Lightweight RPO Lenses vs US Milspec - READ BEFORE YOU BUY

Lightweight RPO Lenses vs US Milspec - READ BEFORE YOU BUY

The latest fad sweeping the night vision industry are lightweight optics. While nothing new (the NV Devices BNVD-UL has been sporting RPO lightweight optics since 2019), this concept has seen recent resurgence. Lightweight optics promise a reduction in operator fatigue by reducing overall system weight, but this reduction in weight comes at the expense of some optical trade-offs.

This is a follow-up discussion to the YouTube video published in Q2 2022

This is the updated video featuring RPO 3.0


In the following series of images, we demonstrate some of these trade-offs in a side-by-side comparison between US Milspec optics and RPO 2.0 and the latest 3.0 using tightly-controlled testing conditions to remove bias, and user subjectivity.

The test setup:

  • Sony A7S3 with custom PVS-14 lens adapter attached to Sony G-Master 16-35mm F/2.8 lens (fixed white balance, picture profile, shutter speed, frame rate, ISO)
  • Photonis Echo PVS-14 (same intensifier and housing used in all tests)
  • PVS-14 and camera lenses are independently focused using focus-peaking to achieve the best focus each time

In the first test we have the standard black and white checkerboard contrast and light-level test:

There is substantial light fall-off at the edges of the image. This is because the RPO 3.0 lens features a reduced exit diameter. Not only does this physically limit the amount of light that would reach your eye, but optimal viewing distance has also been decreased. A technical discussion with RPO engineers confirmed that the eye relief is set at 25mm or 1 inch which is a very short distance away from your eye. This means that using any sort of eye protection including low profile goggles (eg: Oakley M-frames) will result in perceived vignetting or occlusion of the edges of the image. The situation is further exacerbated as the profile of the eye/face protection increases (gas masks, etc). The decreased eye relief is confirmed via the in-person experience, as pushing the device further away from the operator's face results in loss of image in the periphery. Photos  circulating online showing no visible vignetting with RPO 3.0 lenses are likely taken with the camera pressed up to the eyepiece - an unrealistic use-case.

You can see the difference in viewing window size below:

The reduction in signal is further confirmed when we analyze the footage using Adobe Premiere's vector scope in raw Luma:

Notably not only does the periphery of the image have lower signal, but the peak is also lower comparing between Carson Industries and RPO 3.0.

The same footage, now analyzed using the RGB Parade:

Besides some blue fringing at the peak, the Carson still scores higher than RPO 3.0 across the board.

Other measured tests conducted using the Hoffman ANV-126A-001 test rig have stated that the RPO lenses (both 2.0 and 3.0) feature "higher light transmission". The reason that may be the case is that our particular PVS-14 lens adapter setup places the point of optical convergence away from the "ideal" 25mm eye relief that the RPO 3.0 was designed for. This is due to the shorter overall length of the RPO 3.0 eyepiece preventing our lens adapter from being mounted closer to the optimal 25mm distance. However, it should be pointed out that the distance between the PVS-14 lens and camera lens is fixed for all tests, and the Carson industries yields a consistently superior image, shown below in real-world testing. In practice, this means that to get a comparable image to US Milspec, a system with RPO lenses must be forced closer to the operator's eyes.

As gain is scored on the Hoffman ANV-126A as a singular, overall value, one other reason for the claim of higher light transmission may be due to the higher flare giving higher light levels but lower overall useable information (think Michael Bay / JJ Abrams lens flares). This can be seen below:

In the images above, a few key notes can be taken away:

  • The RPO lens set provides substantial light fall-off at the edges in a fixed setup. This means that the RPO will inevitably force the user to place the device at an unnaturally close distance to the user's face and eyes, reducing the operator's ability to use their critical peripheral information (if trying to look around an NV device)
  • The RPO lens set also has ring-shaped flare artifacts and reduced levels of detail surrounding light sources (confirming that higher light levels, while technically meaning "higher light transmission", actually provides the user less information)

In an effort to try to recover some of the lost periphery, the RPO 2.0 eyepiece is swapped back in. It should be noted that the RPO 2.0 eyepiece lens cell does not have the reduced exit diameter as the RPO 3.0 and the resulting image confirms that the peripheral edges of the image is retained and vignetting is removed compared to Carson US Milspec.

There is however, a downside to the RPO 2.0 eyepiece and that is rainbow artifacts at the periphery (1 o'clock)

It is also noticeable that RPO 3.0 and the "2.5" combo exhibits odd crown-shaped light flaring artifacts seen below:

These oddly-shaped artifacts are not present on US Milspec lens systems:

Conclusions:

While weight reduction and increasing operator comfort can be important, this currently comes at the cost of decreased detection capability, distracting image quality, reduced operator eyebox, and increased cost. While the results above are conducted in real world lighting scenarios, it should be noted that they were conducted in an urban environment. Rural or forested areas will show reduced difference in user perception.

The lack of factual side-by-side comparisons and limited pool of available information and ability to try these lens systems in person has led to the perceived notion that "more expensive" simply equals "better" when the data and images shown above show otherwise. As our original review stated, whether or not reduced weight is worth the trade-offs, should be a highly user- and application-driven decision.

Back to blog