Solid state electronics have provided lighter weight night vision units that work better than the old-fashioned gear that used photomultiplier tubes, but there was an even older technology as [Our Own Devices] shows us in a recent video. The Metascope Type B was a first-generation passive night vision viewer that relied on moonlight.
The video shows a 1946 technical paper from the Office of Scientific Research and Development with [Vannevar Bush] credited as the institute’s director. If that name sounds familiar, you may remember that he foresaw hypertext (inspiring both [Doug Englebart] and the creation of the Web).
The Type B was an improvement over the older Type A, which had been tested during the invasion of North Africa in 1942. The type A weighed less than two pounds and was much smaller than the type B. However, it didn’t work very well, so they stopped making them and did a redesign, which is what you see in the video. The type B weighed almost 5 pounds.
To use the metascope, you had to “charge” it with light and then wait. Eventually, you’d need to charge it again. The type B allowed you to charge one phosphor plate while using another one. When that plate became weak, you could swap the plates to continue using the device.
If you aren’t keen on the history, you can skip to just before the 15-minute mark of the video for the hardware examination. He doesn’t open the device, but that’s probably wise, given the nature, age, and rarity of the metascope.
Modern image sensors are very sensitive to infrared, and normal cameras usually have filters to keep them out. Not that you can’t remove it, of course. If you want to see something more modern, [Nick] built his own AN/PVS-14 night vision scope and you can too.
“Solid state electronics have provided lighter weight night vision units that work better than the old-fashioned gear that used photomultiplier tubes”
That’s news to me. The old night vision units that I am thinking of didn’t use quite the same thing as what you would now find if you searched for a photomultiplier tube. If you look for a photomultiplier tube now you will find something that has multiple stages of amplification but no particular spatial resolution or image-forming ability. So it’s great if you want to detect single photons coming off of a scintillator crystal so you can make yourself a gamma detector, but it just isn’t an imager unless you were to make a sweep function of some kind. I’m going to assume that you are including other tubes with photocathodes, since that makes more sense.
The first generation of those, which is all I have ever used, is the kind of night vision device that accelerates electrons given off by a photocathode across a voltage potential landing onto a phosphor screen. Very simple, and it can run for however long you maintain the voltage potential (not just for a few seconds before recharging), but it has low amplification. It’s generally considered to need either moonlight or a convenient source of supplementary illumination, but I think some people may skip mentioning the second part. Some “gen 1” solutions used multiple stages back to back for more amplification, but it’s got tradeoffs. Something that works better is to make a microchannel plate to put in between the photocathode and the phosphor. It basically is a massive array of the same electron multipliers you find in a photomultiplier, meaning you can keep the spatial information for the image while adding amplification. The other thing to do is to not just provide a fixed voltage to the photocathode but instead vary it to deal with brightness fluctuations. Otherwise you can damage the tube and also keep yourself from seeing anything if you accidentally look at something bright even for an instant.
As far as I knew, the best image intensification was still a tube using a photocathode whose electrons are amplified by a microchannel plate and directed towards a phosphor, rather than anything using a solid state CMOS sensor. (Well, assuming “best” means “works in the darkest conditions” but excludes thermal infrared, which is a whole different thing.) I know there’s silicon photomultipliers and such, so I could believe there’s good solid state sensors now for amplifying visible or near infrared, but I haven’t seen them for imaging yet. It looks like there’s something called BSTFA now, which might finally be a digital, color sensor with good enough sensitivity although I didn’t look in detail.
State of the art night-vision systems are still using photomultipliers because the sensitivity is just better. The top systems combine it with some augmented thermal imaging or the like, but the basic sensor is still just an advanced tube
The best digital night vision doesn’t come close to modern Gen 3 Elbit or L3/Harris intensifier tubes.
The internet aligns with real life. I just found 2 night vision goggles @ a thrift store for $7. For being around 6 years old they work better than i expected with my previous experience being night vision mode on camcorders. Now i get a little history.
I collect Night vision, and it is a fascinating subject, with a lot of experimentation with different technology, tweaks, phosphors, cathode types and designs and optical paths. I especially English cascade Gen1+ EEV P8079 tubes, simply because of how good some of them still are, and how nice the contrast and image is. Sure, its a beast of tube weighing 0.9kg, but it’s just nice to look through. Another favourite is Philips XX1332C which have a 50mm optics. It’s just 30lp/mm resolution, but still higher total resolution than even modern 80lp/mm small tubes, and the monstrous optics collect a lot of light. You wouldn’t want to wear it on the head, but in a vehicle or tank?
Modern digital camera sensor is okay, especially as cheap as it can be, but in starlight I always select an image intensifier tube, since they really can’t be beaten. Thermal have their use, but a completely overcast day where the sun couldn’t be seen, and if the nature have seen rained, or there is snow, thermal is pretty useless for navigation and moving about. For part of the world, that is a third of the year. Focal plane thermal have now become cheap enough that it is the best and cheapest night vision, but it have it’s cons, and a major is that it doesn’t see visible light, so it won’t see LED lights, headlights and such, and they suck way more battery power. I expect dual spectrum digital camera plus thermal camera fused units to become the go-to unit over the next 20 years, at least when cost is an issue. For a lot of countries military even more so, as budgets matter, and giving every combat soldier a head unit and a day-night weapon sight, is far more important than just issuing really good to a few, and you know the brass is going to be sold on video streaming and video recording, and ‘shooting from cover’.
I have an abraham tank night vision site that has exalant gain and clarity but it weighs over 10 lbs and is a periscope design. Is it possible to take out the tube and make it usable in a lighter chassis?
“Modern” night vision is a toy. The minimum performance for a toy can be almost zero, hence the low cost.
REAL night vision gear still uses photomultipliers, because they are much MUCH better at letting the user actually see in the dark as naturally as possible. And they do it WITHOUT simply being a flashlight and a fancy camera.
Real night vision gear has very low latency.
It also isn’t limited to single digit frame rates.
Why?
Because people’s live depend on it.
Try driving with those fancy 4k night cameras.
The 3fps @ 150ms latency is going to put you into a wall.
This old tech is pretty cool, but not even remotely in the same space as a modern toy. This was made for saving/taking lives.