The goal of this example is to identify the Narcissus effect in a thermal camera operating in the infrared spectrum. The Narcissus effect occurs when a detector inside an infrared system sees a reflection of itself, resulting in a dark spot in the center of the image. Additionally, radiation reflected from each surface of the optical system also returns to the detector, affecting the image quality of the system. Therefore, studying this effect using Ansys Speos software is of interest, as it enables the visualization of the resulting image and the contribution of each optical component to the image quality, essential for the design of these type of optical devices.
[[NOTES:]] Software Prerequisites
Narcissus effect can be modeled in any Ansys Speos version of the software. However, to be able to follow this example, the following version of the tool needs to be installed on your computer:
- Ansys Speos 2024 R2 or later versions.
Overview
Understand the simulation workflow and key results
The Narcissus effect is the main source of noise that thermal imaging systems, such as IR cameras, can face, affecting their global performance. This effect has its cause in the difference of temperatures between the detector (lower temperature) and the rest of the system. Due to the detector sensitivity to thermal variations, when internal reflections occur in the propagation of light through an optical system, the own image of the detector is going to be reflected appearing in the center of the image, highlighting its contribution above those of other components as well as that of the observed object itself.
The main contribution to the image of the Narcissus effect can be solved, in most cases, by the selection of suitable antireflection coatings.
In this example we will be able to reproduce the Narcissus effect in an IR camera, analyzing the contribution of the sensor itself in comparison to the rest of the optical elements. For this purpose, the thermic source type available in Speos would be applied to each one of the elements of the camera.
The workflow above illustrates the step-by-step process for analyzing the Narcissus effect in an IR camera using Ansys Speos.
- Defining thermic sources: the correct definition of the thermic sources in the optical system is crucial when visualizing the Narcissus effect. Each component of the IR-camera of our example will have an associated thermic source, defined by a temperature value. These sources are responsible for the Narcissus effect appearing in the sensor plane.
- Setting up the inverse simulation: after defining the sources, the remaining elements for computing an inverse simulation will be generated. In addition to the sensor and the selection of geometries, local meshing will be applied to the four lenses comprising the IR camera.
- Simulation results. Image quality analysis: the irradiance map generated after computing the inverse simulation will enable us to visualize the Narcissus effect. Moreover, thanks to the layer options, we will analyze each source's contribution separately.
-
Simulation results. Ray path analysis: thanks to the Light Expert tool available when defining a simulation, we will be able to finish our analysis of this effect with the visualization of the ray path inside the IR camera.
This article will only cover the visualization of the Narcissus effect in Ansys Speos software, analyzing this effect considering both lens and mounting systems. For more details about the analysis of Narcissus effect only focusing on lens system in Ansys Zemax OpticStudio, you can consult OpticStudio Narcissus analysis macro.
Run and Results
Instructions for running the model and discussion of key results
[[NOTES:]] Associated files
In the downloadable data, you will find two main folders: START and END . In this example. we will follow the step-by-step instructions using the data available in the START folder. You can find the complete project in the END folder, where you will only need to run the simulation to obtain the final results.
Step 1: Defining thermic sources.
To be able to run a simulation in Ansys Speos to analyze the Narcissus effect, the initial step is to define the sources and sensors to be included.
In this example, it is particularly important to define thermic sources for each component of the IR camera as well as for the test-chart we are using as an object. Ideally, we would like to see only the test-chart in our simulation results, but in the reality the camera system components emits also light in far infrared, so we need to define them as thermic sources too, in order to include them all in the simulation.
For this purpose, each thermic source will have a different temperature definition depending on the corresponding component. By doing this, we will be able to simulate all the contributions from different elements that are going to be detected by our sensor. For more information about the definition and specification of thermic sources in Speos we recommend Far infrared with Ansys Speos software course available on Ansys Learning Hub . You can also consult Creating a Thermic Source - Speos User's guide .
