soundsong1.m

% src/soundsong1.m

function wav = soundsong1(BPM,key,flag,song,fs,ADSR,har)
% soundsong(BPM,key,song,fs,ADSR,harmonics)
% 输入:
%   <float> BPM: beats per minute
%   <char> key: 歌曲的调
%   <int> flag: -1表示降半个音阶(?), 1表示升半个音阶(?)
%   <n-by-2 matrix> song: 乐谱, 第一列为音符唱名, 第二列为各音符持续拍数
%       休止符用-inf表示
%   <float> fs: 采样频率
%   <1-by-4 matrix> ADSR: adsr包络控制
%   <containers.Map> har: 各音对应各阶谐波分量幅度
% 返回值:
%   <row vector> wav: 合成的歌曲

tpb = 60/BPM;   % time per beat (seconds)
wav = [];       % intialize wav

attack = ADSR(1);
decay = ADSR(2);
sustain = ADSR(3);
release = ADSR(4);

for i = 1:size(song,1)
    t = 0:1/fs:(tpb*song(i,2));     % time sequence
    f = freqmap(key,song(i,1),flag);
    if ~isKey(har,song(i,1))
        harmonics = [1];    % default
    else
        harmonics = har(song(i,1));
    end
    w = harmonics*sin(2*pi*f*(1:length(harmonics)).'*t);  % add harmonics
    if song(i,1) == -inf
        wav = [wav, zeros(1,length(t))];
    else
        w = adsr(attack,decay,sustain,release,w,t);
        wav = [wav,w];
    end
end

wav = wav/max(wav);     % normalization

sound(wav,fs);

end