Well, UV-light was deemed not really tactically advantageous at night b/c there is none to amplify. The skyglow, the light that can be amplified peaks towards the SWIR region, not UV...  

Early multialkali photocathodes (Gen0,1 and early Gen2) were all very sensitive to UV light... but the tech evolved (like the S25 Exteded Red PC) to see more into the red and NIR, not into the UV side of things... b/c of the advantageous skyglow the makes passive observation possible.

So, why do you keep implying that Gen3 or Gen2 I2 devices are not full spectrum? A Gen2/3 with a 720nm IR cutoff filter will still work, at a reduced performance, but will work just fine.

Like I've said, my advice is don't use anything that emits light while going up against other humans that might have night vision capabilities.

Wasn't the point of inventing NVDs to see at night without giving away your position? otherwise, why not just use a flashlight then?

I don't understand what you mean by filters here, but the question remains: why would anyone cripple a perfectly good Gen3 scope with a filter? With that said, there is only one filter I would ever put in my Gen3 scope, that is an H-Alpha filter for astronomy purposes.

G.

I bothered to do the duplicate photos in my time (as promised), and used a moon calendar to pick the least favourable time and even checked the exposure (New moon, 0.1%) - it was a 40km drive.  

Just getting the thread back on track - the comparisons are up above (also I've blocked gman).  

The data, as always speaks for itself.  

The spectral ranges of the different types of NV device are well understood.

The peak of the visible spectrum is around 500nm and the best a GaAsP tube can do is a QE of 50% (but is near useless in the NIR) and a typical GaAs G3 is down at less than 30% in the visible spectrum and tails off quickly into blue/UV.  

The performance of G3 tubes is skewed to the NIR, with peak sensitivity around 800-850nm (NIR), thats why most illuminators/IR designators are set around 840nm/850nm.  None of this a mystery.  

The Aurora Pro is a good device within is operating window, its peak performance being somewhat less than a modest G2 device.  The OPSIN looks like it will improve on this somewhat, how much is yet to be established.  

It should be noted the Aurora Pro is used by a number of government agencies both in the US and overseas.  

FBI Aurora Pro

This is no surprise to me as its an excellent surveillance and reconnaissance tool.

Any suggestion that SiOnyx is trying to con people is nonsense.

We are in agreement here, "its not as good as even modest Gen2." Which is what I've been trying to say since this whole thing started.

The next scope who knows, but I am not so sure it will be a major improvement... considering the price is going to be substantially higher, I feel this is a con.

The Pro is borderline a con, at the current cost at least, especially considering that I've purchased/built Gen2/3 gear for less than the cost of a Sionyx... so IMO, every time I see charts/specs I fall back to the 90s and imagine some sleazy conman trying to sell me a high performance Gen0 Ruski device...

Personally I wouldn't recommend it as your first device... get a used Gen2/3 device IMO...

Now, the cost would've been justified IMO if it would've had a higher resolution sensor, then I would've probably kept it for visual assisted astronomy.


G.

Hey, ... :) UV is ONE tool in the box. And with every tool you have to know how to use it, it´s strengths and it´s weaknesses. Yes, UV is active and i would´nt wave a UV-laser around in a warzone, but in comparison to a flashlight it´s lowkey active AND has the power to remoteswitch some things from an undetectable passive to an easy detectable active state, like these f-ing PIR-sensors people are plastering their houses and gardens with and which can be really fun to identify and remoteactivate them and the connected lights to drive the houseowners nuts ...

And with every tool you want that tool to work 100% for you and not being a double-agent in your hand, right? Let´s take a knife. When i use a knife i want to be one 100 percent sure that the edge on this companion is THE ONLY edge it has and that there´s not a hidden second one which slaughters me while using it. This was the reason i try to "warn" people of that exact property of standard UV-shortpasses. That these leak IR-light which annuls the lowkey-activness of your UV-lightsource in certain environments.

Maybe some misunderstanding here regarding the definition of "full-spectrum". Yes your are right in characterising tubes as full-spectrum. But: they are full-spectrum on(ly on) the receiving side, on the output-side they are flat-monochrome, they do not provide any information regarding the quality (qualities) of the light (lights) which are present in the scene you are looking at with a tube.

