How to read a mraw file

I have a video in .mraw and I would like to read it and exploit it but with the function I found, does not work because it is .cih and i have .cihx if someone could help me please

thank you in advance

[Mathilde Schneider]

Hi Lylia,

I'm facing the same issue, and using the readmraw doesn't seems to work. Have you managed to figure it out?

Regards,

Mathilde

[Phil Kreth]

Mathilde,

Please see my reply below. I edited SEP's readmraw.m function to handle .cihx files (as well as corner cases for newer .cih files).

Cheers,

Phil

Answer 1

[Shadaab Siddiqie]

From my understanding you want to read .mraw file. You can refer to the code bellow.

I=readmraw('Photron_mraw_example',[1 10]);
for n=1:1:10
imshow(I.Images.RawImages(:,:,n),[0 3000]);
pause(.1);
end

Here readmraw function is from Photron MRAW File Reader.

Answer 2

[Phil Kreth]

Hi everyone. I have edited the readmraw function that was written by "SEP" to handle either .cih or .cihx files. The originally function is posted on the FileExchange here - https://www.mathworks.com/matlabcentral/fileexchange/42408-photron-mraw-file-reader

Below is a copy of the new readmraw.m file that you can use.

function imgs = readmraw(filename, numImgs)
%   readmraw.m 
%   READMRAW Read Photron MRAW files into MATLAB
%   
%   imgs = READMRAW('C:\Photron\Filename.mraw', [a,b]) loads images a
%   through b from 'C:\Photron\Filename.mraw' into matrix imgs. 
%
%   Remarks
%   -------
%   This function must be handed the common *.cih(x) and *.mraw file name 
%   and the range of images to be loaded (MATLAB may not handle the entire
%   image range for large files). 
%   NOTE: Both the *.cih(x) file and the *.mraw file are utilized
%   Autor: SEP                                Creation Date: Jun 20, 2013
%   Editor: Phil Kreth                    Modification Date: Sep  9, 2024
%
%   Examples
%   --------
%   % Load all images
%   imgs = readmraw('C:\Photron\Moviefile.mraw', 0); 
% 
%   % Load images 10 through 50
%   imgs = readmraw('C:\Photron\Moviefile.mraw', [10,50]);
% 
%   % Load image 10
%   imgs = readmraw('C:\Photron\Moviefile.mraw', 10);
% 

fid1 = fopen(sprintf('%s.cih',filename(1:end-5)),'r');
if fid1 < 0
    fid1 = fopen(sprintf('%s.cihx',filename(1:end-5)),'r');
    cihx = true;
else
    cihx = false;
end
fid2 = fopen(sprintf('%s',filename),'r');
if fid1 < 1 || fid2 < 1
    error(['Could not locate .CIH or .CIHX header for file: ''' filename '''']);
end

if ~cihx % CIH FILE
    % Read Header Information
    Header = textscan(fid1,'%s','delimiter',':');
    Header = Header{1};

    color_ind = find(contains(Header, 'Color Type')) + 1;
    if strcmp(cell2mat(Header(color_ind(1))), 'Color')
        color = true;
    else
        color = false;
    end

    bit_ind = find(contains(Header, 'Color Bit')) + 1;
    bits = str2double(cell2mat(Header(bit_ind(1))));
    if color
        bits = bits/3;
    end
    bit_depth = sprintf('ubit%d', bits);

    frame_ind = find(startsWith(Header, 'Total Frame')) + 1;
    Total_Frames = str2double(cell2mat(Header(frame_ind(1))));

    width_ind = find(contains(Header, 'Image Width')) + 1;
    Width = str2double(cell2mat(Header(width_ind(1))));

    height_ind = find(contains(Header, 'Image Height')) + 1;
    Height = str2double(cell2mat(Header(height_ind(1))));

    % fps_ind = find(contains(Header, 'Record Rate(fps)')) + 1;
    % fps = str2double(cell2mat(Header(fps_ind(1))));
    
else % CIHX FILE
    % Read Header Information
    Header = textscan(fid1,'%s','delimiter',{'<','>'});
    Header = Header{1};

    color_ind = find(contains(Header, 'type')) + 1;
    if strcmp(cell2mat(Header(color_ind(1))), 'Color')
        color = true;
    else
        color = false;
    end

    bit_ind = find(contains(Header, 'bit')) + 1;
    bits = str2double(cell2mat(Header(bit_ind(1))));
    if color
        bits = bits/3;
    end
    bit_depth = sprintf('ubit%d', bits);

    frame_ind = find(contains(Header, 'totalFrame')) + 1;
    Total_Frames = str2double(cell2mat(Header(frame_ind(1))));

    width_ind = find(contains(Header, 'width')) + 1;
    Width = str2double(cell2mat(Header(width_ind(1))));

    height_ind = find(contains(Header, 'height')) + 1;
    Height = str2double(cell2mat(Header(height_ind(1))));

    % fps_ind = find(contains(Header, 'recordRate')) + 1;
    % fps = str2double(cell2mat(Header(fps_ind(1))));
end

