[ main ][ links ][ about ][ help ][ discussions ][ projects ]

Armasight SPARK CORE review - Hyper or just hype?

A review of the Armasight SPARK CORE Gen1 Night Vision Device, by David Kitson.

Armasight's CORE is one of those technologies that really didn't seem like a technology at all. I've certainly questioned their claims in the past, and have been frustrated by their use of acronyms suggesting technology outside of the typical marketting hype. Recently, Dimitry Rocklin, CEO of Armasight, provided an opportunity for me to test-drive one of their CORE products. The SPARK system is their Gen1 solution aimed at entry-level users who are trying to achieve Night Vision goals without the budget to go for Gen2 or Gen3 equipment.

Many of the claims made about the product by different reviewers have claimed performance similar to Gen2 equipment, something I am highly skeptical of, so it's been a good chance to get to the bottom of what CORE really is and to examine it with an analysing eye. Armasight provided their product as a whole system – that is, a complete part intended for use on rifles, or helmet mounted. The complete system allows for helmet-mounting, handheld use or rifle mounting.

However, in reading this review, it's important to state clearly and unambiguously at the beginning, this is a First Generation night vision device, and this is a Gen1 technology review, and before reading through to see how it performs, it's important to state the the achilles heal of Gen1 equipment is it's gain level. Unless it's used in a cascade tube ( exponential gain Gen1 ) then it's not going to provide much amplification. As a result, All Gen1 single-stage Night Observation Devices should be viewed in light of active systems – that is, they require Infrared light from an illuminator to function well. Even an older, less capable Gen2 will usually outperform it by an order of magnitude. Gen1 is a short-range system intended to be used with illuminators, usually at ranges well-under 100m and typically within 30m.

Armasight SPARK CORE

Armasight's SPARK system is a Gen1 monocular. It's main claim to performance is the CORE image intensifier, however it also has quite a few other particular features. The first to notice is that it has three hard-mount points. These are similar to the NVT-14 style ( or MUM or GT-14 ) and are relatively standard. Two of these are placed in the expected position, which is useful, since it means the device can be head mounted with a helmet. The third is mounted on the IIT cover, and is useful for additional equipment mounting, such as IR illuminators.

The unit is well designed. Unlike most Gen1's, it has the look and feel of a miltary unit, despite it's large size, and has a very professional feel to it that most Gen1 simply lacks. However for each “good idea” in it's build, there seems to be a design flaw. Firstly, the objective looks like it's PVS-14 style ( good ) but it's too small and threaded to small to take PVS-14 style parts ( bad ). It's a C-mount thread ( good ) but the flange on the upper is too long by a few millimeters to take other C-mount lenses ( bad ). The ocular has a strong focus knob and larger than usual eyepiece ( good ) but the eyepiece diameter is only 19mm (bad, well, it could have been better ). It has rubber eyecup and objective cover on retaining cords ( good ) but they are difficult to adjust and the cords sometimes get in the way of the IR (bad). It has a PVS-14 style power knob ( good ) but no “pull on IR” (bad. Well, not really bad, because, lets face it, if you have a Gen1, IR discipline means nothing ).

Because of this, the Spark feels like every time the designer moved in the right direction, he managed to somehow get it wrong. Despite this, it's important to consider that these are not really problems with the unit – just missed opportunities. Sizewise it's small for a Gen1 and looks good enough to make an amateur airsofter happy. It's larger-than-usual size is due to the size of the Gen1 tube, which is smaller than many Gen1's but still quite long. More on that later.

Mounting options are the usual Gen1, however because of the standard mounting points, this unit can be worn with a USGI helmet adapter. This is actually somewhat useful as there are many valid reasons to head-mount a suitable Gen1, even if it's lacking gain. Also, Armasight seem to sell adapters for many common parts ( eg, Afocal lenses ) so it does have potential to address some of the unit's design shortcomings.

Armasight AIM PRO.

The SPARK fits directly into Armasight's AIM PRO system. This is a mountable reticle projection unit that adapts the SPARK CORE directly to a rifle for night-time shooting. It couples with their 810nm illuminator via an adapter and with the SPARK fitted between to provide a complete shooting accessory. The AIM PRO, unlike many other Gen1 accessories, is a general Night Vision accessory, capable of working with any generation and with native mounts for well-known Gen2/3 units such as the MUM2 monocular and, no doubt, much of the Armasight range. With adapters, it will also support the AN/PVS-14 and has an exceptional range of adjustment and a safety switch to allow it to turn on into DAY mode.

