Those of us who bought or are thinking of buying the X-sight, have been following along in the threads detailing the various bugs, firmware shortcomings, and power consumption of this new product pretty carefully. But I wanted to post something about the other side of the story - why you might want to hang on to, or even buy, one of these units. I think a lot of guys are very focused on the flawed rollout of the product, and the inconsistent firmware. It's just not a complete product yet, and some folks expect that when they plunk down their money (and rightly so). But not so much with the younger folks though, since most of them have been buying electronics like this for years, and fully understand the vagaries of "firmware". And the scope is frustratingly usable as is, thought it seems reasonable to expect it to be a lot better in the near term.

First of all, while this is not the first video-scope to market, the use of an Ambarella chip opens a big door, and makes the x-sights potentially a disruptive technology (meaning, what people expect and want is going to change because of it). This wont be the last consumer video scope you see, not by a long shot. Expect smaller and more powerful digiscopes, with more and more features, and lower and lower prices.

In the near future, perhaps 3-10 years, I would expect to see CMOS camera sensors running in 4K mode, that can "see" visible light and long-wave IR on the same chip (the X-sight "sees" short wave IR and visible light) at a reasonable price point. OLED displays are getting more and more dense as well, so viewfinders will be clearer and more accurately reflect the actual image the camera is seeing, in very small form factors. And the smaller the chips get, the easier they are on batteries.

I can envision a "scope" about the size of a pack of a cigarettes that can "see" thermal at night or day, provide magnification ranges from 1x to 15x clearly, range a target, automatically adjust the aiming point based on the load fired, inclination and tilt of rifle, weather conditions, spin effect, coriollis effect, whatever. It could identify target species, detect motion across a wide angle FOV, overlay colors onto "hot" items in the display. It could auto-compute the lead and holdover if you are tracking a moving target based on your actual swing speed, keep a database of your favorite loads, and automatically set the zero for each load. It could eliminate mirage, analyze your technique, and play your favorite porn in a picture in picture display while you hunt. Nearly anything an iPhone can do, these package camera systems can do. It opens a lot of doors (some awesomely useful, some incredibly stupid, but choice is a good thing).

I've spent a little time looking at the unit now, and it's clear that its based on an Ambarella system-on-a-Chip, just like the popular GoPro cameras, and a whole host of advanced security cameras. It also contains additional sensors with perhaps unknown capabilities, WiFi, and GPS. At minimum it likely includes a magnetometer chip (compass, hopefully 3-axis). Accelerometers might be useful as well, but have seen no evidence they are included. I sure wish they had included a laser for rangefinding, but hey, it's a first gen product.

That's just a few things,there are more. Whether you want to go this way or not is up to you, but if you want to get on the cutting edge, expect to bleed a little. It's not a place for those lacking patience. On the other hand, gearheads should love this stuff. I do.

With that said, here's what I think ATN should focus on to get this scope where it needs to be, and to make shooters genuinely excited about what it can do - even if it may be a while before all features can be coded and debugged. And this is low-hanging fruit with the existing hardware. Are you listening Sachal? Fix the bugs on existing stuff, and gimmee some goodies!

1. Add a toggleable Shoot mode and Scan mode. Tailor the information displays to what the shooter is doing, rather than trying to cram all modes into a single screen (overlay). Maybe move the camera functions to the menus only, and tie the shoot/scan toggle to the left arrow (this is first and foremost a scope, not a camera - priorities!). In Scan mode, turn on compass, image stabilization, and any other feature appropriate to the mode. Scan mode is where you hit the shooter with lots of info. In Shoot mode, turn off most info, turn off image stabilization, and turn on zero auto-holdover for a shooters chosen range (if shooter configures the feature "on"). In shoot mode, the less distracting crap on screen, the better, Shooters should focus on shooting, not gadgetry.

