please can someone help me with running this script

Hi there, I am new to matlab and I having trouble running this script. I was designed the emulate the motion of a record groove but i can't seem to get it to work. I import the audio data and set the time vector as required but it just returns

Function definitions are not permitted in this context.

Please help

sig_left=data;

t=(0:0.0000441:30);

function [vertex, faces; vertex_color] = LPsim(t, sig_left; 

sig_right;track_skip; groove_depth, Rmax, rpm)

rpm=33.3333;

% 33 1/3 rpm industry standard

Rmax=0.1524e6;

% Maximum radius (at time t=0), in microns (6"=0.1524m)

groove_depth=50;

% Groove depth (in microns)

       track_skip=150; 

% Distance between tracks (in microns)

     sig_right=sig_left;

% Assume monoaural input

RPS=rpm/60;
% Rotational speed of record (revolutions per second)

% Some pre-computed constants

cos45=sqrt(2)/2;

sin45=sqrt(2)/2;

% Pre-allocate memory for large vectors (speeds performance)

nVerts=length(t)*3;

vertex=zeros(nVerts, 3);

vertex_color=zeros(nVerts, 1);

faces=zeros((length(t)-1)*2, 4);

% Construct verticies

% We are using the following conventions:

%   x: lateral position, from center of rotation

%   y: position on the lacquer, from center of rotation

%   z: depth, z=0 at the surface

%

% The record cutter revolves around on the (x-y) plane.

%

% Compute neutral stylus position

x0=(Rmax-RPS*t*track_skip).*cos(RPS*t*2*pi);

y=(Rmax-RPS*t*track_skip).*sin(RPS*t*2*pi);

% Mix in the signal

x=cos45*(sig_left+sig_right)+x0;

z=sin45*(sig_left-sig_right)-groove_depth;

% Transpose to place them in row order

x=x';

y=y';

z=z';

% Construct 3 vertices: one at the upper left corner of the V-groove,

% another in the bottom apex of the V-groove, and lastly, one for the

% upper right corner. Both upper corners have a height (z) of '0' by

% definition (where z=0 at the surface of the lacquer).

vertex(1:3:end-2, :)=[x-z, y, zeros(length(z), 1)];

vertex(2:3:end-1, :)=[x, y, z];

vertex(3:3:end, :)=[x+z, y, zeros(length(z), 1)];

% Color-code groove based on depth (leave all other vertices at '0'

% since they are set to be at the surface (hence the zeros(length(z))...)

vertex_color(2:3:end-1)=z;

% Connect verticies to form faces

% Each time point generates a triangle, and every 2 triangles generates a face

% Therefore, given 'n' time points, you will have (n-1)*2 faces

% It turns out that since the verticies are numbered in order, for a 3

% triangle setup you will have (where the verticies are numbered from left

% to right, front to back)

% [1 2 5 4]

% [2 3 6 5]

% [4 5 8 7]

% [5 6 9 8]

% [ . . . ] and so on

% Compute the number of faces

nFaces=(length(t)-1)*2;

% Determine the index for the first vertex of next-to-last triangle

ntl_vertex=(length(t)-1)*3-1;

% Vectorized output

faces(1:2:nFaces, :)=[1:3:ntl_vertex; 2:3:ntl_vertex+1; 5:3:ntl_vertex+4; 

4:3:ntl_vertex+3]';

faces(2:2:nFaces, :)=faces(1:2:nFaces-1, :)+1;

Thanks for your help

Answer 1

[Harold Bien]

Hi Leon,

It's hard to tell from the posted source, but a few things of note. First, you have lots of places where you used ';' when it's supposed to be ',':

For example: "function...(t, sig_left, sig_right,..." I'm not sure how MATLAB will interpret ";" in the argument list of the function.

Second: pay attention to dimensions of vectors as noted above by Leon. You have t defined to be a column vector - it will be of dimension 1xN where N is number of samples. Your signal is from another variable "data" which I do not know the layout of the vector (and I suspect it may be in row order, i.e. of dimension Nx1). From the comments, I can tell that x, y, and z are expected to be in column vector format (1xN) as I transpose the variables to yield the row vector (Nx1) in the lines "x=x'; y=y'; z=z';"

The line where the error occurs expects x, y, and z to be row vectors of dimension Nx1 as they are concatenated as columns.

For you convenience, I'll attach below the source code from the supplemental info but in electronic format (<http://www.rsc.org/suppdata/lc/b4/b409478f/b409478f.pdf)>. Note again that you can just run the function as-is without editing LPsim.m using the following command-line syntax:

[vertex, faces, vertex_color]=LPsim(t, data, data);

% Test case data=randn(1,100); t=0:1:99; [vertex, faces]=LPsim(t, data, data);

Finally, I do have to apologize for the difficulties in using the code. This was during my early days of MATLAB and I should have been more explicit about the required dimensions (1xN) of the input arguments.

[Edit: In the transfer, there was an inadvertent line-break in the comments which break LPsim.m which I fixed and re-uploaded].

Answer 2

[Geoff Hayes]

Leon - the error message is telling you have inserted some code before your function definition. Your above code is

 sig_left=data;
 t=(0:0.0000441:30);

 function [vertex, faces; vertex_color] = LPsim(t, sig_left; sig_right;track_skip; groove_depth, Rmax, rpm)

The sig_left and t are inputs to your function and so should not be pasted into the file that has your function defined. The only lines of code that can precede a function definition are comments.

What you need to do is remove those two lines of code so that the first line of your LPsim.m file is

function [vertex, faces; vertex_color] = LPsim(t, sig_left; sig_right;track_skip; groove_depth, Rmax, rpm)

Then, in the Command Window, type

 sig_left=data;
 t=(0:0.0000441:30);
 [vertex, faces; vertex_color] = LPsim(t,sig_left);

The remaining inputs appear to be initialized in the first few lines of the function and so could probably be removed.

Try the above and see what happens!

[Leon]

thanks for your help geoff, it has now has a new error

Error using +

Matrix dimensions must agree.

Error in LPsim (line 88)

x=cos45*(sig_left+sig_right)+x0;

I have used a mono audio file so sig_left and sig_right are identical i am confuse please help

[Geoff Hayes]

Leon - what are the dimensions of the arrays sig_left, sig_right, and x0? As x0 is initialized as

 x0=(Rmax-RPS*t*track_skip).*cos(RPS*t*2*pi);

and t is an array, I suspect that x0 is not of the same dimension as sig_left and sig_right.

[Leon]

Hi Geoff thanks again for you help. i am making progress but now i am getting this error

Dimensions of matrices being concatenated are not consistent.

Error in LPsim

vertex(1:3:end-2, :)=[x-z, y, zeros(length(z), 1)];

Thanks

leon

[Geoff Hayes]

Leon - you are trying to concatenate together matrices that do not have the same dimension. What are the dimensions (sizes) of x, z, and y?

[Geoff Hayes]

Leon's answer moved here.

They are all 1323696x1

[Geoff Hayes]

Leon - put a bepreakpoint at this line (the one that is throwing the error) and run your function. When the debugger pauses at this line, determine the size/dimension of each variable by typing the following in the command window

 size(x)
 size(y)
 size(z)

What do you see?