The full-spectrum i mean on the other hand is a device which hands the full-spectrum capabillity from the input to the output so the user receives much more information of what is going on on the stage.

Agreed.

The same as with the tools. There´s no reason to restrict the tools in your box. Filters can be very helpful. For example: when using NV in town or similar contrast-rich environments, i always use this IR-pass on my device. Using these has the same effect as using a log-profile on a digital camera. It flattens the contrasts significantly by keeping away the tubes from throtteling because of some stupid highlights in the scene with the effect that you can see MUCH deeper into dark spaces.

---

And else: Can we please keep the hate out of this discussion?

---

When the technology develops it is much better to see color, no doubt... but until then, I2 devices are a better choice.

Well, I understand the concept, but carrying too many tools could also be a handicap.

Digital, at the moment, its just not there for on-foot soldiers. For vehicles, digital sensors are the way to go, no size restriction, cooled sensors and much more capable all around... with sensor fusion as well...

G.

Especially when it´s about speed of relative movements between NV-devices and the scenes or targets these are looking at. And even this X27-device producing admittedly breathtaking footages, there seems to be something "fishy" about it´s results (or better: the achievement of those), because when this device is actuall excecuting some "fast" relative movements to the scenery, you can see the stars for example morphing into stripes, which is a hint this device, although already performing in mostly very well lit fullmoon condtions, additionally using "long" shutterspeeds to keep the noise-floor down and a phenomenon which would`nt happen with Analog in even much worse lighting-conditions (?).

Absolutely, the X27 probably has its drawbacks, size being one of them...  I don't know much about this device beyond the videos I've seen... but that gives you a glimpse of the stuff that we don't know about.

By Noise Floor you mean, EBI? Well, I personally like tubes with the lowest EBI possible for stargazing.

It is clear now that multispectrum is the way forward, aka. sensor fusion... having great I2 or thermal alone is no longer as big of an advantage as it was, say, 30 years ago.


G.

I bothered to do the duplicate photos in my time (as promised), and used a moon calendar to pick the least favourable time and even checked the exposure (New moon, 0.1%) - it was a 40km drive.  

Just getting the thread back on track - the comparisons are up above (also I've blocked gman).  

The data, as always speaks for itself.  

The spectral ranges of the different types of NV device are well understood.

The peak of the visible spectrum is around 500nm and the best a GaAsP tube can do is a QE of 50% (but is near useless in the NIR) and a typical GaAs G3 is down at less than 30% in the visible spectrum and tails off quickly into blue/UV.  

The performance of G3 tubes is skewed to the NIR, with peak sensitivity around 800-850nm (NIR), thats why most illuminators/IR designators are set around 840nm/850nm.  None of this a mystery.  

The Aurora Pro is a good device within is operating window, its peak performance being somewhat less than a modest G2 device.  The OPSIN looks like it will improve on this somewhat, how much is yet to be established.  

It should be noted the Aurora Pro is used by a number of government agencies both in the US and overseas.  

FBI Aurora Pro

This is no surprise to me as its an excellent surveillance and reconnaissance tool.

Any suggestion that SiOnyx is trying to con people is nonsense.

Conditions: starlight with some cloud cover
Note: water is a good reflector

Note: water is a good reflector

Note: OPSIN and Nightwave share the same detector

Same black silicon cmos but difference lens and "programming" between opsin and nightwave.

https://youtu.be/j2xCytF2YEc

Looks pretty good, now we need IRL testing with different NL conditions and framerates.

Let me ask, how much does the sensor efficiency and all the data shown in those fancy charts really account for in the overall system performance? B/c given how much better these chart imply the Sionyx really is, well, if the "sensor efficiency" and all those stats truly were the critical defining factors of performance here then the Sionyx would perform a better than at least a cheap Gen2, wouldn't it? Reality tells us that if the device is not as good as cheap Gen2, then the information in those charts is irrelevant.

Are you being paid by Sionyx by any chance? b/c you can't seriously insist in defending this device the way you are, not after admitting that its not even better than a cheap Gen2... unless, perhaps, you are being paid to do so.

