GitHub/dsd-fme
1
0
Fork 0
mirror of https://github.com/lwvmobile/dsd-fme.git synced 2026-11-13 01:07:24 -05:00
Code Issues Packages Projects Releases Wiki Activity

Fix Float Baseline; Notes on P25p2 Audio;

Browse Source
This commit is contained in:
lwvmobile
2023-08-04 18:46:49 -04:00
parent a93177eb61
commit 363c99ceca
3 changed files with 55 additions and 41 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
src/dmr_bs.c
View File

@@ -517,8 +517,8 @@ void dmrBS (dsd_opts * opts, dsd_state * state)
if (opts->mbe_out_fR != NULL) closeMbeOutFileR (opts, state);
//clear out any stale audio storage buffers
memset (state->f_l4, 0.1f, sizeof(state->f_l4));
memset (state->f_r4, 0.1f, sizeof(state->f_r4));
memset (state->f_l4, 0.0f, sizeof(state->f_l4));
memset (state->f_r4, 0.0f, sizeof(state->f_r4));
memset (state->s_l4, 0, sizeof(state->s_l4));
memset (state->s_r4, 0, sizeof(state->s_r4));

68
src/dsd_audio2.c
View File

@@ -71,10 +71,9 @@ void playSynthesizedVoiceFS3 (dsd_opts * opts, dsd_state * state)
float stereo_samp2[320]; //8k 2-channel stereo interleave mix
float stereo_samp3[320]; //8k 2-channel stereo interleave mix
//note, always set a stereo float mix to a baseline value and not zero!
memset (stereo_samp1, 0.1f, sizeof(stereo_samp1));
memset (stereo_samp2, 0.1f, sizeof(stereo_samp2));
memset (stereo_samp3, 0.1f, sizeof(stereo_samp3));
memset (stereo_samp1, 0.0f, sizeof(stereo_samp1));
memset (stereo_samp2, 0.0f, sizeof(stereo_samp2));
memset (stereo_samp3, 0.0f, sizeof(stereo_samp3));
//dmr enc checkdown for whether or not to fill the stereo sample or not for playback or writing
encL = encR = 0;
@@ -181,12 +180,12 @@ void playSynthesizedVoiceFS3 (dsd_opts * opts, dsd_state * state)
FS3_END:
memset (state->audio_out_temp_buf, 0.1f, sizeof(state->audio_out_temp_buf));
memset (state->audio_out_temp_bufR, 0.1f, sizeof(state->audio_out_temp_bufR));
memset (state->audio_out_temp_buf, 0.0f, sizeof(state->audio_out_temp_buf));
memset (state->audio_out_temp_bufR, 0.0f, sizeof(state->audio_out_temp_bufR));
//set float temp buffer to baseline
memset (state->f_l4, 0.1f, sizeof(state->f_l4));
memset (state->f_r4, 0.1f, sizeof(state->f_r4));
memset (state->f_l4, 0.0f, sizeof(state->f_l4));
memset (state->f_r4, 0.0f, sizeof(state->f_r4));
//if we don't run processAudio, then I don't think we really need any of the items below active
state->audio_out_idx = 0;
@@ -212,6 +211,23 @@ void playSynthesizedVoiceFS3 (dsd_opts * opts, dsd_state * state)
}
//NOTE: On FS4 and SS4 voice, the longer the transmission, the more the function will start to lag
//the entire DSD-FME loop due to the skipping of playback on SACCH frames (causes noticeable skip when it does play them),
//this isn't a major problem, since the buffer can handle it, but it does delay return to CC until the end
//of the call on busy systems where both VCH slots are constantly busy with voice
//the longer the call, the more delayed until returning to the control channel
//NOTE: Disabling voice synthesis clears up the delay issue (obviosly since we aren't having to wait on it to play)
//disabling voice in only one slot will also fix most random stutter from the 4v in one slot, and 2v in the other slot
//NOTE: The same skip may be occurring on the main and v2.1b branches of DSD-FME as well, so that may be due to the 4v/2v and
//playing back immediately instead of buffering x number of samples or 4v/2v to get a smoother playback
//NOTE: When using capture bins for playback, this issue is not as observable compared to real time reception due to how fast
//we can blow through pure data on bin files compared to waiting for the real time reception
//its usually a lot more noticeable on dual voices than single (probably due to various arrangements of dual 4v/2v in each superframe)
//float stereo mix 4v2 P25p2
void playSynthesizedVoiceFS4 (dsd_opts * opts, dsd_state * state)