This allows it to be used as a true day-night aiming device.

Coupled with the AIM PRO, the SPARK does provide a fairly functional hunting and shooting platform and although it looks like the reticle prism should get in the way, the objective lens looks easily around it and the light losses should be well less than reflex sights, which lose a lot of light through the reticle reflector.

The AIM PRO attaches to the rifle via a MIL-STD-1913 rail segment. It does however require a longer rail than many similar-sized items, and the QD mechanism is limited in grab and will fail to correctly adapt to smaller segments such as may be used to adapt an 11mm or 13mm rail to Picatinny. Armasight supply a rail adapter, which is just long enough my a few bare millimeters, but generally, it's designed for long, continuous segments of rail 2.5 inches and longer. Segments shorted than 1.5 inches will not engage at all.

The AIM also features a rail grabber for the monocular, which attaches well, though again, through a slight design limitation, can be adjusted but fails to grab similar rails completely. It's not a show-stopped, but it does highlight very small changes that would have yielded an improvement in design, especially as the system provides a ready upgrade path to Gen3 devices with similar rails as US rails are often slightly smaller than Russian rails of the same dimension.

In terms of function, the AIM PRO gave no issues. The adjustment tools provided are adequate and the mechanism is better than many others I've seen and tried. That said, despite two attempts to field test it, factors outside of the test prevented any accuracy testing of the AIM PRO ( I will attempt to update this review later to include more field testing ). At this point, however, I have no reason to thing it will not perform as expected.

The adjustment of the AIM PRO illumination levels however is exceptional, and reticles were visible, clear and well-defined in all tests. Some unwanted reflections were apparent, both in-reticle and immediate-field reflections, but otherwise it all worked as expected and the range was sufficient for Gen1, Gen2 and Gen3. Gen3 testing, in particular, gave excellent passive results.

Getting to know the product.

The SPARK/AIM-PRO combination seems to be the standard base pair within the Armasight range. The product arrived, clean and neatly packaged. It was definitely more tacticool than tactical, but Armasight have shown taste there and kept things utilitarian, the whole setup stowing away in two similarly-marked pouches, which I thought well suited to the task. Each is lined with extras ( batteries, keys, information pamplets ) and sized well. The IR810 illuminator is only boxed, but then it's small enough to fit in either pouch and two pouches is enough.

First impression is that where something should be, it's there. Lens caps, retaining rings, etc. Unfortunately, these later got in the way.

First test was down at a friend's farm – we went for a walk looking for roos in the paddocks on a moonless night. At first, there was a little light. At about moon-light level, late dusk, I took some pictures and was amazed at how detailed they were. The images came out well enough to fool anyone they were Gen3, had they not known I was using a lot of skyglow. Pictures can be deceiving and there were few clues that it was a Gen1. The aiming reticle shows up well and we began with some range tests, working out at what ranges we could spot each other and notice a significant change in albedo of the clothes we were wearing with the IR illuminator. The best range we got to was about 120 yds, according to his laser rangefinder, which was more than I was expecting.

< A real SPARK CORE image. It's about moon-dark - but very diffused. It looks really good and is a genuine un-altered image. In higher light, the image is pretty good. Trees @400m from memory. >

< Another image in a slightly different direction. Also very sharp. >

Then we went for a walk in the dark. It wasn't over-cast-remote-rural dark, but it was dark enough that we had to dark-adapt our eyes for at least five to ten minutes just to see the paths we were walking on. The monocular didn't provide an image that helped much, but did give similar vision to non-adapted eyes.

We spotted a roo at about 100m, and several more at 200m. The one at 100m definitely spotted us as we flicked the IR beam ( 810nm ) in his direction, but despite being light-shy, he didn't run and the eyeshine helped locate him. Appearance was well made out and we could see he was a big roo ( and a he... No, I can't tell, but my friends could, and I know from experience he would have been correct ).

At that range, the reticle dot filled the roo's head, but had we needed to place a shot, he felt it would have been possible, so we decided that cats, dogs and foxes out to 100m would be fairly easy to drop, even with a 1x lens. A 3x lens would improve accuracy and might be needed for really small targets like rabbits.