My unsolicited advice to anyone who reads this while looking to buy a quality NVD scope, and happens to be on the fence: well, head to the EE in this forum (great resource, BTW) and buy a used Gen3 device there, even a FOM restricted (exportable tube) will blow this toy out of the water, and there will be NO disappointment afterwards, plus it will hold its resale value very well if you want to dump it for cash later down the road.

---------------

In regards to the OPSIN lens, sure, a faster lens will give some performance, but not much, even the best possible f-Stop improvement will only yield a couple dB to the input signal... at best.

G.

Thanks for the video.

Well, looks promising, but I am rather cautious at this point, b/c I saw the same kind of videos for the pro and the raving reviews and here I am now... so, without knowing the conditions its difficult to tell how much of an improvement, if any, it really is. That kind of image quality and fps is what I recall getting under some moon with mine...

I must point out tho, that the I2 video they used was the worst possible quality, and IMO not really representative of what the technology they claim to be better than really is. If you stick a thin film Omni VII next to that it will be crystal clear at that light level... Heck, I don't think even my Gen2 M911 looks that bad in NL5... but hey, its a marketing video.

I really wish they would've increased the sensor size..  at least it would've been useful as a daylight camcorder too.

The battery pack looks a bit on the bulky side IMO... and 8 hours runtime for a single monocular, means 4 hours for a bino... maybe haul a bigger battery on a backpack? And what is the runtime of the unit without external battery?

G.

Well look, at the end of the day they want to show their product performing good to customers. So showing actual practical results are not the best approach at first for a company regardless of how bad we want to see actual performance.

I too think they might have done a few tricks, maybe the footage was recorded at 30Hz to minimize noise and boost the brightness. Although it may vary slightly with FW updates, just like the Aurora.

As for the dual tube version that they are going to release in the future, I think they'll definitely think of something. 8 hours run time is not a lot for most people, and 4 is even worse.

Perhaps, wait till its released and tested by professionals.

Sub 1 mlux performance is good.  0.5 mlux capable cameras cost 20k.

This why (I imagine) no one takes you seriously - you already know a 15K Anker, that fits in a Mohawk with room to spare, and is light AF, can power a Mono for 3-4 eight hour nights. The 10Ks are even smaller, and you can fit two in a Mohawk. Numerous Analog Binos accept remote battery packs, as do COTIs and HMT Thermals - so using one, isn't a "ding". As far as internal - it's equal, or longer, than most Thermals on internals.

Your use of terms like "toy", "slide show", "backpack, etc... is what does your credibility in, on a tech forum - the bias is almost palpable. JMO

So, just because my opinion is different, and I express myself different, then nobody takes me seriously... is that what you are implying? so I have to "be a good bro, and take it in the rear end, play ball... etc"

The current Sionyx is a toy, and its my opinion. Why? b/c it doesn't perform as good as old Gen2. and if something isn't as good as Gen2 I call it a toy. Now, the OPSIN, we have yet to see... but I was hyped up and let down by the previous device already, so this time around I much rather be very cautious and be mistaken than hyped up and let down yet again.

All I made was an accurate observation, and when I said "backpack" I refer to the fact that I do not want to haul battery packs in my head, and I much prefer to put battery packs on a plate carrier or on a backpack than in my head, or are you implying now that nobody should take me seriously b/c I suggested putting the battery pack in a backpack now?

Regardless of what you and I think, the fact remains: the device really requires an external pack to have any decent runtime. And if you want binos then you really need that a large pack... I also understand the Anker power bank is not that big (I own similar packs for other applications), but its just one more thing to haul around, and for me, its rather inconvenient to say the least.

I hear no comments about the piss poor quality Gen3 video chosen in the OPSIN video, because yeah, its perfectly fine to show a washed out scaled down image of the worst image intensifier NVD device ever built... and call it a day...  this is the kind of shady tactics ATN and those 90's resellers used to push to sell their Gen0 stuff. I've seen this before. I understand they are trying to make a point, but comparing mature MILSPEC tech that has been used to fight wars since the 70s, to a device that has never been field tested or in used in any combat is just preposterous... Find a stat in which our device is better, and make it all about that feature, slander and discredit all other opinions.