In the table below, the temperatures used in this example are shown. These values can be modified to adjust to different use cases. The first five components belong to the IR camera system, while the last one, Thermic object source is referring to the test-chart we are using as an object.
Component | Temperature |
Thermic objective mount source | 290k |
Thermic rings lenses source | 287K |
Thermic occulting baffle source | 285K |
Thermic sensor baffle source | 280K |
Thermic sensor source | 270K |
Thermic object source (with vertical lines as test-chart) |
275K |
- Open Ansys Speos.
- Open IR_camera-Narcissus_effect.scdocx in the downloaded data (Narcissus_effect_on_IR_camera> START).
- In the Light Simulation tab > Sources, select Thermic. A new thermic source should be appearing on the simulation panel. Double click on the Thermic source to open the definition panel:
- Change the name of the source to “Thermic objective mount”.
- Change the Flux Type to Radiant.
- In Emittance, modify the temperature to the value corresponding to the objective mount (check the table available at the beginning of this section).
- In the 3D view, on the left corner, click on the icon and select IR-CAMERA_MOUNT on the structure tree. The objective mount should be highlighted in the 3D view. Then, validate the active selection. All the selected geometries (26 in total) should appear in the definition panel > Emissive faces.
- Click on compute to visualize the result for the source definition.
- Create another five thermic sources, one for each one of the elements of the IR camera. These components, along with their corresponding temperatures, are specified in the table at the beginning of this section. The definition of the rest of the parameters for these thermic sources should be the same as the one already created for the objective mount.
[[NOTE:]] For the thermic source definition for the object, only the surface in the background must be selected as an emissive face. In addition, the reverse normal must be selected to ensure that the source is emitting in the direction of the IR camera.
Once all the thermic sources have been defined for this IR camera example, the following step would be to set the rest of the simulation parameters and geometries.
Step 2: Setting up the inverse simulation.
Sensor definition
In this example, an inverse simulation will be set up. Additionally, to the six thermic sources already defined in the previous step, an irradiance sensor should also be defined.
- In the Light Simulation Tab Sensors, click on Irradiance sensor.
- In the 3D view, on the left corner, click on the icon for origin of the axis system and then, click on the Structure tree >IR-CAMERA_SENSOR > Sensor origin.
- For the X and Y direction, keep the default selection. For the integration direction, select the Z axis and set the reverse direction to True to align it with the light's propagation through the system.
- In the sensor's definition panel, change the type to spectral and select Source as the layer type to filter visualization results by the created thermal sources.
- Set the X and Y range and the Wavelength parameters as shown in the image below.
Local meshing
To obtain more accurate simulation results, we will apply local meshing parameters to the four lenses in this example. This process is necessary to model with precision the lens shape during simulation. To simplify this process, we will first create a group containing all the surfaces where the local meshing will be defined. By doing this, we will achieve greater accuracy on the lens surfaces by defining a finer meshing, while other geometries, such as the mechanical components of the system, will be simulated with a less precise meshing.
For more information about Meshing in Ansys Speos we recommend Speos Meshing Best Practices .
- In the 3D view, hide all the IR camera components except for the four lenses.
- While holding the Ctrl key, start selecting in the 3D view all the surfaces with curvature of each lens (2 per lens, see the image below). If you have difficulties selecting some of the hidden faces, you can temporarily hide some of the other lenses to simplify the selection process. Another option is to hold Ctrl, select one face, and then scroll with the mouse wheel. This will move the selection through all the visible faces of the geometries, allowing you to select them. With all the surfaces selected, go to the Groups panel and click on the “Create NS” icon. A new group will appear. Rename it as “Lenses”.
- Back in the Light Simulation tab, in the Geometry properties section, click on Local Meshing. A new icon should appear on the simulation panel.
- Define the parameters for the local meshing for the lenses as shown.
- Click on the icon in the left corner of the 3D view and select the Lenses group previously created. The local meshing will be applied only to these faces.
- Finally, to be able to see the local meshing in the 3D view, right-click on the selected geometries in the definition panel, then click on Preview Meshing.
