xticklabel in the midle of the spectrogram

Hello all,

I have several .wav files with 3minutes each that I used to compute a spectrogram.

Now I want to put the name of each .WAV file in the xticklabels in the midle of the correspondent part of the espectrogram.

Does anyone have an idea how this can be done?

Thank you in advance.

Answer 1

[Voss]

Something like this maybe:

% some plot
plot(rand(1,1000));

% some file names:
name = {'a wav file.wav','another wav file.wav'};

% some region boundaries along the x-axis
x = [0 400 1000];
xline(x(2:end-1),'r','LineWidth',2); % so you can see the boundary/ies

% set xticks in the middle of each region, labeled according to 'name'
set(gca(),'XTick',(x(1:end-1)+x(2:end))/2,'XTickLabel',name);

[Ricardo Duarte]

Hello @Voss

This partially solves the questions...

But how can I set the boundaries automatically acording to the size of the wav. Let's say, at the midle of each wav.

Thanks

[Voss]

Post some data and code, and I'll see what I can do.

[Ricardo Duarte]

S=-174;
concatenatedAudio = [];
tic;
%Loop through each audio file
for i = 1:length(fileList)
    %Load the audio file
    [audio, Fs] = audioread(fileList(i).name);
    allDates{i} = fileList(i).date;
    %Concatenate the audio data
    concatenatedAudio = [concatenatedAudio; audio];
end

concatenatedAudio=single(concatenatedAudio);

%% Spectrogram calculation
% Example spectrogram parameters
if strcmp(windowType,'Hann')
    window = (0.5 - 0.5*cos(2*pi*(1:Fs)/Fs));%hann
elseif strcmp(windowType,'Hamming')
    w = (0.54 - 0.46*cos(2*pi*(1:Fs)/Fs)); %hamming
end

[S, F, T, P] = spectrogram(concatenatedAudio, window, [], [lowfreq:Fs/2] ,Fs);

% Plot the spectrogram
PdB=10*log10(abs(P'))-Si;
figure; surf(T,F,PdB'); view(2); shading flat;set(gca,'YScale','log')
axis xy;  % Invert Y-axis for better visualization
colormap('parula');  % Choose a colormap
title('Spectrogram of Combined Audio Files');
xlabel('Time (s)');
ylabel('Frequency (Hz)');
colorbar
ax = gca; % axes handle
ax.YAxis.Exponent = 0;
xlim([T(1) T(end)]);
ylim([F(1) F(end)]);
caxis([0 max(max(PdB))]);
ylabel(colorbar,['PSD [ dB re ' num2str(1) ' \muP'...
    'a^2 Hz^-^1 ]'],'fontname','arial','fontsize',14);
PSDTime=toc;
fprintf('PSD time elapsted: %s',num2str(PSDTime))
fprintf('s','\n')
fprintf('\n')

[Ricardo Duarte]

The wav files are too big to upload here

[Voss]

You can keep track of the size of each wav data set (the number of samples) when you read each file. That is the row vector n_samples below. Then the boundaries are x = cumsum([0 n_samples]/Fs);

Si=-174;
concatenatedAudio = [];

tic;

n_files = numel(fileList);
n_samples = zeros(1,n_files);

%Loop through each audio file
for i = 1:n_files
    %Load the audio file
    [audio, Fs] = audioread(fileList(i).name);
    
    % keep track of the size of each data set:
    n_samples(i) = size(audio,1);
    
    allDates{i} = fileList(i).date;
    %Concatenate the audio data
    concatenatedAudio = [concatenatedAudio; audio];
end

concatenatedAudio=single(concatenatedAudio);

%% Spectrogram calculation
% Example spectrogram parameters
if strcmp(windowType,'Hann')
    window = (0.5 - 0.5*cos(2*pi*(1:Fs)/Fs));%hann
elseif strcmp(windowType,'Hamming')
    window = (0.54 - 0.46*cos(2*pi*(1:Fs)/Fs)); %hamming
end

[S, F, T, P] = spectrogram(concatenatedAudio, window, [], [lowfreq:Fs/2] ,Fs);

% Plot the spectrogram
PdB=10*log10(abs(P'))-Si;
figure; surf(T,F,PdB'); view(2); shading flat;set(gca,'YScale','log')


% set the xticks and xticklabels
x = cumsum([0 n_samples]/Fs);
set(gca(),'XTick',(x(1:end-1)+x(2:end))/2,'XTickLabel',{fileList.name});


colormap('parula');  % Choose a colormap
title('Spectrogram of Combined Audio Files');
xlabel('Time (s)');
ylabel('Frequency (Hz)');
colorbar
ax = gca; % axes handle
ax.YAxis.Exponent = 0;
xlim([T(1) T(end)]);
ylim([F(1) F(end)]);
caxis([0 max(max(PdB))]);
ylabel(colorbar,['PSD [ dB re ' num2str(1) ' \muP'...
    'a^2 Hz^-^1 ]'],'fontname','arial','fontsize',14);
PSDTime=toc;
fprintf('PSD time elapsted: %s',num2str(PSDTime))
fprintf('s','\n')
fprintf('\n')