Pixels = Width*Height;

fclose(fid1);


% Define Image Range
if numImgs == 0             % load all the images 
    first_frame = 1;
    frames = Total_Frames;
elseif length(numImgs) == 1 % load a single image
    first_frame = numImgs;
    frames = 1;
else                        % load a specified range of images
    first_frame = numImgs(1,1);
    last_frame = numImgs(1,2);
    frames = last_frame-first_frame+1;
end


% Load Images
bytes_offset = (first_frame-1)*Pixels*bits/8;
if color
    bytes_offset = bytes_offset*3;
end
fseek(fid2, bytes_offset, 'bof');

if bits > 8
	data_fmt = 'uint16';
else
	data_fmt = 'uint8';
end

if color
    imgs = zeros(Pixels*3, frames, data_fmt);
    for n = 1:frames
        imgs(:,n) = fread(fid2, Pixels*3, bit_depth, 0, 'b');
    end
    imgs = [imgs(1:3:end,:); imgs(2:3:end,:); imgs(3:3:end,:)]; % separate color channels
    imgs = reshape(imgs, [Width*Height 3 frames]); % reshape to separate color channels
    N = [Width Height 3 frames];
    imgs = permute(reshape(imgs, N), [2 1 3 4]); % standard reshape and permute
else 
    imgs = zeros(Pixels, frames, data_fmt);
    for n = 1:frames
        imgs(:,n) = fread(fid2, Pixels, bit_depth, 0, 'b');
    end
    N = [Width Height frames];
    imgs = permute(reshape(imgs, N), [2 1 3]);
end

fclose(fid2);

Cheers,

Phil

*Note: I updated this code on September 9, 2024 to handle an edge case that was found when trying to read all images with a version 3 CIH file. Sometimes, you may have encountered an error if you saved a couple copies of the MRAW and one of those had fewer frames than an earlier copy. Loading all frames with a 0 entered for the second argument could produce an error as the code would have looked for the larger number of frames instead of the correct one.

[Phillip]

Phil -- thanks for sharing this, you've saved me likely a few hours work!

Sincerely,

Another Phil.

[Gabriele Pisetta]

Amazing script! works flawlessly!

Thanks

[Jonny Cheng]

Hello Phil,

Thanks for your contributions on the scripts.

I have modified the code to read the cih file for a Photron SA-Z color camera. However, for the same camera with cihx file. The image shows wired color and lower resolution. Could you check my code and tell what is wrong?

function imgs = readmrawcihx2(filename, numImgs)

% readmraw.m

% READMRAW Read Photron MRAW files into MATLAB

%

% imgs = READMRAW('C:\Photron\Filename.mraw', [a,b]) loads images a

% through b from 'C:\Photron\Filename.mraw' into matrix imgs.

%

% Remarks

% -------

% This function must be handed the common *.cih(x) and *.mraw file name

% and the range of images to be loaded (MATLAB may not handle the entire

% image range for large files).

% NOTE: Both the *.cih(x) file and the *.mraw file are utilized

% Autor: SEP Creation Date: June 20,2013

% Editor: Phil Kreth Modification Date: Aug 31, 2021

%

% Examples

% --------

% % Load all images

% imgs = readmraw('C:\Photron\Moviefile.mraw', 0);

%

% % Load images 10 through 50

% imgs = readmraw('C:\Photron\Moviefile.mraw', [10,50]);

%

% % Load image 10

% imgs = readmraw('C:\Photron\Moviefile.mraw', 10);

%

fid1 = fopen(sprintf('%s.cih',filename(1:end-4)),'r');

if fid1 < 0

fid1 = fopen(sprintf('%s.cihx',filename(1:end-5)),'r');

fid2 = fopen(sprintf('%s.mraw',filename(1:end-5)),'r');

cihx = true;

else

fid1 = fopen(sprintf('%s.cih',filename(1:end-4)),'r');

fid2 = fopen(sprintf('%s.mraw',filename(1:end-4)),'r');

cihx = false;

end

%fid2 = fopen(sprintf('%s.mraw',filename(1:end-5)),'r');

if fid1 < 1 && fid2 < 1

error(['Could not locate .CIH or .CIHX header for file: ''' filename '''']);

end

if ~cihx % CIH FILE

% Read Header Information

Header = textscan(fid1,'%s','delimiter',':');

Header = Header{1,1};

color_ind = find(contains(Header, 'Color Type')) + 1;

if strcmp(cell2mat(Header(color_ind(1))), 'Color')

color = true;

else

color = false;

end

bit_ind = find(contains(Header, 'Color Bit')) + 1;

bits = str2double(cell2mat(Header(bit_ind(1))));

if color

bits = bits/3;

end

bit_depth = sprintf('ubit%d', bits);

frame_ind = find(contains(Header, 'Total Frame')) + 1;

Total_Frames = str2double(cell2mat(Header(frame_ind(1))));

width_ind = find(contains(Header, 'Image Width')) + 1;

Width = str2double(cell2mat(Header(width_ind(1))));

height_ind = find(contains(Header, 'Image Height')) + 1;

