Error: Warning: The end of file was reached before the requested samples were read completely. Total number of samples read is 15138816.

I am getting an error which says that the file I am reading is too short to read the requested number of samples. However, I know this cannot be the case as I can check the length of the file itself by looking at it before processing. Further, I have inserted lines into my code which means it should only attempt to read the file if it is a certain length. Why is this happening?

%Input variables:
directory={'H:\SoundTrap\wav files\GoatIsland\008_GoatIsland_5280';}; 
tlo=1.0;
thi=299.0;
filelength='5min';

%% For loop to run through series of folders of .wav files:
for j=1:length(directory)
    
    clearvars -except j directory tlo thi filelength; %clear all variables between folders
    
    folder=char(directory(j));
    d=dir(fullfile(folder, '*.wav')); %list all .wav files in path folder
    files=length(d);                %number of files in folder
    
    output=[];      %create empty output array
    clear('A');     %clear output array between folders
    row=1;          %tool to store output data on new row of output matrix
                    %'A' at each iteration of loop

    %get sampling frequency of first (i.e. all) files in folder
    firstfile=fullfile(folder, char(d(1,1).name));
    [xbit,fs]=audioread(firstfile, [1 2]); 
    
   
%% Input variables (using fs from first file)  
    lcut=0;           %low frequency cutoff (will adjust later)
    hcut=72000;       %high freq cutoff
    
    %Select 2 mins or 5 mins of each .wav depending on data/purpose...
    %tlo=;          
    %thi=;                      
    nlo=fs*tlo;       %multiply fs by tlo to get starting point                    
    nup=fs*thi;       %do the same for end point  
    N=fs/1;           %segment length
    window=fs;        %frequency bin size
    overlap=window/2; %amount of overlap between segments (50%)
    nfft=fs;          %fft length
                        

    for i=1:files                           %for each file
        disp(d(i).name);                    %display filename
        filename=fullfile(folder, d(i).name); %get full filename
        wavinfo=audioinfo(filename);        %get file info
                 
    %Run PSD only if file is 2 mins long (1:59):
    %i.e. 119seconds*144000samples
        if (wavinfo.TotalSamples>=nup)     
        
        [xbit, fs]=audioread(filename, [nlo,nup]);
    %read in file
    
        xbit=detrend(xbit);
    %removes DC offset
    
        [spectra, f]=pwelch(xbit,window,overlap,nfft,fs); 
    %perform pwelch
    
        spectraout=10*log10(spectra); 
        clear('spectra');
    %convert output to dB by multiplying by 10*log10
    
        calcspectra=spectraout-S;
        clear('spectraout');
    %apply calibration (S)
    
        output(row,:) = calcspectra;
        clear('calcspectra');
    %fill successive rows with output from %pwelch
    
        row = row + 1; 
    %move onto next row of output matrix
     
    
    %If file is too short (rare):
        else
        %fill data row with zeros
        output(row,2:72001)=zeros;          %fill row with zeroes
        
        formatSpec=('Error: %s has an unexpected file length! Not processed\n');
        fprintf(2, formatSpec, filename);        
        
        row=row + 1;

        %list of filenames where rows were filled with zeroes is issued in
        %a separate _shortfiles.csv
        end
    end
    
    %code goes on to save output of each loop in .csv file...
end