C-compiler setup in MATLAB R2009a 64-bit version

I am trying to run an embedded matlab code in the 64-bit version of MATLAB R2009a and am getting a 'Make error' as shown below,

------------------------------------------------

Making simulation target "gammatoneFilter_sfun", ...

D:\00_Project\Matlab\Moi\v6\slprj\_sfprj\gammatoneFilter\_self\sfun\src>"d:\program files\matlab\r2009a\sys\lcc\bin\lccmake" -f gammatoneFilter_sfun.lmk The system cannot find the path specified.

---------------------------------------------

I understand this is be because the compiler is not installed using the 'mex -setup' command but on doing so unlike the 32-bit version there are no compilers listed to install.

1) I have both the 32 & 64 bit versions installed in D-drive and in separate folders D:/Program Files/Matlab_64bit & D:/Program Files(x86)/Matlab_32bit. Is this causing any problem?

2) Am I missing any package that helps install the 64-bit version of lcc/gcc? If so which one and where can I find it?

3) Since I received 'Simulink memory allocation error' when running the code in 32-bit version, I am trying to run the code in the 64-bit version. Is this right way? or should I write a better code & run it in 32-bit version?

Thanks in advance

Answer 1

[Kaustubha Govind]

I am not all too familiar with the Simulink memory allocation error (perhaps your model deals with large amounts of memory?), but I think attempting to run the model on 64-bit is definitely worth a try.

The free LCC compiler that 32-bit MATLAB ships is a 32-bit compiler only, and cannot generate 64-bit binaries - this is why 64-bit MATLAB does not ship with it. You can install MSVC 2008 Express which is a free compiler and is supported with R2009a (see full list here http://www.mathworks.com/support/compilers ). Please pay special attention to footnote#6 on the same page, to ensure that you install x64 compilers for MSVC 2008.

Once you have installed the compiler, run "mex -setup" and run your model again.

[Rajaram B S R]

Thank you very much for helping me out here. I did the installation and could list the compiler in the 64-bit version setup.

But there is a small problem, as I would be using the analog input block in my previous block which needs the data acquisition toolbox from the 32-bit version, I could not go ahead with this. Thanks anyways for the info.

Answer 2

[saikiran Puranam]

%% Startup and Globals

clear all

clc

%% Filter Parameters

width = 3840;

height = 2160;

n = 1000;

sig = 5;

% Read in the image

A = imread('canyon.jpg');

% Convert it to grayscale to make things slightly simpler

A = rgb2gray(A);

% Display it.

imshow(A);

% Convert the image to floating point numbers

A = double(A);

% Create a gaussian kernel (similar to a sinc wave in 2D, gives us

% lowpass characteristics

H = fspecial('gaussian',n, sig);

% Do the filtering via direct convolution (and time it with tic/toc)

tic

A_conv = conv2(A,H,'same');

toc

% Display the result

figure;

imshow(A_conv, []);

%% FFT Convolution

% Dot he filterting via FFT convolution (and time it with tic/toc)

tic

A_freq = fft2(A,height+n-1,width+n-1);

H_freq = fft2(H,height+n-1,width+n-1);

Out_freq = A_freq.*H_freq;

Out = ifft2(Out_freq);

toc

% Display the result

figure;

imshow(Out,[]);

[saikiran Puranam]

I have an error in this code in imshow(A); line

[saikiran Puranam]

in matlab 2020a version