Then, comments like "you've never owned one" "no one takes you seriously" or "have your vision checked" strengthen my resolve to make my opinion heard, in hopes of steering people clear of this thing. I've come to the realization that there is one worse state of mind than being ridiculed, its called being disappointed. So, I guess I'll have to endure ridicule in the hopes it helps save someone else from the disappointment, and the money too.

Speaking of disappointment, yes, I am a very disappointed former customer, so of course I am biased. B/c why wouldn't I? I don't say good things about something when my experience with that something was bad.

For less than the cost of those two Sionyx that someone kept around, I've put together another Gen3 Omni VII PVS-7... which can see perfectly well down to NL5+ levels no problems... and doesn't need charts or special pictures to justify my purchase.

G.

Great video, look forward to seeing some real world tests. Hopefully the supply chain issues are resolved and these will be available soon.

AlphaOneDesigns on IG, just posted a Sionyx produced video (I assume) on the OPSIN - a lot of details.

@mickdonaldson

Mick have you got a link for this?

I reached out to Bill B at SiOnyx to see if there was any further information given the apparent radio silence.

Direct from SiOnyx, the OPSIN will be released in October - the date is as yet to be finalised.

I don't - just saw it on that guys page. October 31st it is!

I have a paper with the spectral response of what is most likely the sensor going into the OPSIN (2021 paper so data matches), and the QE improvement is no where near what you are suggesting.  It also includes pictures of indoor images at a supposed 1 mlux level, but no info on the spectra of the lighting, so potentially a very misleading image. QE is not nearly the most important thing for digital sensors.  It's all about the read noise/dark current, and it's very difficult to find any information on noise levels for Sionyx products.

Either way EBAPs is the near future and had already clearly surpassed tubes in the early 2010's at low/mid light levels even in high subjective pilot flight testing.  The longer term future is mutli-spectral tuned MCT or InSb sensors.

Bingo, misleading is the name of the game with Sionyx.

Yep, totally agree, QE is clearly not the most important thing for this device, not when it can't even match a lowly Gen2 with supposedly crappier QE...

G.

Looks like EBAPS is already in use by the military...

https://optics.org/news/11/1/24

I have no way of knowing what the QE of the new detector will look like, just an educated guess - most likely they will increase the area under the curve by working at the margins rather than the peak.  There are some good gains in irradiance at longer wavelengths that can make a big difference to overall performance.  In fact the integral of the irradiance peaks at ~ 1050nm an 1100nm provides more ~ 4 times more light than the whole irradiance spectrum between ~350nm and 900nm.  So just increasing the QE in the range from 1000nm to 1200nm could provide good rewards.  

Quantum efficiency is a measure of a detectors ability to turn a photon into a electron.  The ability to do that with high efficiency over a large useful spectral range is important.  Another metric often used, is sensitivity which is measured in mA/w and as with QE the ability to have high sensitivity over a useful spectral range is important.  

The integral of QE or sensitivity directly reflects the potential of a detector.  The issue that EBAPS faces is restriction of wavelength.

I'll get one jut to see what it does - not sure what the delivery schedule will be.  However, if I can get one quickly - I'll run it up against the Pro and a cpl of decent IITs.  :)

If you mean by wavelength at the input side, well, that can be changed by using a different PC, according to Intevac's documentation. All the way from UV to SWIR. The EBAPS already seem to outperform I2 devices, probably due to lower noise since there is no MCP, so I don't see any issue there...

EBAPS appears to be the EO camera in the Apache helicopter, and currently in development for the US military under contract for pilot goggles, as opposed to Sionyx which isn't. So if you want to be combat effective, follow what the military does... EBAPS is.... someday... $$$$

G.

Can you post a link to the paper?

TFB posted some images from Shot Show in January that were allegedly done with 1mw of illumination.

Awesome - U da man.

Who said that or where had it been announced a new sensor going into that Opsin-device? To my knowledge Sionyx proprietary sensors are gathered in the XQE (enhanced quantum efficiency)-family which consists of the XQE-0570, -0920 and -1310 sensor(s), with the 1310 used in the Aurora-actioncam-series and no successor to the 1310 anywhere to be mentioned.

