/* * Handle loading and decoding of sounds through SDL_sound. This file * is portable; actual playing is handled in RageSoundManager. * For small files, pre-decode the entire file into a regular buffer. We * might want to play many samples at once, and we don't want to have to decode * 5-10 mp3s simultaneously during play. * * For larger files, decode them on the fly. These are usually music, and there's * usually only one of those playing at a time. When we get updates, decode data * at the same rate we're playing it. If we don't do this, and we're being read * in large chunks, we're forced to decode in larger chunks as well, which can * cause framerate problems. * * Known problems: * I hear a click in one speaker at the beginning of some MP3s. This is probably * something wrong with my SDL_sound MAD wrapper ... * * TODO: * Rate (speed) * Configurable buffer sizes (stored in SoundManager) and so on * * We need (yet) another layer of abstraction: RageSoundSource. It'll just * implement the SDL_sound interface (in a class). Two implementations; * one, used normally, that just wraps SDL_sound; and another that is given * a list of sounds and time offsets and transparently mixes them together * at the given times, filling the gaps with silence. This should be an easy * way to handle autoplay tracks in keyed games. Normal background music can * be passed to it with an offset of 0 (or the gap, however it works out). */ #include "stdafx.h" #include "RageSound.h" #include "RageSoundManager.h" #include "RageUtil.h" #include "RageLog.h" #include "RageException.h" #include "RageTimer.h" #include "SDL_sound-1.0.0/SDL_sound.h" #ifdef _DEBUG #pragma comment(lib, "SDL_sound-1.0.0/lib/sdl_sound_static_d.lib") #else #pragma comment(lib, "SDL_sound-1.0.0/lib/sdl_sound_static.lib") #endif const int channels = 2; const int samplesize = 2 * channels; /* 16-bit */ const int samplerate = 44100; /* If a sound is smaller than this, we'll load it entirely into memory. */ const int max_prebuf_size = 1024*256; /* The most data to buffer when streaming. This should generally be at least as large * as the largest hardware buffer. */ const int internal_buffer_size = 1024*16; /* The amount of data to read from SDL_sound at once. */ const int read_block_size = 1024; RageSound::RageSound() { ASSERT(SOUNDMAN); LockMutex L(SOUNDMAN->lock); static bool initialized = false; if(!initialized) { if(!Sound_Init()) throw RageException( "RageSoundManager::RageSoundManager: error initializing sound loader: %s", Sound_GetError()); initialized = true; } stream.Sample = NULL; position = 0; playing = false; Loop = false; AutoStop = true; speed = 1.0f; stream.buf.reserve(internal_buffer_size); m_StartSeconds = 0; m_LengthSeconds = -1; /* Register ourselves, so we receive Update()s. */ SOUNDMAN->all_sounds.insert(this); } RageSound::~RageSound() { Unload(); /* Unregister ourselves. */ SOUNDMAN->all_sounds.erase(this); } RageSound::RageSound(const RageSound &cpy) { ASSERT(SOUNDMAN); LockMutex L(SOUNDMAN->lock); stream.Sample = NULL; full_buf = cpy.full_buf; big = cpy.big; AutoStop = cpy.AutoStop; m_sFilePath = cpy.m_sFilePath; m_StartSeconds = cpy.m_StartSeconds; m_LengthSeconds = cpy.m_LengthSeconds; Loop = cpy.Loop; position = cpy.position; playing = false; speed = cpy.speed; if(big) { /* We can't copy the Sound_Sample, so load a new one. * Don't bother trying to load it in a small buffer. */ stream.buf.reserve(internal_buffer_size); Load(cpy.GetLoadedFilePath(), false); } /* Register ourselves, so we receive Update()s. */ SOUNDMAN->all_sounds.insert(this); } void RageSound::Unload() { if(IsPlaying()) Stop(); Sound_FreeSample(stream.Sample); stream.Sample = NULL; m_sFilePath = ""; stream.buf.clear(); full_buf.erase(); } void RageSound::Load(CString sSoundFilePath, bool cache) { LOG->Trace( "RageSound::LoadSound( '%s' )", sSoundFilePath.GetString() ); Unload(); m_sFilePath = sSoundFilePath; Sound_AudioInfo sound_desired; sound_desired.channels = channels; sound_desired.format = AUDIO_S16SYS; sound_desired.rate = samplerate; Sound_Sample *NewSample = Sound_NewSampleFromFile(sSoundFilePath.GetString(), &sound_desired, read_block_size); if( NewSample == NULL ) throw RageException( "RageSound::LoadSound: error loading %s: %s", sSoundFilePath.GetString(), Sound_GetError() ); /* Try to decode into full_buf. */ big = false; if(!cache) big = true; /* Don't. */ /* Check the length, and see if we think it'll fit in the buffer. */ { int len = Sound_Length(NewSample); /* This will fail with EAGAIN if it'll take a while. We only want * to do this if it's fast. */ if(len != -1) { float secs = len / 1000.f; int pcmsize = int(secs * samplerate * samplesize); /* seconds -> bytes */ if(pcmsize > max_prebuf_size) big = true; /* Don't bother trying to preload it. */ else full_buf.reserve(pcmsize); } Sound_Rewind(NewSample); } while(!big) { int cnt = Sound_Decode(NewSample); if(cnt < 0) throw RageException("Read error on %s: %s", sSoundFilePath.GetString(), Sound_GetError() ); /* XXX (see other error-handling XXX) */ /* Add the buffer. */ full_buf.append((const char *)NewSample->buffer, (const char *)NewSample->buffer+cnt); if(full_buf.size() > max_prebuf_size) { full_buf.erase(); big = true; /* too big */ } if(NewSample->flags & SOUND_SAMPLEFLAG_EOF) break; } if(big) { /* Oops; we need to stream it. */ stream.Sample = NewSample; Sound_Rewind(stream.Sample); } else { /* We're done with the stream. */ Sound_FreeSample(NewSample); } position = 0; } void RageSound::SetStartSeconds( float secs ) { m_StartSeconds = secs; } void RageSound::SetLengthSeconds(float secs) { m_LengthSeconds = secs; } /* Start playing from m_StartSeconds. TODO: a way to start playing * from the current pos (unpause) */ void RageSound::Play() { LockMutex L(SOUNDMAN->lock); if(playing) Stop(); playing = true; SetPositionSeconds(m_StartSeconds); // Tell the sound manager to start mixing us SOUNDMAN->StartMixing(this); } void RageSound::Update(float delta) { if(playing && big) stream.FillBuf(int(delta * samplerate * samplesize)); } /* Fill the buffer by about "bytes" worth of data. (We might go a little * over, and we won't overflow our buffer.) */ int RageSound::stream_t::FillBuf(int bytes) { ASSERT(Sample); bool got_something = false; if(Sample->flags & SOUND_SAMPLEFLAG_EOF) return got_something; /* EOF */ while(bytes > 0) { if(buf.size()+read_block_size > buf.capacity()) break; /* full */ int cnt = Sound_Decode(Sample); if(Sample->flags & SOUND_SAMPLEFLAG_EOF) return got_something; /* EOF */ if(Sample->flags & SOUND_SAMPLEFLAG_ERROR) { /* There was a fatal error; get it with Sound_GetError(). * XXX: How should we handle sound errors? We can't * just return error, since we're in a separate thread. * Most of the time we should probably just warn and move * on (no big deal), but the gameplay screen should query * periodically and do something more intelligent when we * fail (so we don't play out the rest of the song in * silence) ... */ throw RageException("Read error: %s", Sound_GetError() ); } buf.write((const char *)Sample->buffer, cnt); bytes -= cnt; got_something = true; } return got_something; } /* Called by the mixer: return a block of sound data. * Be careful; this is called in a separate thread. */ int RageSound::GetPCM(char *buffer, int size, int sampleno) { LockMutex L(SOUNDMAN->lock); /* If the sound is paused, just fill the buffer with silence. * Hmm. Pausing is annoying, since we'll get startup latency if * we keep our buffer playing; we should stop the stream completely * when we pause ... */ ASSERT(playing); int bytes_stored = 0; pos_map[0] = pos_map[1]; pos_map[1] = pos_map[2]; pos_map[2] = pos_map[3]; pos_map[3].clear(); /* "sampleno" is the audio driver's conception of time. "position" * is ours. Keep track of sampleno->position mappings for two GetPCM calls. * * This way, when we query the time later on, we can derive position * values