Height = str2double(cell2mat(Header(height_ind(1))));

% fps_ind = find(contains(Header, 'Record Rate(fps)')) + 1;

% fps = str2double(cell2mat(Header(fps_ind(1))));

else % CIHX FILE

% Read Header Information

Header = textscan(fid1,'%s','delimiter',{'<','>'});

Header = Header{1};

color_ind = find(contains(Header, 'type')) + 1;

if strcmp(cell2mat(Header(color_ind(1))), 'RawBayer')

color = true;

else

color = false;

end

bit_ind = find(contains(Header, 'bit')) + 1;

bits = str2double(cell2mat(Header(bit_ind(1))));

if color

bits = bits/3;

end

bit_depth = sprintf('ubit%d', bits);

frame_ind = find(contains(Header, 'totalFrame')) + 1;

Total_Frames = str2double(cell2mat(Header(frame_ind(1))));

width_ind = find(contains(Header, 'width')) + 1;

Width = str2double(cell2mat(Header(width_ind(1))));

height_ind = find(contains(Header, 'height')) + 1;

Height = str2double(cell2mat(Header(height_ind(1))));

% fps_ind = find(contains(Header, 'recordRate')) + 1;

% fps = str2double(cell2mat(Header(fps_ind(1))));

end

Pixels = Width*Height;

fclose(fid1);

% Define Image Range

if numImgs == 0 % load all the images

first_frame = 1;

frames = Total_Frames;

elseif length(numImgs) == 1 % load a single image

first_frame = numImgs;

frames = 1;

else % load a specified range of images

first_frame = numImgs(1,1);

last_frame = numImgs(1,2);

frames = last_frame-first_frame+1;

end

% Load Images

bytes_offset = (first_frame-1)*Pixels*bits/8;%

if color

bytes_offset = bytes_offset*3;

end

fseek(fid2, bytes_offset, 'bof');

if bits > 8

data_fmt = 'uint16';

else

data_fmt = 'uint8';

end

if color

imgs = zeros(Pixels*3, frames, data_fmt);

for n = 1:frames

imgs(:,n) = fread(fid2, Pixels*3, bit_depth, 0, 'b');

end

imgs = [imgs(1:3:end,:); imgs(2:3:end,:); imgs(3:3:end,:)]; % separate color channels

imgs = reshape(imgs, [Width*Height 3 frames]); % reshape to separate color channels

N = [Width Height 3 frames];

imgs = permute(reshape(imgs, N), [2 1 3 4]); % standard reshape and permute

else

imgs = zeros(Pixels, frames, data_fmt);

for n = 1:frames

imgs(:,n) = fread(fid2, Pixels, bit_depth, 0, 'b');

end

N = [Width Height frames];

imgs = permute(reshape(imgs, N), [2 1 3]);

end

fclose(fid2);

[Phil Kreth]

Jonny,

I know I'm super late with my reply here, but I can't find what's going on with your code unless I have an example file to work with. I don't have access to a color SA-Z, so it's challenging to guestimate what I might need to modify. Can you share a CIH(X) and MRAW with me? Anything that's got a couple of frames would be fine, and I would recommend uploading to a public resource like Google Drive.

Thanks,

Phil

[Phil Kreth]

So, I just found out what's going on. Apparently, the "type" can be more than just "Mono" or "Color". The other options are "Raw" and possibly "RawBayer". I was unaware of the "Raw*" formats until I came across one today. One of our 12-bit mono files got saved out of PFV4 with the "Raw" option checked under the MRAW options dialog. Reopening this file in PFV4 itself actually displayed the images incorrectly even.

I am not sure how to handle these formats as even PFV was not displaying them properly. If you come across this error, drag and drop your cihx file into MATLAB to view its contents. Not far down from the top, you will see lines similar to these:

<colorInfo>

<type>Mono</type>

<bit>12</bit>

</colorInfo>

The <type>XXXX</type> line is what I was referring to. The mrawreader() function is only expecting that type to be "Color" or "Mono". If your camera is a color camera, you can change "Raw" or "RawBayer" to "Color" and it will work.

Cheers!

Answer 3

[Turbulence Analysis]

Hi,

With this function, I am getting the below error

Unable to perform assignment because the size of the left side is 589824-by-1 and
the size of the right side is 17977-by-1.

Error in readmraw (line 126)
        imgs(:,n) = fread(fid2, Pixels, bit_depth, 0, 'b');
 

[Phil Kreth]

Hi there, Turbulence Analysis,

Sorry for the late reply. I don't monitor this site.

It looks like the code was anticipating a much larger image than what was being fed in through the fread() command. The interesting thing about the fread() output there being 17977x1 is that 17977 is a prime number, so there's no way it could be divided up (by bit depth, by color vs grayscale, or by any width & height combination). What were the characteristics of the file you were trying to feed in (W, H, Number of Frames, Bit Depth, Color vs Grayscale)?

Best,

Phil

[Clarice du Plessis]

Hi Turbulence Analysis

I got the same error when I specified the filename as .cihx. Changing it to .mraw solved the issue for me.

Regards

Clarice