< The roo can be seen top-of-reticle >

< The eyeshine helps a little - it's easier to spot with the eye than the camera... Range approx 100m >

The walk back allowed us to determine that the beam from the IR could be seen about 200 to 300 meters away, on highly reflective trees and the like, but recognition wouldn't have been possible at those ranges. Just because light can be seen doesn't mean there's enough to make out the necessary detail. That goes for spotting too. The kangaroos at 200m were clearly visible to the Gen3 we took, even without IR – passive was enough. Yet with the IR, the roo at 100m could just be made out, but the ones at 200m were invisible. Also, the middle of the screen needed to be used. As the unit was panned, the kangaroo would blend into the background at the screen edges, meaning it had to be in the middle third of the screen to make out.

Bench testing.

Collecting results, the next test involved putting the CORE under some scrutiny. I hooked it into the tester, and the results under high light ( moonlight ) conditions were exceptional, but the results under low-light were zero. No image could be made out at all. Not even the general shape. This wasn't unexpected, but the high light results were very clear. My recent digitals could see the high-light readings on frame combination settings, but not at 30 Hz. The SPARK CORE however make it look like a Gen3. Several tests were like this – when it worked, it worked well. When it failed, it failed hard.

The next test was for distortion. I printed a matrix pattern and took some pictures, noticing something that we started to spot during the first field test – The two most prevalent things to notice were that the middle third of the screen, an area of around 10 degrees, was the sharpest and that the distortion increased and resolution decreased rapidly beyond this point.

< A comparison with a low-light shot with a Gen3 and the Armasight lens show the lens is actually fairly good - the distortion comes from the Gen1 tube. >

The other thing I noticed was that often the image was good enough to see clearly with my eye, but the camera couldn't make anything out. The extremely low background noise and clean image often give the impression that the unit was working like a Gen3 – and I'll say Gen3 because it has the sort of image you see on a PVS-14 with the gain turned down – like looking through green glass, except I'll call this one a coke-bottle lens,because of the distortion and blurriness in the edge of the image.

So I adjusted the amount the door was open behind me, letting in light from the next room, so that I could get good photos.

At this point, the distortion had be curious, so I removed the objective (front) lens from the SPARK and noticed it was threaded 1” 32tpi... Almost C-mount when measured, but a few millimeters too long to support aftermarket C-mount lenses, so no cheap F1.4 50mm lenses for this model :( Quite a shame that. However I could fit the SPARK lens to one of my Gen3's. The result was actually quite good. The SPARK uses a F1.7 lens 35mm lens, which really isn't suited to the Gen1 – which could benefit from a good piece of cheap glass in F1.4 or F1.2... However despite the noticeable vignetting, it produced a clear image on the Gen3 in low-light with almost no visible distortion. Nice lens Armasight...

However now it was confirmed that the source of the distortion was the tube. Looking in, I saw that the rumors of a proximity focused tube was incorrect, as the anode cone was clearly visible through the brownish photocathode. At this point, I realized any chance of it being a Gen1+ were gone – It was just like every other Russian Gen1, with almost the same internal build. In this case, the Ceramic in the CORE didn't really mean anything of note.

< Pincushioning caused by the electron lens setup with electrostatic inversion >

Still, it had the performance, and it was good. A combination of high resolution and high sensitivity was giving the SPARK the edge over other Gen1's I have tried.

Bench testing complete, it was time to head out for a second field test – this time somewhere darker.

Katanning Tests.

Katanning is where I do my NV testing. Located in the heart of the Western Australian wheat belt, it's one of the darkest places on earth that's still populated. Great for astronomy. Even better for testing NV gear and I have plenty of experience with how it looks through different Gen1 and Gen2 units as well as Gen3. So when the lights go down, I'm not so reliant on my eyes to tell me how dark it is.

Well, we went walking until it got really dark, then to show my son how badly a Gen1 performs next to a Gen2, I pulled out the SPARK.

“Dad, this is pretty good,” he says, holding up the SPARK.

“What do you mean, pretty good?” I asked.