from the sampleno values returned from GetPosition. * * We need to keep two buffers worth of values, since we might loop at * the end of a buffer. */ /* Now actually put data from the correct buffer into the output. */ while(size) { int got; if(position < 0) { /* We havn't *really* started playing yet, so just feed silence. How * many more bytes of silence do we need? */ got = -position * samplesize; got = min(got, size); memset(buffer, 0, got); } else if(big) { /* Feed data out of our streaming buffer. */ ASSERT(stream.Sample); got = min(int(stream.buf.size()), size); stream.buf.read(buffer, got); } else { /* Feed data out of our full buffer. */ int byte_pos = position * samplesize; got = min(int(full_buf.size())-byte_pos, size); got = max(got, 0); if(got) memcpy(buffer, full_buf.data()+byte_pos, got); } if(!got) { /* We need more data. Find out if we've hit EOF. */ bool HitEOF = true; if(big) { /* If we don't have any data left buffered, fill the buffer by up to * as much as we need. */ if(stream.buf.size() || stream.FillBuf(size)) HitEOF = false; /* we have more */ } else { unsigned byte_pos = position * samplesize; /* samples -> bytes */ if(byte_pos < full_buf.size()) HitEOF = false; /* we have more */ } /* If we've passed the stop point (m_StartSeconds+m_LengthSeconds), pretend * we've hit EOF. */ if(m_LengthSeconds != -1 && float(position)/samplerate > m_StartSeconds+m_LengthSeconds) HitEOF = true; if(!HitEOF) continue; /* We're at EOF. If we're not looping, just stop. */ if(Loop) { /* Rewind and start over. */ SetPositionSeconds(m_StartSeconds); continue; } if(AutoStop) break; /* We're out of data, but we're not going to stop, so fill in the * rest with silence. */ memset(buffer, 0, size); got = size; } /* Save this sampleno/position map. */ pos_map[3].push_back(pos_map_t(sampleno, position, got/samplesize)); int got_samples = got / samplesize; /* bytes -> samples */ /* This block goes from position to position+got_samples. */ const float FADE_TIME = 1.5f; /* XXX: Loop shouldn't set fading; add a Fade_Time member? * * We want to fade when there's FADE_TIME seconds left, but if * m_LengthSeconds is -1, we don't know the length we're playing. * (m_LengthSeconds is the length to play, not the length of the * source.) If we don't know the length, don't fade. */ if(Loop && m_LengthSeconds != -1) { Sint16 *p = (Sint16 *) buffer; int this_position = position; for(int samp = 0; samp < got_samples; ++samp) { float fSecsUntilSilent = (m_StartSeconds + m_LengthSeconds) - float(this_position)/samplerate; float fVolPercent = fSecsUntilSilent / FADE_TIME; fVolPercent = clamp(fVolPercent, 0.f, 1.f); for(int i = 0; i < channels; ++i) { *p = short(*p * fVolPercent); p++; } this_position++; } } bytes_stored += got; position += got_samples; size -= got; buffer += got; sampleno += got_samples; } return bytes_stored; } /* After we finish via the GetPCM return value, this is called by * RageSoundManager once the sound is actually flushed, which means * we're *really* stopped. */ void RageSound::SoundStopped() { LOG->Trace("stop completed"); Stop(); } void RageSound::Pause() { if(!playing) return; // Tell the sound manager to stop mixing this sound. SOUNDMAN->StopMixing(this); playing = false; } void RageSound::Stop() { /* Tell the sound manager to stop mixing this sound. */ SOUNDMAN->StopMixing(this); if(big) { ASSERT(stream.Sample); Sound_Rewind(stream.Sample); stream.buf.clear(); } playing = false; position = 0; for(int i = 0; i < 4; ++i) pos_map[i].clear(); } float RageSound::GetLengthSeconds() { if(big) { ASSERT(stream.Sample); int len = Sound_Length(stream.Sample); if(len == -1 && stream.Sample->flags & SOUND_SAMPLEFLAG_EAGAIN) { /* This indicates the length check will take a little while; call * it again to confirm. */ len = Sound_Length(stream.Sample); } if(len < 0) return -1; /* XXX: put a Sound_GetError() error message somewhere */ return len / 1000.f; /* ms -> secs */ } else { /* We