2. Add real ballistics calculation - First, use Cherokee over the Wifi link, to allow us to easily enter data into the scope, select things, etc. Entering ballistics info through the keypad would be a complete pain, so just accept that shooters need a phone, pad, or PC to set up the ballistics functions. And webpage setups are dirt simple and don't suck up NAND space. Let shooters enter their gun setup - height of scope above bore, rate of twist, bore diameter, etc. in one area (system setup?). Add another page to enter load data, and save it right there on the terminal. Let us name each dataset (ex. "Hornady 120g SST"). Gather up muzzle velocity, BC, etc. for each load. Number it, and limit the number of slots to something manageable like 10. When the shooter wants to select the load for use, he just saves it, and the data is pushed into the scope's corresponding numbered slots. Let the scope store data and zero for 10 scopes. Shouldn't require much memory. Once a slot is made active on the scope, any user changes to the fundamental zero point, are saved to that slot. That way, selecting a slot recalls the zero set by the shooter, and loads the ballistics calculator data. To use, a shooter just menus and picks which slot to use, the sets the range to target manually using a text roller or whatever looks good from the SDK. Tie the set range function to the right arrow. Maybe draw small hashmarks on the vertical crosshair to indicate other calculated ranges besides the currently set range. (-100, -200 -300 +100 +200 etc). I know you guys are working on something like this, but hey - trying to get my dibs in for the way *I* would like it to work!

3. How about using that 3-axis mag chip, and giving us a scope cant (level) meter in shooting mode?

Yes, we all want the scope that does everything we want it to do, with the longest battery life, smallest size, lightest weight, easiest to use, best glass, rated for beyond .50 bmg recoil, at the least price.

Why not make every scope be a merged NV/thermal Trackingpoint scope, but with programmable custom reticles?

Nobody can afford it.

While there are frustrations I am in 100% agreement with your comments.

Not a Tech guy, are you.

Tapper -

What convinces you that the X-sight has the chip you aare referring to? Got pictures of it? It is possible that the coding is shared but an alternate chip is in use, either way you are the second to say it has it so just looking for undisputed fact.

As for people expecting teething pains due to it being tech based and correlating that to other devices. I say caution is a must because this particular device guides a bullet downrange with the intent to kill its target. Glitches are hardly acceptable in my opinion and hacking (homebrew) of the scope may actually present an opportunity for laws to be made against the act.

I am in agreement with you on technologies development and what the future could look like. I actually hope it doesn't pan out in some regards as I feel animals won't stand a chance against novice hunters and if missing becomes less frequent than learning may occur less often. Also I don't see police (LE) allowing the tech to get that sophisticated for obvious reasons.

I like your built in level idea.

Not a tech guy here either by the way

HTXH

Use Notepad on your computer, and open the xsight.bin file with it. Scroll a bit, and you'll see why we know it's an Ambarella board. I looked at the chips, and based on what I saw in the cut sheets, I'd hazard a guess it's either an A5 or A5S package. Could be a later REV. The code mostly ports between REVs. I think they added a 3-axis magnetometer on the i2c or GPIO bus. They're using an 800x600 OLED for the viewfinder, probably connected on the boards HDMI port. Dunno what camera they are using, but it's probably a vanilla CMOS chip with the IR cut filter removed, or possibly servo'ed with another filter. If it was something special, it would show up in their marketing pitch.



I can't disagree with that, but likewise it is reasonable to expect that any first generation electronic device will experience teething pains. Risk vs. Reward. And someone has to be first with it. Hopefully ATN understands the need to get good solid code in these units. I would be heartily surprised, if their geeks weren't being sweated like hothouse chickens right about now. And technology is the great democratizer, it's not just what this unit can do, but what the price point does - brings reasonably priced NV to guys that can't afford $3,000 thermals. It can also significantly extend the range, of an average shooter that doesn't know an MOA from a Mil from a mosquito. You're going to see $200 dollar versions of this unit coming out of China post haste.

I can't speak pro or con about the liability of wh4ck1ng up their code, but it would take a long time, and it would be much faster to just write your own code from scratch.



