Re: [mcx-users] Unexpected initial photon weights (detp.w0) when using pattern3D

[Fang, Qianqian]

Hi Jesse,

Can you disable normalization in your emission simulation? MCX by default normalizes all launched photon energy to 1 for all source types.

---

From: mcx-...@googlegroups.com <mcx-...@googlegroups.com> on behalf of Jesse Lam <jesse...@gmail.com>
Sent: Friday, July 24, 2026 12:22 PM
To: mcx-users <mcx-...@googlegroups.com>
Subject: [mcx-users] Unexpected initial photon weights (detp.w0) when using pattern3D

 

You don't often get email from jesse...@gmail.com. Learn why this is important

Hi Dr. Fang,

Thank you for continuing to develop MCX! I have been using it personally for years and have enjoyed diving deeper into the complexity.

I am currently interested to use the pattern3d emission source with an absorbed energy deposition map as the input. However, when testing, I was unable to match the output weights with initial weights in a simulation with no attenuation. I provided a script to replicate what I'm seeing:
- Run an initial energy simulation with no scattering to get a basic energy deposition map

- Use the deposition map for a "pattern3d" emission (input)

- Emit the light perfectly collimated and with no attenuation back to the detector (output)

From this setup, I expected the input = output, since there is no attenuation. I first checked whether sum(detp.w0) was equal to sum(eng.data(:)), which was my input. However, I did not see that. I noticed the following:

1. Calculating total photon weight from Reflectance does match well with the Launch weight.

2. However, sum(detp.w0) did not match the Launch weight and was also not consistent over radii despite the total input energy being unchanged.

I remain puzzled, and I was hoping you could shed some light on how detp.w0 is saved when using a pattern3d source. Any hints would be greatly appreciated.

Thank you.

Kind regards,

Jesse

clear;

%% Basic options

rad=.2:.2:1; %set of increasing disk radii

voxsize=.05; %voxel edge length (mm)

cubelen=4; %total simulation size (mm)

%% Building cfg for energy deposition

cfg.nphoton=1e6;

cfg.unitinmm=voxsize;

% Volume parameters

sx=cubelen./voxsize;

sy=cubelen./voxsize;

sz=cubelen./voxsize;

cfg.vol=uint8(ones(sx,sy,sz));

cfg.vol(:,:,1)=0;

cfg.isreflect=0;

cfg.isspecular=0;

cfg.seed=1;

for a=1:length(rad)

% source parameters

cfg.srctype='disk';

cfg.srcparam1(1)=rad(a)./voxsize;

cfg.srcparam1(2)=0;

cfg.srcpos=[sx/2 sy/2 1];

cfg.srcdir=[0 0 1];

% detector

cfg.detpos=[sx/2 sy/2 1 sqrt(sx^2+sy^2)/2]; %det size = top surface

cfg.tstart=0;

cfg.tend=1e-8;

cfg.tstep=1e-8;

cfg.outputtype='energy';

cfg.savedetflag='dpwxv';

% run sim

cfg.prop=[0 0 1 1; .1 0 1 1];

eng=mcxlab(cfg); % energy map to be used as input

%% Building cfg for emission

cfg.outputtype='fluence';

cfg.issaveref=1;

cfg.issaveexit=1;

cfg.isspecular=0;

% update source

launch_w0(a)=sum(eng.data(:)); %expected initial weight

cfg.srctype='pattern3d';

cfg.srcpattern=eng.data./max(eng.data(:)); %pattern3d input float [0 1]

cfg.srcpos=[1 1 1 launch_w0(a)]; %specify initial weight

cfg.srcdir=[0 0 -1]; %photons launch back to detector

cfg.srcparam1=[sx sy sz];

% run sim

cfg.prop=[0 0 1 1; 0 0 1 1]; %no attenuation

[flu,detp]=mcxlab(cfg);

% results

detected_weight(a)=sum(eng.data(:)).*sum(flu.dref(:).*cfg.unitinmm.^2);

w0(a)=sum(detp.w0);

end

%% Plotting

figure(1); clf;

subplot(3,1,1);

plot(rad,launch_w0);

ylabel('Weight');

title('Launched Photon Weight');

subplot(3,1,2);

plot(rad,detected_weight);

ylabel('Weight');

title('Detected Photon Weight');

subplot(3,1,3);

plot(rad,w0);

ylabel('Weight');

title('Sum of detp.w0');

--
You received this message because you are subscribed to the Google Groups "mcx-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to mcx-users+...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/mcx-users/5d4b53bc-3736-4075-a41b-b099b76dbc27n%40googlegroups.com.