Defining the inverse simulation
For the visualization of the Narcissus effect in our IR camera, we are going to define an inverse simulation.
- On the Light Simulation tab, on the Simulation area, click on Inverse simulation.
- On the Simulation panel, double click on the Inverse.1 simulation generated to modify its parameters as shown. Rename the simulation as Inverse.narcissus to be able to identify it. Don't forget to enable the Light Expert option changing the value to "True", leaving the LPF max path parameter as default.
- Right click on the inverse simulation and go to options. In the inverse simulation tab, check the Dispersion option.
- On the left corner of the 3D view, click on the icon to select the geometries to be included in the simulation. From the structure tree, select: IR-CAMERA_MOUNT, IR-CAMERA_LENS1, IR-CAMERA_LENS2, IR-CAMERA_LENS3, IR-CAMERA_LEN4, IR-CAMERA_RINGS, IR-CAMERA_SENSOR, IR-CAMERA_OBJECT_SOURCE, IR-CAMERA_PROTECT-GLASS. Then, verify the selection. All 31 geometries should appear in the definition panel of the inverse simulation.
- Replicate the selection process with the six thermic sources previously created and with the irradiance sensor.
- Click on compute and wait for simulation results.
The time required to compute the inverse simulation, which needs to be run on CPU (it's not supported on GPU), depends on the hardware specifications. In this example, obtaining the results shown in step 3 and 4 took on CPU approximately 2 hours on 20 cores , with all parameters for the simulation defined in the previous steps, which we highly recommend for accurate results. However, several actions can be taken to reduce the computing time if needed. These are describe in the Important model settings section of this article.
Step 3: Simulation results. Image quality analysis.
Once the simulation is completed, two main files will be generated: an .html report and an .xmp map. The first file contains all the information related to the simulation's performance and analysis, including computing analysis, percentage of error, and other specifications. In this example, we will focus on the Inverse.narcissus.Irradiance.xmp file, which allows us to visualize the Narcissus effect on the sensor plane in terms of irradiance.
- Open the .xmp map generated.
- In the View tab, the photometric units option is selected by default. Change it to radiometric units to be able to see infrared wavelength range.
- In the color map tab, True Color is selected. Change the color map to one of your preferences. In this example, we recommend Black to white (custom color) for a better visualization of the results.
- Open the level scale and adjust the maximum and minimum value for irradiance until the results are visible in the irradiance map. For the visualization of all wavelengths and layers we recommend irradiance values between 75-80 W/ m² , but these settings may need to be adjusted depending on the type of color map selected.
The irradiance map allows the visualization of the narcissus effect in the IR camera of this example. The image below shows that, due to this effect, the sensor does not receive the expected uniform irradiance signal from the object source. Instead, both the center and corners of the image display contributions from other elements of the IR camera.
The source layers defined in the irradiance sensor allow us to visualize the contribution of each source to the image separately. This enables a comprehensive study of the amount of energy provided by each element of the IR camera, identifying the main contributors to the Narcissus effect.
The images below display the irradiance map of the different sources defined in this example, using the Black to White (custom color) color map. To correctly visualize each layer, it is crucial to adjust the level scale to match the irradiance values of each contribution until the displayed results are similar to those shown below.
The tool Measures offers the option to do a more comprehensive study of the contribution of each component to the Narcissus effect in the IR camera. By selecting the area of interest, you can obtain different measurement values such us the average, maximum, minimum, etc. This allow us to analyze the amount of irradiance provided by each thermic source.
By combining the layer definition with the measures tool, we can compare each contribution irradiance value. In the two images below, we see that, with the temperatures defined in this example for the thermic sources, while the sensor's average irradiance value is of 0.2 W/ m², the average value for the objective mount is an order of magnitude higher.
In the following links of our Speos User's guide you can find more information about simulation results: Visualizing Results and HTML Simulation Report .
Step 4: Simulation results. Ray path analysis.
