][i] = (double) snr32[i];
                   }
                }
 
                II_transmission_pattern(scalar, scfsi, &fr_ps);
                II_main_bit_allocation(ltmin, scfsi, bit_alloc, &adb, &fr_ps);
 
                if (error_protection)
                   II_CRC_calc(&fr_ps, bit_alloc, scfsi, &crc);
 
                encode_info(&fr_ps, &bs);
 
                if (error_protection) encode_CRC(crc, &bs);
 
                II_encode_bit_alloc(bit_alloc, &fr_ps, &bs);
                II_encode_scale(bit_alloc, scfsi, scalar, &fr_ps, &bs);
                II_subband_quantization(scalar, *sb_sample, j_scale,
                                      *j_sample, bit_alloc, *subband, &fr_ps);
                II_sample_encoding(*subband, bit_alloc, &fr_ps, &bs);
                for (i=0;i<adb;i++) put1bit(&bs, 0);
          break;
 
/***************************** Layer 3 **********************************/

          case 3 : break;

       }
 
       frameBits = sstell(&bs) - sentBits;
       if(frameBits%bitsPerSlot)   /* a program failure */
          mjpeg_error("Sent %ld bits = %ld slots plus %ld",
                  frameBits, frameBits/bitsPerSlot,
                  frameBits%bitsPerSlot);
       sentBits += frameBits;
    }

    close_bit_stream_w(&bs);

	mjpeg_info("Num frames %ld Avg slots/frame = %.3f; b/smp = %.2f; br = %.3f kbps",
	   frameNum,
           (FLOAT) sentBits / (frameNum * bitsPerSlot),
           (FLOAT) sentBits / (frameNum * samplesPerFrame),
           (FLOAT) sentBits / (frameNum * samplesPerFrame) *
                               s_freq[info.sampling_frequency]);

	mjpeg_info("Encoding to layer %d with psychoacoustic model %d is finished", info.lay, model);
	mjpeg_info("The MPEG encoded output file name is \"%s\"",
			   encoded_file_name);
    exit(0);
}