If so, then please consider to upload bigger versions of your scene-shots because these former "postage-stamp"-types made it hard to really judge and compare the content(s) presented in them. And if possible additional data on the centi-, milli-, microlux actually present in the scene which would be more helpful than the usual lightlevel-marker "can see the penis/cannot see the penis" ...

(precautionary: this post, especially the penis-part, was meant to be non-offending)

No offense taken... :)

The sensor isn't capable of more resolution, hence "postage-stamp" it is what it is.

G.

Depends on if you're interested in spec arguments, or if you want to go off what the human eye is actually seeing while in use. The best way, for most end users to judge, is to take both analog and digital units out on the same night, put them in a side-by-side rig, and record through both occulars with identical cameras/settings. Doesn't matter what the xLux or NL is, just compare to the analog unit. This guy does that:

Carpe Nocturnum

Even then, do a search on Google like -  site:ar15.com "it looks way better in person" and read the 100's of posts from analog guys saying just that - it definitely is a thing. Taking that Sensor data, and displaying it on a 0.39" 768p OLED (before Occular Lens Mag) gives a way better real world presentation, than a posted 720p image/video. Even GHOz' alleged "postage stamp" images, look worse than they did in real life.

No doubt. But, at least me, i have no Sionyx to walk and compare it, whichfore i have to ask for/rely on third-party material of which i hope tries to do the best job in teleporting the original scene to the viewer and regarding this purpose, i think the pics GroundhogOZ is uploading are a bit too mousecinema-like with additional comparison-pics of for example of an analog device squeezed into and thus even smaller than the main pic.

I´m familiar with the nocturnum-channel. But not been there for some time. Thanks for reminding me of it.

That dude has an amazing channel/skill set - he's on IG too. As you can see from the vids - he's a fan. Fair enough on you needing pics to make determinations - for some reason I thought you had one at some point, my bad.

The newer XQE1350 silicon is listed in the Opsin spec sheet.

See here:

https://tvprzemyslowa.pl/data/links/d74e6ffbd405199b41fbfbc1d0fb0a18/26004_4188.pdf

It’s also mentioned on the Nightwave product page

https://www.sionyx.com/pages/nightwave

There’s a spec sheet for the OEM version here

http://www.militram.com/oem-light-0-0005-lux-image-sensor-digital-night-vision-sionyx/

@mickdonaldson  Unfortunately there are no easy answers, for example the images from the SiOnyx devices port out as 1260x720 JPEG images with a file size of ~ 300k.  The implication is they are significantly compressed in the first instance.  Then they are uploaded to a hosting site, which in turn ports to AR15.com - in other words the images are compressed again resulting in further degradation.  

The best solution is direct side by side comparison (because the degradation is the same for all images on one "slide") as you suggested but even then there are differences in resolution that are simply a function of the different technologies e.g. IIT ~ 5MP equivalent Vs thermal or Pro, both of which are significantly lower resolution.  

Another point to note, there are a lot of photos taken down IIT in the NV forum, most are in relatively high light conditions e.g. back of the house into the yard or down the street type photos or peripheral to towns with lots of skyglow.  What these photos really reflect are lp/mm variations of the different devices and across the FOV and halo rather than the inherent quality of the device down at NL5.

I have always taken the approach of trialling the different devices in trickier settings e.g. where light levels are very low and in the case of thermals, where the thermal contrast is relatively poor.  The result is less than perfect images as you would expect.

From the link provided by @boatswain (1mW is approx 4mlux over 25m^2 - 4mlux is NL4, so moonless starlight equivalent)

We are pleased to introduce a new OEM Low light (0.0005 Lux) image sensor for digital night vision.
The CMOS sensor was chosen by the US Army’s Night Vision and Electronic Sensor Directorate (NVESD) as the night vision camera for digital night vision for the Integrated Visual Augmentation System (IVAS) project.
We can offer the camera module in several configurations as board level or with a C-mount lens holder.
The main features and specifications are:

