Case Study 7: Zemax LED Source To Diffuse Object Imaging

Goal

This case study demonstrates a more realistic source/object split than the specular proxy case:

• load a vendor Zemax .DAT LED source;

• route the LED through a 45 degree beam splitter;

• illuminate a true Diffuse Object surface;

• use guided Lambertian scatter to return light through the splitter;

• pass the return path through a simple BK7 imaging lens;

• inspect source throughput and camera-plane irradiance.

The screenshots in this tutorial are generated from the live Tk UI with:

python -m KrakenOS.UI.capture_zemax_led_diffuse_case_study_screenshots

Load The Layout

1. Start the UI with python -m KrakenOS.UI.layout_editor.

2. Choose Layouts -> Beam Splitters / Folds -> Zemax LED Beam-Splitter Imaging.

3. Keep Trace mode = Non-Sequential Preview.

4. Keep NS probabilistic coating split off.

The first table row is an Illumination Source scene row backed by the OSRAM green Zemax rayfile:

attachment/LED/rayfile_LSG_T676_20200827_Zemax/rayfile_LSG_T676_green_100k_20200827_Zemax.DAT

The Object row is reference geometry only. The useful optical chain is: LED source -> beam splitter -> diffuse object -> beam splitter return -> BK7 imaging lens -> Image plane.

The layout combines a vendor rayfile source, beam splitter, diffuse object, imaging lens, and detector plane.

Read The Scatter Paths

Click Update. The 2D plot uses physical path labels plus scatter branch rays. The diffuse object is the green vertical surface on the left. The blue curves are the imaging lens surfaces on the return path.

The return rays are not mirror reflections from an Object Target proxy. They are child branches spawned by the Diffuse Object row.

Use Path View

Open the table toolbar Path view dropdown and select:

Path 2: 45 deg 50/50 beam splitter to 45 deg 50/50 beam splitter via Diffuse object target

This isolates the source-to-object and object-return leg.

The Diffuse Object row is the physical scattering target.

Now select:

Path 5: 45 deg 50/50 beam splitter to Image plane via Splitter rear exit face, Imaging lens front, Imaging lens back

This isolates one useful image path through the splitter exit face and lens. Several path entries may terminate at the Image plane because diffuse scatter creates many deterministic child branch codes.

Path view remains useful even when the branch-code list is long.

Inspect Diffuse / BRDF Settings

Right-click the Diffuse Object row and choose Coating / Polarization -> Diffuse / BRDF Settings....

This preset uses the dependency-free built-in Lambertian model. Guided target surface is set to the splitter return aperture so the deterministic diffuse child rays can be traced through the useful camera branch.

Audit Source Illumination

Choose Actions -> Source Illumination Report. Set Target to:

3: Diffuse object target

The report shows that the vendor LED source reaches the diffuse object and reports source throughput and footprint size at the target.

Now change Target to:

6: Image plane

The Image-plane report verifies that source-driven diffuse return rays reach the detector after splitter and lens losses.

Run Image-Plane Analyses

Set:

Analysis path = All paths
Analysis surface = 6: Image plane
Detector bins = 64

Click DetMap and Update.

DetMap shows the Image-plane power distribution from the useful source-to-diffuse-object return rays.

Click Illum and Update.

Illum uses the same source-hit records to show image-plane illumination and centroid diagnostics.

Run The Python Example

The same layout has a scriptable example:

python KrakenOS/Examples/Examp_Zemax_LED_Beam_Splitter_Imaging.py

The script prints the rayfile source path and a per-ray trace summary, including whether each sampled LED ray hits the diffuse object and reaches the Image plane.

What This Proves

This case study combines several KrakenOS UI capabilities in one workflow:

• vendor Zemax rayfile source import;

• source/object separation;

• deterministic beam-splitter branch spawning;

• true Diffuse Object scattering instead of a specular proxy;

• guided Lambertian target sampling;

• path filtering for a branch-heavy non-sequential scene;

• detector power and source-illumination analysis at the Image plane.

Common Mistakes

I expected one clean return ray from the object.

A diffuse surface creates many child rays. This is expected and is the point of replacing the specular proxy with Diffuse Object.

The Analysis path dropdown contains many Rscatter entries.

Those are deterministic scatter branch codes. For the camera result in this tutorial, keep Analysis path = All paths and choose the Image plane as the analysis surface.

The Source panel ray count did not change the explicit rayfile source row.

This layout declares an explicit scene source in SETTINGS["scene_sources"]. Edit the source row or Scene Source Manager when changing that source.

I changed Guided target surface to None and fewer rays reached the camera.

Unguided diffuse sampling sprays rays over the scatter cone. Guided target sampling keeps the UI preview deterministic while still weighting rays with the selected scatter model.