Help with plotting histogram from audio file

Hi! I've recently joined a physics lab and am trying to write some code. I'm looking to create a histogram of pulse heights at a given time interval but I'm getting the following. I'm new to MatLab and don't quite understand what's going wrong. Does anyone think that they might be able to help? Here are pictures and the code. Thanks!

%Height Matrix
function [dataMatrix] = HeightWithFrequency(audioFileLocation, secondsPerInterval, t)
sampleInfo = audioinfo(audioFileLocation);
%sample rate in hz
sampleRate = sampleInfo.SampleRate;
intervalSize = sampleRate*secondsPerInterval;
%total length of sample in seconds
te = t + intervalSize;
%end time included in the histogram ( includes range of data from t to te)
start = floor(t*sampleRate);
%starting sample number
finish = ceil(te*sampleRate);
%ending sample number
[y, ~] = audioread(audioFileLocation, [start, finish]);
%assigns y to be the first channel of the audioread
dataMatrix = findpeaks(y, 'MinPeakProminence', 0.06);
%matrix of peaks
dataMatrix = dataMatrix';
end

longrun = '/Users/.../.../.../.../.../fileName.flac';
time = 300;
interval = 30;
sizes = HeightWithFrequency(longrun, interval, time);
histogram(sizes, 30)
xlabel('size (units)')
ylabel('frequency')
title(['Frequency of cell sizes at time', num2str(time)]);

[Walter Roberson]

Your code is not account for the possibility that te might be after the end of the audio.

[AZ0]

Thanks so much for the feedback! Just edited it so that it takes the minimum between the total length of the sample and the finish time but it's still giving me issues. Do you know if this makes sense?

te = t + intervalSize;
%end time included in the histogram ( includes range of data from t to te)
start = floor(t*sampleRate);
%starting sample number
len = floor((sampleInfo.Duration * sampleRate)/2);
%total length of sample in seconds
finish = min(ceil(te*sampleRate), len);
%ending sample number

The new errors are:

Error using findpeaks
Expected Y to be a vector.

Error in findpeaks>parse_inputs (line 201)
validateattributes(Yin,{'double','single'},{'nonempty','real','vector'},...

Error in findpeaks (line 135)
  = parse_inputs(isInMATLAB,Yin,varargin{:});

[Walter Roberson]

Every time you calculate something using sample rates and so on then you have difficulties with uncertaintiies and rounding problems. You should be using the TotalSamples propertly in your min(). Note that the range to be passed is in samples not time.

Answer 1

[Sandeep Mishra]

Hi AZ0,

I encountered a similar error using the ‘findpeaks’ function while running the code snippet in MATLAB R2024a.

Upon debugging the code, I discovered that the output from the ‘audioread’ function returns an m*n matrix, where ‘m’ represents the number of audio samples and ‘n’ indicates the number of audio channels present in the file.

To determine the number of channels in your audio file, you can utilize the ‘sampleInfo’ variable as shown below:

% Channel Info
numberOfChannels = sampleInfo.NumChannels;

Since the ‘findpeaks’ function takes a vector as input, a suitable solution is to extract peaks from a specific audio channel. You can use the following code snippet to obtain peaks from the first audio channel:

dataMatrix = findpeaks(y(:,1), 'MinPeakProminence', 0.06);

For more information, refer to the following MathWorks documentation:

1. ‘audioread’ function: https://www.mathworks.com/help/releases/R2024a/matlab/ref/audioread.html#btiabil-1-y
2. ‘audioinfo’ function: https://www.mathworks.com/help/releases/R2024a/matlab/ref/audioinfo.html#:~:text=character%20vector-,NumChannels,-Number%20of%20audio
3. ‘findpeaks’ function: https://www.mathworks.com/help/releases/R2024a/signal/ref/findpeaks.html#bufbbs1-data

I hope this helps you in resolving the issue.