Due to enabling the Light Expert option when defining our inverse simulation, an extra file appeared under the .xmp file in the simulation panel, an .lpf file. By opening this file, we are going to be able to display in the 3D view the path of the rays, with the possibility of filtering by area and surface of interaction.
- Open the .lpf file generated. The Irradiance map will open with a selected area by default whose size and position can be modified.
- The Light Expert tool offers the option to filter the displayed rays by selecting required or/and rejected faces of the system to draw the light path. We are going to click on the icon on the 3D view for required faces and select the four lenses of our system.
- For the definition of the Light Expert parameters, we are going to have face filtering by Or, with a ray number of 1000 and 50 mm of ray length for the drawing. The rotation will be set as 0 mm.
- In the structure tree, select the plane to display a cross section view of our system and have a better visualization of the ray path inside the IR camera. Any other plane of the user's preference could be selected as well by clicking on the third icon in the 3D view, and then selecting the desired reference.
- For the Irradiance map settings, we are going to replicate the process followed in step 3 of this article.
To be able to visualize the ray path with the settings we have stablished, all we need to do is go to the irradiance map and select an area, then the rays contributing to the result on the selected are a will appear in the 3D view, as shown in the images below. We can change the area's size and position, and also the plane in the 3D view for the visualization of the rays. With the selection of required and reject faces, we are also able to filter the displayed rays depending on the type of interaction.
In summary, thanks to the possibility of defining thermic sources and a wide operational wavelength range in the infrared spectrum, Ansys Speos software provides a workflow for visualizing the Narcissus effect in an optical system. The layer option in the irradiance map simplifies the analysis of this effect, allowing the user to separate the results by source contribution. The Light Expert tool offers a more comprehensive study with the visualization of the light path inside the IR camera, helping the user to understand the rays interaction that could be contributing to this effect in each area.
The analysis of the Narcissus effect in a IR system is essential in its design, since fully understanding the causes of this effect within the optical system is the first step toward correcting it in subsequent design stages.
Important model settings
Description of important objects and settings used in this model and instructions for updating the model based on your device parameters
Simulation computing time
One of the critical parameters when running an inverse simulation is the number of passes defined and, more precisely, how long it takes to compute a single pass.
On one hand, an inverse simulation pass is highly dependent on two sensor parameters that we can modify to adjust the computing time if necessary: pixel and wavelength sampling. The higher the x and y pixel sampling, the smoother the visualization results. This also applies to wavelength sampling. By reducing these parameters, inverse simulation can compute a pass in a shorter time without impacting signal/noise ratio. However, reducing these parameters reduces results accuracy.
On the other hand, in this example a limit of 1500 passes was established. To reduce the time needed for the simulation decreasing the number of passes is an option as well. Nevertheless, it is important to take into account that this action will increase the noise level of the results.
Another factor contributing to the computing time is the meshing definition. In this example, we have define local meshing for the lenses of the IR camera in study. The Sag value, 0.01 mm, can be increased if necessary to reduce the initialization time of the simulation.
In summary, when managing simulation computing time, it's essential to find and adapt the best balance between parameter definitions for each specific use case.
More information about simulation best practices in Ansys Speos Software can be found in Simulation Parameters for Visualization Best Practices – Ansys Optics .
[[NOTE:]] Signal to noise ratio (SNR)
Signal to noise ratio (SNR) characterizes the quality of a measure, in this case, in terms of an optical system. In order to obtain this value, we need to know the level of signal to compare it to the level of unwanted noise, allowing us to express SNR in terms of power of signal and power of noise:
$$SNR=\frac{P_{signal}}{P_{noise }}$$
Additional resources
Ansys related documentation
- Stray Light Analysis – Smartphone Camera – Ansys Optics
- Using FLIR Thermal Cameras in Speos to generate temperature maps – Ansys Optics
- Ansys Speos Spectroscopy - Ansys Learning Hub
- Analyzing Aero-optics Effects - Ansys Learning Hub