He hands it to me, and I start to see where the myths about them outperforming Gen2 come from. Because of the low noise floor, and no amplification noise, the image is very faint, but visible, and definitely an improvement over unaided vision. So I pull out the AU/PVS-14 and turn the gain way down as low as it goes and swap between them, and it's about the same. A PVS-14 should be below 100x system gain when fully turned down – pretty close to the Armasight SPARK CORE. So if you want to fully understand how a SPARK CORE will operate on any given night, and you have a PVS-14, then turn the gain knob all the way down and that's about right. That's what a SPARK CORE looks like.

< Low light in Katanning - I had to amplify this with a Gen3 tube to take the picture. It is through the SPARK CORE though. >

< In another direction - Still picking up a lot of light, and again, amplified by Gen3 stage >

Well, you can see with a PVS-14 even turned all the way down... So I took a picture, using a Gen3 tube to amplify the image and you can see the noise, but it's pretty dark. It looks like a full moon from the shadows. Actually, it's moon-dusk on less than a quarter-moon and even the PVS-14 isn't showing the shadows up like that. Mind you, both together are pretty bright, and I easily get a snapshot. Not a perfect (infinity) S/N as claimed, but closer to about 10:1 I think, under very dark conditions. Still, not that bad, and I am impressed.

It's then that we discover the CORE won't attach to the rifle rails. It needs a longer rail segment ( 1.7 inches minimum, 2.5 inches optimum ) to connect to – A pain, because there's no reason for it to be like that.

< The SPARK CORE and AIM PRO mounted on an AR6. The rail length is insufficient to allow a suitable mount. >

So we finish testing there, but generally I find my Son is seriously liking the CORE at this point, and he starts pointing out to me that it's perfectly suited to the kinds of ranges we shoot at with the AR6. Though in some final tests I notice that it seems impossible to get rid of immediate-field reflections and make a note to investigate on return.

Back to final testing.

Well, the only remaining tests are the head-mount and to investigate some reflections. I pull out the standard mount I use and insert it into a USGI slot of a Rhino and the SPARK slots right in, like it's meant to be there. The spacing around the rails is pretty much standard too, so optical alignment is no problem. There's a proper attachment for that, but I don't have it so can't comment on it, but the one I have is fine.

Walking around is OK – It's about 30 degrees FOV – less than the 40 I'm used to but I don't notice a lot of difference. However the distortion is terrible. It feels like the last time I was drunk and makes it really hard to use both eyes, but once I get into a darker area, the feeling of distortion disappears – so in darker circumstances, my brain begins to use both eyes again. Optical alignment is actually pretty good on the tested unit, and I find both images align immediately, at least in the centre.

< Head mounted testing - the SPARK CORE worked quite well and was more comfortable to wear than I expected >

Resolution is excellent, but the image is very dark, especially inside. I turn on the IR, then notice a major flaw in the design. Either the plastic is carrying IR or the LED inside isn't shielded, but the IR light introduces a HUGE amount of noise into the image – perhaps reducing the S/N to around 4:1 or so – Some image detail completely vanishes.

I places some duct tape over the LED to confirm, and the S/N drops to near-units as the image in darker areas disappears. This is something that needs to be fixed. It's bad enough that the rubber bungee from the lens cap and the battery retaining strap are constantly throwing shadows across the view, but this kills it entirely. I get the feeling it needs a few good coats of black paint or something inside. I didn't investigate that issue further, but will ask Armasight to comment on that.

< Immediate-field reflections from the IR810 on the prism that projects the reticle. It starts out like this, and gets progressively worse with higher power. >

Finally, I check external illuminator immediate-field reflections. They are pretty severe because the IR810 has no shade out front – something it badly needs given how far back it's mounted on the rifle. The barrel isn't so much a problem, but the prism that projects the reticle causes serious reflection. A small piece of duct-tape to shade the prism corrects the situation and I'm able to see the reticle clearly, but the final problem presents itself. The illuminator can't boresight with the reticle.

Now, admittedly, I haven't zero'ed the unit, so perhaps I would be luckier if I did, then again, perhaps it would be further out. I can't criticise that too much however since it's a fairly common limitation. Expanding the beam covers the reticle again, but it loses intensity. This is something that individual users will just have to address individually.