{
@@ -263,13 +279,12 @@ void playSynthesizedVoiceFS4 (dsd_opts * opts, dsd_state * state)
if (opts->slot1_on == 0) encL = 1;
if (opts->slot2_on == 0) encR = 1;
//note, always set a stereo float mix to a baseline value and not zero!
memset (stereo_samp1, 0.1f, sizeof(stereo_samp1));
memset (stereo_samp2, 0.1f, sizeof(stereo_samp2));
memset (stereo_samp3, 0.1f, sizeof(stereo_samp3));
memset (stereo_samp4, 0.1f, sizeof(stereo_samp4));
memset (stereo_samp1, 0.0f, sizeof(stereo_samp1));
memset (stereo_samp2, 0.0f, sizeof(stereo_samp2));
memset (stereo_samp3, 0.0f, sizeof(stereo_samp3));
memset (stereo_samp4, 0.0f, sizeof(stereo_samp4));
memset (empty, 0.1f, sizeof(empty));
memset (empty, 0.0f, sizeof(empty));
//run autogain on the f_ buffers
agf (opts, state, state->f_l4[0],0);
@@ -280,7 +295,7 @@ void playSynthesizedVoiceFS4 (dsd_opts * opts, dsd_state * state)
agf (opts, state, state->f_r4[2],1);
agf (opts, state, state->f_l4[3],0);
agf (opts, state, state->f_r4[3],1);
agf (opts, state, empty,2); //running empty may cause auto gain control issues, and also any auto gain applied here will invalidate the memcmp
agf (opts, state, empty,2); //running empty may invalidate the memcmp, ideally both this empty and an empty f_ should come out with the same result anyways
//interleave left and right channels from the temp (float) buffer with makeshift 'volume' decimation
for (i = 0; i < 160; i++)
@@ -344,12 +359,12 @@ void playSynthesizedVoiceFS4 (dsd_opts * opts, dsd_state * state)
END_FS4:
memset (state->audio_out_temp_buf, 0.1f, sizeof(state->audio_out_temp_buf));
memset (state->audio_out_temp_bufR, 0.1f, sizeof(state->audio_out_temp_bufR));
memset (state->audio_out_temp_buf, 0.0f, sizeof(state->audio_out_temp_buf));
memset (state->audio_out_temp_bufR, 0.0f, sizeof(state->audio_out_temp_bufR));
//set float temp buffer to baseline
memset (state->f_l4, 0.1f, sizeof(state->f_l4));
memset (state->f_r4, 0.1f, sizeof(state->f_r4));
memset (state->f_l4, 0.0f, sizeof(state->f_l4));
memset (state->f_r4, 0.0f, sizeof(state->f_r4));
//if we don't run processAudio, then I don't think we really need any of the items below active
state->audio_out_idx = 0;
@@ -383,8 +398,7 @@ void playSynthesizedVoiceFS (dsd_opts * opts, dsd_state * state)
int encL;
float stereo_samp1[320]; //8k 2-channel stereo interleave mix
//note, always set a stereo float mix to a baseline value and not zero!
memset (stereo_samp1, 0.1f, sizeof(stereo_samp1));
memset (stereo_samp1, 0.0f, sizeof(stereo_samp1));
//TODO: ENC Check on P25p1, DMR MS, etc
encL = 0;
@@ -424,12 +438,12 @@ void playSynthesizedVoiceFS (dsd_opts * opts, dsd_state * state)
FS_END:
memset (state->audio_out_temp_buf, 0.1f, sizeof(state->audio_out_temp_buf));
memset (state->audio_out_temp_bufR, 0.1f, sizeof(state->audio_out_temp_bufR));
memset (state->audio_out_temp_buf, 0.0f, sizeof(state->audio_out_temp_buf));
memset (state->audio_out_temp_bufR, 0.0f, sizeof(state->audio_out_temp_bufR));
//set float temp buffer to baseline
memset (state->f_l4, 0.1f, sizeof(state->f_l4));
memset (state->f_r4, 0.1f, sizeof(state->f_r4));
memset (state->f_l4, 0.0f, sizeof(state->f_l4));
memset (state->f_r4, 0.0f, sizeof(state->f_r4));
//if we don't run processAudio, then I don't think we really need any of the items below active
state->audio_out_idx = 0;
@@ -890,7 +904,7 @@ void agf (dsd_opts * opts, dsd_state * state, float samp[160], int slot)
int i, j, run;
run = 1;
float empty[160];
memset (empty, 0.1f, sizeof(empty));
memset (empty, 0.0f, sizeof(empty));
float mmax = 0.90f;
float mmin = -0.90f;
@@ -966,7 +980,7 @@ void agf (dsd_opts * opts, dsd_state * state, float samp[160], int slot)
// int i, run;
// run = 1;
// float empty[160];
// memset (empty, 0.1f, sizeof(empty));
// memset (empty, 0.0f, sizeof(empty));
// float mmax = 0.75f;
// float mmin = -0.75f;