·       2nd Gen SiOnyx extreme low light technology
·       0.5 mLux Moonless Night Performance
·       anytime safe operation
·       Monochrome or Color
·       2D noise reduction
·       Auto exposure / Auto gain
·       Uncompressed RAW video output
·       MIPI CSI2 output interface
·       Serial and i2c control interface
·       Unmatched low power consumption
KEY PARAMETERS
·       SENSOR SiOnyx XQE-1350 1.3MP CMOS
·       MINIMUM ILLUMINATION 0.5mLux @ 30 fps and F1.28
·       MOONLESS NIGHT PERFORMANCE 1X – 3X
·       OPTICAL FORMAT 1-inch (15.7mm), 9.5 micron pixel size
·       AUTO EXPOSURE RANGE 22usec to 500msec
·       VIDEO OUTPUT MIPI CSI2 – 4 lane – 400 MHz
·       VIDEO TYPE Uncompressed 12 bit
·       IMAGE LATENCY < 120 microsecond
·       SHUTTER TYPE Progressive, Rolling
·       Video Resolution 1280×1024
·       Frame Rate 120, 90, 60, 45, or 30 FPS
·       Connector Flex cable
·       Comm Port UART serial and i2c
·       IR Cut Filter Manual filter (not included)
·       Lens Mount Circuit Boards only
·       Operating Temperature -40°C to +49°C
·       Power Supply 3.3, 1.8, & 1.2 VDC
·       Power Consumption 1.05W @ full frame rate
·       Dimensions (mm) 25(w) x 25(h) x 32(l)mm
·       Weight <34 g

That spec gives an ACQUIRE model D:R:I for a human sized object of 1140m, 280m and 140m whilst maintaining a good FOV (50% probability)

If you force the model to 90% probability with a FL of 15mm you get a D:R:I of 590m, 150m and 70m for a FOV of 42 degrees x 34 degrees.

Well, 1.05 Watts at full frame rate just for the sensor means that a Lithium AA battery, with about 5 Watt hour (3200 mAh, 1.5v nominal)... will run the sensor for ~5 hours, but now the question is: how much does the rest of the electronics consume? I am sure it will be more than 1.05W... runtime will be less than ideal.

In addition to poor battery runtime without requiring large battery packs, @1.05 Watts also means there will be a LOT of heat dissipation at the sensor, so this device will never be able match I2 to see below certain threshold, will have lots of noise (see photos) (or even see at all) b/c its just too hot (in I2 this would be roughly equivalent to having very high EBI).

This is why high quality digital Astrocameras have active cooling peltier elements, b/c when you cool that CMOS sensor down to -20C then it can see really well even the faintest DSO stuff... but again, active cooling is possible in vehicles... not really practical for on-foot.

G.

Thank You! :)

Thank You for these explanations.

And how about taking pics through the ocular of the Sionyx? Which would be comparing apples to apples, (analog)screen to (pixel/digital)screen, the screens which are the main human/machine interfaces and which deliver the final user-experiences (although i can imagine this to be tricky with the fiddly Sionyx-ocular)?

But: I think it was the carrier of the Sure Shot Night Vision-channel who took the effort to conduct trough-the-(also fiddly)eypiece videos of the Iray-MH25 and the results have been an outstanding representation of what the user REALLY gets to see on that screen.

Which min. f-stop does the Opsin provide?

Hi Schrotthaufen,

i found f1.4 on different sources.

Brother, I'm a semi-technical, end-user - my eyes cross when I read most of the slide-rule micron sh*t I've always found your images to be close to what my eyes see with digital NV and thermal - no complaints from me. As is often the case on the internet, there is a lot of "teach" vs "do" in this thread. Now if you'll excuse me, I have to top off my 2.5" x 3.5" 6.7oz Anker 10K (L3 Harris 31 Batt Pack is 7.5oz) HMT Bump Rig, for an upcoming shoot. I wish I didn't need to also add about twenty 220gr .300BLK rounds to the Mohawk as well, to balance it out - but such is life.

Holy shit if Mick is semi-tech, I'm a no-tech, monkey with the AK.  I have no fuckin clue what you guys are talking about here.  But I like looking at the pictures.  Just show me which way the fucking batt goes in and I'm off.