Performance of the CORE

After a few weeks of practical testing, I come back again to a question, and we get to the heart of the Armasight claims – the CORE image intensifier. Core stands for Ceramic Optical Ruggedized Engine, an Armasight acronym, and it's what provides the image for the SPARK. On first view, the CORE specifications look impressive.

These seem to be reflected in the difference between Armasight Gen1 and CORE on their website. Gen1 comes with typical figures of around 350-400uA/LM sensitivity ( vs 200 uA/lm for Gen1 ) and the tested unit came in over 500 uA/LM – and their main claim, that the CORE uses Ceramics in the manufacturing process rather than Glass that most Gen1 are made of.

The claim of ceramics is an interesting one. Armasight claim 500 G's for CORE with 200 G's for glass. The first thing to mention is that Gen1 and Gen1+ has been using glass for a long time, and Gen1+ glass tubes are rated well in excess of 500 G's. Some Gen2 devices made for weapons recoil that come out of Russia are also glass, so the claim that ceramic is somehow better in this respect isn't very accurate. Ceramic does however allow different manufacturing techniques. In this case, the Core uses a ceramic space to keep the two high-voltage sections apart. Otherwise it's just a normal Gen1 with a construction not dissimilar from the method used to make mx9964 tubes.

In evaluating CORE, the following criteria is used. This is broken down by category so as to provide an effective overview and to include the influence of housing technologies on the overall design.


In technically evaluating the CORE, it was tested for performance and general quality and while it only performed as a Gen1 device, there were moments when I was significantly impressed that I had any function at all. The light levels were well below where the camera could capture an image, but the fact that the amplification that was occurring was clear and sharp left me able to see more with the SPARK than without it on both test nights and also under starlight only conditions.

Having tested and evaluated it, CORE is what I always suspected. A simple Gen1, with no enhancing technology. It would be wrong though to say it's the same a Gen1. It is a very good Gen1 and the focus and attention on the resolution of the screen and the photocathode response do actually make a significant different. With a well-designed housing to take advantage of it, I was continually surprised that, under many conditions I expected a complete failure of the test, it continued to provide an image and one that I was able to use.

Range with external illuminators was exceptional and the SPARK CORE easily provided sufficient image to identify targets at 100m, with enough resolution to do so with 1x magnification.

Coupled with the AIM PRO, it seems that Armasight have created something I have never seen anyone else do successfully. They've built a model by which a Gen1 system can be upgraded to a Gen3 system at reasonable costs and with relatively low-losses in the process. With an estimated $500 street-price for the SPARK CORE, it does provide a functional system for someone to equip their rifle with a complete system, obtain head-mounted equipment and move around in darkness, especially if they are considering upgrading to a PVS-14 or MUM ( adapter required for PVS-14 ).

<  The same system, this time with a DBT-44 rev2 Gen3 monocular in 2x configuration - Clips straight in... An easy upgrade to Gen3 >

That said, the SPARK CORE would also provide a low-end solution to people with reflex and red-dot sights as well, if they supported suitable NV models of operation.

It's not a magic solution to NV problems. Although some image is available under dark conditions, it won't peer into dark areas without enough light, and doesn't lend itself to practical uses such as law enforcement where passive operation is critical. It's only gen1 and it's bulky compared to modern NV. It has as many flaws as it has features.

But for a Gen1, I'd be lying if I said it didn't impress me. And I can see it being useful to people who end want to start with a Gen1 unit and upgrade to Gen3.

So in the end, is CORE Hype or Hyper? Well, given that it's just Gen1 technology, I have to go with hype. That said, it's the undisputed King of Gen1 devices I've seen in the past and stands head and shoulders above them all. It proved useful and usable in situations and conditions I expected it to fail in. I wouldn't have expected anyone to come up with a useful and usable Gen1 device, but Armasight have succeeded there.

So for people who really don't need to see past a few dozen feet, who can put up with the distortion for head-mounting, or for those who just have to shoot a few coyotes out to 100yds, but need to identify what they are shooting at, it is actually suitable. The fact that the illuminator and sighting system work well for other Gen2 and Gen3 applications also proves a bonus, as is the fact it support other systems.

Not suitable for tactical applications, but definitely suited to low-budget active night-vision requirements.

Core website:

Logged in as:guest: 1:guest ** :./data/user-data/reviews-SPARKCORE

No comments presently. End of page.