Actually, I think the statistics on the thing will probably continue to bear out, that novice hunters will still perform like novice hunters, regardless of the current state of technology, and for some time to come.



Well, I make my living pushing bits, but don't hold it against me. I like to shoot hogs too..

I'm 100% on board with what Tapper's saying here, and I hope ATN is listening to his ideas, because they are spot on.

USER INTERFACE is hugely important, so dont just think about features, but how to make them USEFUL to the guy out in the field.

I'm also glad there's a counterpoint to the doom and gloom crowd. When I have more first-hand data to add, I certainly will.

Kinda hoping other folks have some ideas for this sight.

In the future, If ATN could offer a custom firmware tailored to their customers needs, that would be great.
I don't need all the bells and whistles. AkA- WIFI, Compass, Geo Tagging, Clock Etc....
I do like the idea of the BDC reticle, as I plan on mounting it on a .260 one day to do some longer range shooting in Day mode.
If the newest firmware update takes care of the disappearing reticle during zoom, is bright in NV mode and has the reticle
while recording, I'll be satisfied.
When that day comes, I'll buy a 3x unit too.

Oh, I have all kinds of ideas for the X-Sight. Like. Tons. I've just been berated in other threads for talking about it before mine arrives, so I'll stay put till I actually have it, and can post with "authority".

I'd just like to see custom or customizable reticles.

I'd like to have a dot sight like my crossbow has with 3 different dots for 30, 40, and 50 yards.

2. Add real ballistics calculation - First, use Cherokee over the Wifi link, to allow us to easily enter data into the scope, select things, etc. Entering ballistics info through the keypad would be a complete pain, so just accept that shooters need a phone, pad, or PC to set up the ballistics functions. And webpage setups are dirt simple and don't suck up NAND space. Let shooters enter their gun setup - height of scope above bore, rate of twist, bore diameter, etc. in one area (system setup?). Add another page to enter load data, and save it right there on the terminal. Let us name each dataset (ex. "Hornady 120g SST"). Gather up muzzle velocity, BC, etc. for each load. Number it, and limit the number of slots to something manageable like 10. When the shooter wants to select the load for use, he just saves it, and the data is pushed into the scope's corresponding numbered slots. Let the scope store data and zero for 10 scopes. Shouldn't require much memory. Once a slot is made active on the scope, any user changes to the fundamental zero point, are saved to that slot. That way, selecting a slot recalls the zero set by the shooter, and loads the ballistics calculator data. To use, a shooter just menus and picks which slot to use, the sets the range to target manually using a text roller or whatever looks good from the SDK. Tie the set range function to the right arrow. Maybe draw small hashmarks on the vertical crosshair to indicate other calculated ranges besides the currently set range. (-100, -200 -300 +100 +200 etc). I know you guys are working on something like this, but hey - trying to get my dibs in for the way *I* would like it to work!


this is all out there already in the burris eliminator scopes, not as intensive as what you listed but with the ballistics chart and 4,000 loads to enter you can pretty much program any load out there, even hand loads. it would be nice to program in the round like the burris and get a holdover dot similar, simplifies things immensely. I have a burris eliminator III and am about to order a 5-18 x sight (hoping I am tech savy enough to be able to use this thing) as a trial for possibly buying some to have on hand for our paid pig/coyote hunts.

how hard would it be to create an app separate from the scope that has all of this that tells you where to manually move your X/Y coordinates? I mean its all basic math with constants right? if I know the formulas I could whip something up in windows. so ok you zero in at 100 yards x/y is constant for 100 yards just figure out what + or - equates to from original poi and then its simple math right?

It wouldn't be that hard, but would take a bit of time to write it if you had to build a user interface for it. The equations all exist online. Interesting sidebar - How many pixels per MOA at 1X? 5X? How does that play out when shooting at longer ranges?

