3D MATLAB noise – effect of changing Gaussian convolution kernel size

gen_64x64x64_noise.m

i = 64;
for k = 1:2:31
    tic;
    noise{i,k} = noise3Dwrap(i,k);
    imwrite(reshape(noise{i,k},[i i 1 i]),...
        ['noise3Dwrap_s' num2str(i) 'k' num2str(k) '.gif'],...
        'LoopCount',Inf,'DelayTime',0.10);
    timeToProcess{i,k} = toc;
    save('noise3Dwrap64_k1_31.mat')
end

gen_64x64x64_noise_animations.m

i = 64;
for k = 1:2:31
    fftObj = vision.FFT;
     
    % Open the GIF for the current image
    theGIF = imread(['noise3Dwrap_s' num2str(i) 'k' num2str(k) '.gif'],'frames','all');
    for j = 1:size(theGIF,4)
        % Calculate 2D FFT
        % See: http://www.mathworks.com/help/vision/ref/vision.fftclass.html
        I = im2single(theGIF(:,:,:,j));
        J = step(fftObj, I);
        J_shifted = fftshift(J);
         
        % Display false-color image of 2D FFT
        figure(1);
        imshow(log(max(abs(J_shifted), 1e-6)),[]), colormap(jet(256));
         
        % Capture screen shot for GIF
        % See: http://www.mathworks.com/support/solutions/en/data/1-48KECO/index.html
        frame = getframe;
        im = frame2im(frame);
        [imind,cm] = rgb2ind(im,256);
        % Write frame to file
        if j == 1;
            imwrite(imind,cm,['noise3Dwrap2DFFT_s' num2str(i) 'k' num2str(k) '.gif'],...
                'gif', 'Loopcount',inf,'DelayTime',0.10);
        else
            imwrite(imind,cm,['noise3Dwrap2DFFT_s' num2str(i) 'k' num2str(k) '.gif'],...
                'WriteMode','append','DelayTime',0.10);
        end
    end