Hello,
I’m trying to measure illumination via a pinhole camera and I want to understand the effect of each camera property in Omniverse. From tests I’ve conducted, it seems that:
- The
horizontal_aperture
parameter represents the screen width in physical units (and not the pinhole aperture). - The pinhole size appears to be proportional to the
focal_length
with a factor of approximately 1/25 (meaning f/25 f-number).
However, the typical f-numbers for pinhole cameras are much higher (ranging from f/90 to f/360 or more), so I suspect something might be off with my results.
Test Description:
I used a square plane and illuminated it from above with a white light source (distant light). The scene’s illumination was captured (PtDirectIllumation annotator) using a pinhole camera placed directly above the plane’s center. I then compared the sum of the captured illumination to the light emitted by the plane and from this, I calculated the f-number mentioned above. I used isaacsim.
Could anyone provide insights into whether my interpretation is correct? Is the scaling of the pinhole size and the resulting f-number in line with what should be expected in Omniverse’s pinhole camera model, or might there be some other factors or settings influencing the results?
Thanks