The Image Intensification Pipeline
Analog night vision works by amplifying existing light rather than constructing an image from digital data. The process is a chain of conversions, but every step happens at effectively the speed of light, which is what makes a Gen 3 tube feel instantaneous to the eye.
The chain runs roughly like this: photons from the scene enter the objective lens and strike the photocathode at the front of the image intensifier tube. The photocathode converts those photons into electrons. Those electrons are accelerated and multiplied through a microchannel plate (MCP), which dramatically increases their number. The amplified electron stream then strikes a phosphor screen at the back of the tube, which converts the electrons back into photons — now at a brightness level the human eye can use. Those photons exit the eyepiece and hit the user’s retina, where the eye and brain do the rest of the conversion work biologically.
Because every conversion in this pipeline is essentially instantaneous, an analog tube has no perceptible latency. As Isaac put it, the photons “pour through the tube like water.” This is the single largest functional difference between analog image intensification and digital night vision, where the sensor must collect data, hand it to a processor, and push it to a display — all of which introduces delay.
What the Tube Actually Sees
A Gen 3 image intensifier responds to a band of the electromagnetic spectrum that extends from visible light into the near-infrared. This is why analog NVGs work well with 850nm and 940/950nm IR illuminators and lasers, the standard wavelengths used on weapon-mounted IR devices.
Sensitivity drops off as wavelengths get longer. In direct comparison testing with a range of IR LEDs from 1070nm up to 1885nm, an Elbit-tubed PVS-14 showed clear response at 1070nm, with sensitivity falling off rapidly after that. By around 1550nm only the faintest glow was visible, and nothing was detectable above that. Silicon-based digital sensors, by contrast, can see considerably further into the near-IR — which is one of the few areas where digital outperforms analog on raw capability.
Within the visible-to-near-IR range the tube actually uses, sensitivity is very high. The phosphor screen produces a monochrome image — green phosphor (P43) historically, white phosphor (P45) on most current civilian-market tubes. The tube has no native color information; the photocathode simply counts photons regardless of their original wavelength.
Aperture, Gain, and Why Analog Still Wins on Sensitivity
The PVS-14 uses a 27mm f/1.2 objective lens feeding a photocathode roughly the size of the tube’s input window. That is a physically large light-gathering area compared to the tiny CMOS sensors used in current digital units like the Sony Starvis 2. Even before the intensification mechanism does its work, the analog device is collecting more photons.
Combine that larger collection area with the high quantum efficiency of a Gen 3 gallium arsenide photocathode and the electron multiplication of the MCP, and an analog tube continues to outperform consumer-grade digital devices in the lowest light conditions — particularly under canopy, on overcast nights, or anywhere the photon budget is genuinely scarce.
The user controls this amplification with a manual gain knob on the side of the PVS-14. This is meaningful: the operator can directly adjust how aggressively the tube amplifies, trading brightness against noise depending on conditions. There is no automatic processing layer interpreting the scene.
Power consumption reflects how efficient the analog pipeline is. A PVS-14 runs roughly 50 hours on a single AA battery, and longer on premium lithium AAs. Digital devices, which have to power a sensor, processor, and display continuously, require dedicated battery packs to reach comparable runtimes.
How Tube Performance Is Measured
The data sheet that comes with a Gen 3 tube records several specifications that describe how well that specific tube performs its job. The relevant numbers for image intensification are:
- SNR (Signal-to-Noise Ratio): the ratio of true amplified signal to electronic “scintillation” noise in the tube. Higher is better. This is widely considered the single most important spec for low-light image quality.
- RES (Center Resolution): measured in line pairs per millimeter, this describes how finely the tube can resolve detail. Higher is better.
- FOM (Figure of Merit): SNR multiplied by RES. OMNI VIII mil-spec contracts require a 1600 FOM minimum (64 lp/mm × SNR 25).
- EBI (Equivalent Background Illumination): the light the tube produces with the objective fully covered — essentially the tube’s own noise floor. Lower is better, because EBI sets the absolute darkness threshold below which an image cannot be recovered.
- HALO: the size of the bright ring that forms around concentrated point light sources (streetlights, headlights). Lower is better.
A clean spot chart — meaning no cosmetic blemishes in the active image area — is also part of the evaluation, particularly any imperfections in Zone 1 (the center of the field). Tubes sold by T.REX will not have Zone 1 imperfections.
Filmed vs. Unfilmed, and Why It Matters Less Than People Think
Both major Gen 3 sources for the U.S. civilian market — Elbit Systems and L3Harris — produce tubes that meet the same generational structure. The technical differences come down to construction:
- L3Harris filmless tubes remove the ion barrier film between the photocathode and MCP. This typically yields slightly better halo numbers and somewhat higher SNR statistically, at higher cost.
- Elbit thin-filmed tubes retain a very thin ion barrier film. They tend to be slightly less expensive and are generally what’s sold at the YH/VH/PH mil-spec grades.
Both are true Gen 3 and both are adopted by the U.S. government. As Scott from Nocturnality points out, the generational structure exists precisely because the DoD wanted multiple interchangeable sources meeting equivalent minimum performance — so for ground-based use, the practical difference between a good Elbit tube and a good L3 tube in the field is generally small.
T.REX’s batch specs reflect this: Elbit tubes ship at 1734–2462 FOM with SNR 27.1–30.6, and L3 tubes at 1766–2484 FOM with SNR 26.0–34.5. Both batches sit comfortably above the 1600 FOM mil-spec floor, and at this point on the performance curve, additional spec gains produce diminishing real-world returns.
What Image Intensification Cannot Do
The same physics that makes analog night vision fast and sensitive also defines its limits. Because the tube simply amplifies photons, it requires some ambient photons to amplify — starlight, moonlight, sky glow, or active IR illumination. In total darkness (a closed interior with no light leakage), an image intensifier produces nothing useful, which is where thermal imaging takes over.
The image is monochrome. Color night vision through an analog tube is technically possible using spinning color filter wheels at the objective and eyepiece, but it costs one to two stops of sensitivity and adds significant size, weight, and complexity. The current production solution for color is digital — at the cost of latency and lower base sensitivity.
The tube is also vulnerable to bright light damage, particularly to point sources like lasers, though modern Gen 3 tubes have auto-gating circuitry that rapidly cycles power to the photocathode to protect it from temporary overloads like headlights or muzzle flash.
Within those constraints, image intensification remains the reference standard for headborne night vision: zero latency, high sensitivity, low power draw, and a forty-plus-year track record of refinement.