have the whole file loaded; just return the position. */ return full_buf.size() / (float(samplerate)*samplesize); } } float RageSound::GetPositionSeconds() const { LockMutex L(SOUNDMAN->lock); /* If we're not playing, just report the static position. */ if( !IsPlaying() ) return position / float(samplerate); /* If we don't yet have any position data, GetPCM hasn't yet been called at all, * so report the static position. */ { bool HasData = false; for(int i = 0; i < 4; ++i) { if(!pos_map[i].empty()) HasData = true; } if(!HasData) { LOG->Trace("no data yet; %i", position); return position / float(samplerate); } } /* Get our current hardware position. */ int cur_sample = SOUNDMAN->GetPosition(this); /* Concatenate the pos_maps. (Just a cheat to simplify this a little; this * is small, so this isn't very expensive.) */ vector posmaps; for(int n = 0; n < 4; ++n) posmaps.insert(posmaps.end(), pos_map[n].begin(), pos_map[n].end()); /* sampleno is probably in one of the pos_maps. Search through them * to figure out what position this sampleno maps to. */ int closest_position = 0, closest_position_dist = INT_MAX; for(unsigned i = 0; i < posmaps.size(); ++i) { if(cur_sample >= posmaps[i].sampleno && cur_sample < posmaps[i].sampleno+posmaps[i].samples) { /* cur_sample lies in this block; it's an exact match. Figure * out the exact position. */ int diff = posmaps[i].position - posmaps[i].sampleno; return float(cur_sample + diff) / samplerate; } /* See if the current position is close to the beginning of this block. */ int dist = abs(posmaps[i].sampleno - cur_sample); if(dist < closest_position_dist) { closest_position_dist = dist; closest_position = posmaps[i].position; } /* See if the current position is close to the end of this block. */ dist = abs(posmaps[i].sampleno + posmaps[i].samples - cur_sample); if(dist < closest_position_dist) { closest_position_dist = dist + posmaps[i].samples; closest_position = posmaps[i].position + posmaps[i].samples; } } /* The sample is out of the range of data we've actually sent. * Return the closest position. * * There are three cases when this happens: * 1. After the first GetPCM call, but before it actually gets heard. * 2. After GetPCM returns EOF and the sound has flushed, but before * SoundStopped has been called. * 3. Underflow; we'll be given a larger sample number than we know about. */ return closest_position / float(samplerate); } void RageSound::SetPositionSeconds( float fSeconds ) { LockMutex L(SOUNDMAN->lock); position = int(fSeconds * samplerate); if( fSeconds < 0 ) fSeconds = 0; if(big) { ASSERT(stream.Sample); Sound_AccurateSeek(stream.Sample, int(fSeconds * 1000)); stream.buf.clear(); } } void RageSound::SetPlaybackRate( float fScale ) { LockMutex L(SOUNDMAN->lock); speed = fScale; } /* This is used to start music. It probably belongs in RageSoundManager. */ void RageSound::LoadAndPlayIfNotAlready( CString sSoundFilePath ) { SOUNDMAN->lock.Lock(); if( GetLoadedFilePath() == sSoundFilePath && IsPlaying() ) return; // do nothing SOUNDMAN->lock.Unlock(); Load( sSoundFilePath ); SetLooping(); Play(); } void CircBuf::reserve(unsigned n) { clear(); buf.erase(); buf.insert(buf.end(), n, 0); } void CircBuf::clear() { cnt = start = 0; } void CircBuf::write(const char *buffer, unsigned buffer_size) { ASSERT(size() + buffer_size <= capacity()); /* overflow */ while(buffer_size) { unsigned write_pos = start + size(); if(write_pos >= buf.size()) write_pos -= buf.size(); int cpy = min(buffer_size, buf.size() - write_pos); buf.replace(write_pos, cpy, buffer, cpy); cnt += cpy; buffer += cpy; buffer_size -= cpy; } } void CircBuf::read(char *buffer, unsigned buffer_size) { ASSERT(size() >= buffer_size); /* underflow */ while(buffer_size) { unsigned total = min(buf.size() - start, size()); unsigned cpy = min(buffer_size, total); buf.copy(buffer, cpy, start); start += cpy; if(start == buf.size()) start = 0; cnt -= cpy; buffer += cpy; buffer_size -= cpy; } }