From what I understand, it's not that simple. Sachal alluded to the fact that the picture the sensor sees has to be calculated based on the lenses (IE: you CANNOT set your 3x scope to the 5x firmware setting, and have it work correctly)

However, there's other factors involved, especially with longer-range shots.
- Ballistic Coefficient
- Wind, crosswind, double crosswind?
- Bullet Spin Drift
- Coriolis Effect (spin of the earth)
- Temperature (affects burn rate of powder)

Now, that doesn't mean ALL of this needs to be calculated. The shooter still has plenty of work they need to do, and I worry that many shooters will blame the SCOPE for not being smart enough - when really, it's a lack of their skill or a variable they didn't account for.

That said, this is why I also want CUSTOM reticle options. We need reticles that include BDC & Drift marks, to help shooters with these calculations in their minds - due to environmental changes beyond the scope's ability to compensate - without giving the shooter information overload.

I've been thinking about the Pixels->MOA/Mil conversion idea.

I think we can easily MEASURE pixels per MOA/Mil at set ranges (25, 50, 75, 100 yd), using very simple printed graphics.

Just print out a sheet with a 1" grid, using a central point for reticle aiming.

Start video recording at full HD, and step through the magnification ranges.

Then you can analyze frames of the video to gather this info. This data could then be used to create a database or spreadsheet, and from that, we can develop an equasion to calculate this.

To ensure it's accurate, we should pobably have multiple people perform this standardized procedure, and compare the data to make sure the formula works.

More importantly, once you know this data... how do you APPLY it in the field? We'll have to convert the data/algorithm so that it's useful to the guy looking through the eyepiece, at the reduced-resolution 800x600 display.

I think this is the only way to ensure that a custom reticle or something like that will have the intended effect.

After seeing several xsights over here in the Carolinas, I have had quite a few compplaints about the scopes not holding zero.
Each and every one of them have come here with the mount screws loose. After talking with the people via pm and phones they swore that the mounts are tight only to see them not when they have wasted gas and time for a 5 minute fix.
CHECK your mounts and if you havent already, take the screws out and put some loc-tite or even the wifes nail polish on the threads to keep them from backing out.
If you notice your gun not grouping check that first.

Has anyone compared the view of the updated 3x and 5x scopes here ?
I saw a comparison on U tube, but it was with the old version 3x scope.
If ATN gets their act together, I'd like to buy a 3X for my PCP air rifle for close work and put the 5x on a .260 for further ranges.

shoots100,

I don't have a 5x to test for you, but my 3-12x is en route right now, and will be arriving in two or three days. If it's possible I'll try to get you some pics from it so you can compare. The 5x views you've seen on the web should be pretty representative, because those did not need to be re-manufactured.

I plan to post here with my impressions/assessments once it arrives and I put it through some paces.

I swapped out my 5X for a 3X, just a few impressions after one time at the range. Sorry, forgot to take pix, try to get some soon.

The 3X is significantly smaller and lighter. I'd guess maybe a half to 3/4 pound. The magnification appears to be pretty close to a normal 3X, and the FOV is much wider. A much better choice for hog hunting at typical Texas ranges.

The focus works, the ring is stiff as hell, so I'd suggest working it back and forth a few hundred times before mounting it, to loosen it up. You have to use it often on a low-mag scope like this. You also need to set the diopter right or it won't focus clearly.

Just as with the 5x, I'm having problems getting my rifle to group. I'm going to slap an optical on it just to verify that the gun still shoots a little better than MOA, but I suspect the mount. I did tighten and locktite the screws before I put it on, but I wonder if the combination of ill-fitting weaver lug, and mount itself may be problematic.

Otherwise, it functions just like the 5x.

Thanks guys. I've had a 5x for sometime.
Most of the videos out there are from the older 3x scopes and they suck.
After seeing some videos of the new style 3x, I'm know it's going to make my wallet a lot lighter.
With the wider view of the 3x, it will be perfect for the close range shots I'm presented with while hunting with the PCP air rifle in the suburban woods and fields of NY.
That's if the newest firmware version works and users don't have a problem with it holding zero.
My 5x has had no problems holding zero.
I had it mounted on a .20 practical and now a PCP air rifle, and both rifles have practically no recoil.
What rifle do you have your scope mounted on Tapper ?