24
src/dsd_main.c
View File

@@ -494,16 +494,16 @@ if(opts->frame_m17 == 1) //&& opts->audio_in_type == 5
memset(state->m17_meta, 0, sizeof(state->m17_meta));
//set float temp buffer to baseline
memset (state->audio_out_temp_buf, 0.1f, sizeof(state->audio_out_temp_buf));
memset (state->audio_out_temp_bufR, 0.1f, sizeof(state->audio_out_temp_bufR));
memset (state->audio_out_temp_buf, 0.0f, sizeof(state->audio_out_temp_buf));
memset (state->audio_out_temp_bufR, 0.0f, sizeof(state->audio_out_temp_bufR));
//set float temp buffer to baseline
memset (state->f_l, 0.1f, sizeof(state->f_l));
memset (state->f_r, 0.1f, sizeof(state->f_r));
memset (state->f_l, 0.0f, sizeof(state->f_l));
memset (state->f_r, 0.0f, sizeof(state->f_r));
//set float temp buffer to baseline
memset (state->f_l4, 0.1f, sizeof(state->f_l4));
memset (state->f_r4, 0.1f, sizeof(state->f_r4));
memset (state->f_l4, 0.0f, sizeof(state->f_l4));
memset (state->f_r4, 0.0f, sizeof(state->f_r4));
//zero out the short sample storage buffers
memset (state->s_l, 1, sizeof(state->s_l));
@@ -759,16 +759,16 @@ initState (dsd_state * state)
//Upsampled Audio Smoothing
state->audio_smoothing = 0;
memset (state->audio_out_temp_buf, 0.1f, sizeof(state->audio_out_temp_buf));
memset (state->audio_out_temp_bufR, 0.1f, sizeof(state->audio_out_temp_bufR));
memset (state->audio_out_temp_buf, 0.0f, sizeof(state->audio_out_temp_buf));
memset (state->audio_out_temp_bufR, 0.0f, sizeof(state->audio_out_temp_bufR));
//set float temp buffer to baseline
memset (state->f_l, 0.1f, sizeof(state->f_l));
memset (state->f_r, 0.1f, sizeof(state->f_r));
memset (state->f_l, 0.0f, sizeof(state->f_l));
memset (state->f_r, 0.0f, sizeof(state->f_r));
//set float temp buffer to baseline
memset (state->f_l4, 0.1f, sizeof(state->f_l4));
memset (state->f_r4, 0.1f, sizeof(state->f_r4));
memset (state->f_l4, 0.0f, sizeof(state->f_l4));
memset (state->f_r4, 0.0f, sizeof(state->f_r4));
//zero out the short sample storage buffers
memset (state->s_l, 1, sizeof(state->s_l));