I have access to both a new 3x with the adjustable focus and the 5x. I bought the 5x with my Cabelas points, they didn't have the 3xs at the time. My son liked the scope enough he wanted to get one for himself for our upcoming hog hunting trip. I suggested he get the 3x because it would have a greater field of view and it's $100 cheaper. Thankfully he took my advice. (they tend to do that more as they get older)

Not sure when I will have a chance this week to do a comparison, but I think the 3X is a much more usable platform unless you are doing some real long range shooting. The 3x has well over twice the field of view of the 5x, and the digital zoom still gets you in close if you need it.

Funny part is my 5x sits on a 16" carbine and his 3x is on a 20" heavy barrel rifle.

Ever see a scope that didn't advertise the field-of-view? Good luck finding it on ATN's product info. I've looked in vain for it. It's an extremely important number for a digital scope. Here's why:

First some simple facts:
1. The sight itself is actually a fixed magnification sight, since the optical zoom never changes.
2. The camera sensor (pupil?) is 1080p (16x9), or 1920x1080 effective pixels. This never changes, and is only viewable in recorded video.
3. The eyepiece viewer is an OLED chip with 800x600 (4:3) resolution, therefore its image must be cropped and/or interpolated. In most cases, the camera image will be cropped horizontally to fit the 4:3 screen ratio, then interpolated. If it isn't interpolated, it will digitally zoom the image after cropping.
4. There are two FOV measurements that matter - the camera FOV, and the viewfinder FOV. One effects accuracy/zero ability, the other, how hard it may be to get on a given target while aiming.

Let's say, for argument, that the FOV of the 5X x-sight is 16' of horizontal view at 100 yards. 16' x 12"/ft = 192inches. And at that 100 yard FOV, there are 192 MOA of horizonatal arc, and 10 pixels per MOA. If you zoom to 2X, there will be 5 pixels per MOA. At 3X, 3 pixels per MOA. And so on. At 1x, an object seen at 200 yards will subtend 5 pixels per MOA as well.

As you can see, at this FOV, we can light about 5pixels/MOA at 200 yards, 3 pixels/MOA at 300yds, 2 pixels/MOA at 400 yards, and so on, and this introduces significant aiming error at longer ranges.

Bear in mind - digital zooming just means drawing a smaller and smaller box around an area of pixels on the camera sensor, then displaying it on te viewfinder, which has the effect of making the viewed image larger (but does not add resolution to the view - and this is why the images gets "pixelated" at higher magnification)

So what's the real FOV? Only way to know, is to measure it at 1X, and doing this at 100yds simplifies the math. The Rule of thumb - the wider the FOV, the shorter range you can use the scope before aiming error begins to become a real issue.

A typical optical 3X scope would subtend about 32 feet@100yds at 1X (about 4 pixels per MOA).

You can also immediately see why higher resolution camera chips would be highly desirable.

Note: this is true of all digiscopes, not just ATN's.

Tapper,

I think there is one problem in your math.

If the camera is capable of 1920x1080 and the display only shows 800x600, you should be able to get 2x digital zoom before you notice any degradation on the view finder. So I think it is more like 10 pixels per MOA at 1x, 10 pixels per MOA at 2x, and 5 pixels per MOA at 3X digital zooms.

Well, I would never discount the possibility of an error in my math, but it's important to remember that the viewfinder image is just a representation of the true image at the camera. From the point of view of aiming, it only really matters what the camera can resolve (assuming the image at the viewfinder is interpolated correctly. At somewhere more than 2X, the viewfinder should match the camera pixel for pixel with 80 pixels cut off on either side of the image. ((1920/2) -800) /2