New AX Visio 10x32 binocular (2 Viewers)

[Dyrlege]

Good question. My money is on OM Systems being first out.

Don't forget Panasonic...

 

[wllmspd]

Training AI systems requires a huge load of processing that’s been done for you using vast numbers of training images, the device is just running the output algorithms to do the ID. So I’d expect they could be updated to improve the ID on device. If the binos uploaded all the images to swaro/merlin then those could be used to further refine the training (not suggesting this is done or even an option).

Peter

 

[Jason Bugay Reyes]

Or less: i got a (not very) used G9 with 100-400mm telezoom (micro 4/3 so equiv to 200-800mm in 35 mm) and a 14-140mm quite reasonably. An MM3 or MM4 scope at 60mm (or 50) with a zoom eyepiece is in my opinion a good starter, but there are lots of "views" on that matter in various threads. A good binocular is a must.
Cheers
Per

Agree with the 12x42 NL!

 

[Dyrlege]

Agree with the 12x42 NL!

I was suggesting keeping prices down!😂

 

[Dyrlege]

Training AI systems requires a huge load of processing that’s been done for you using vast numbers of training images, the device is just running the output algorithms to do the ID. So I’d expect they could be updated to improve the ID on device. If the binos uploaded all the images to swaro/merlin then those could be used to further refine the training (not suggesting this is done or even an option).

Peter

I'm quite sure Swarovski will follow up use of their app closely and correlate with Merlin, and exploit user experience to improve and further develop the algorithms.

 

[james holdsworth]

There’s a big difference in using information sources (and your own skills) as an aid in identification, both by sight and sound, than a machine telling you what it thinks it is. Sorry, but that kinda spoils the whole point of birding for me.

The premise is a great learning opportunity, as long as the observer moves past the point where they rely on the AI to do all the work. So far, most of the people I know that run Merlin a lot, have not made that move, and may never.

I started to get into plants a few years ago. It would be much quicker to snap a pic and use AI for the ID, but I find I learn more (and more quickly) when I have to do the sleuthing myself. Much more satisfying as well when I finally make a conclusive ID of something new.

 

[tenex]

The premise is a great learning opportunity, as long as the observer moves past the point where they rely on the AI to do all the work. So far, most of the people I know that run Merlin a lot, have not made that move, and may never.

This isn't just laziness, but lack of guidance. Not only does Merlin not explain why it identifies a bird as a particular species, no one can know quite why it does, not even its programmers. It's only been trained on millions of images/recordings identified by humans, so its usefulness is quite limited as a learning tool. There won't be a Visio II that points helpful arrows at distinguishing field marks. People really don't understand how different AI is; it has the neural net, but not awareness or reasoning.

 

[Dyrlege]

There’s a big difference in using information sources (and your own skills) as an aid in identification, both by sight and sound, than a machine telling you what it thinks it is. Sorry, but that kinda spoils the whole point of birding for me.

The premise is a great learning opportunity, as long as the observer moves past the point where they rely on the AI to do all the work. So far, most of the people I know that run Merlin a lot, have not made that move, and may never.

I started to get into plants a few years ago. It would be much quicker to snap a pic and use AI for the ID, but I find I learn more (and more quickly) when I have to do the sleuthing myself. Much more satisfying as well when I finally make a conclusive ID of something new.

I'm sure it is, and for the less gifted, as I obviously am, it does exactly the same as a GPS (checked by map and compass), an ultrasound (checked by clinical examination), any test at all (tested according to sensitivity, precision, accuracy and specificity and including both positive and negative predictive ability), a red-dot sight, in fact any AID to navigation (in the widest meaning of the word and including printed matter). I should also include species splits because of small genome differences (oops, humans suddenly split into 50 different species, but it does for birds, so...).
The AX is an aid, just like a field guide (I've got plenty) or the song recognition apps and data files, or a medical test or a GPS-powered course or whatever.
It gives me photo documentation while I'm using a quite nice (in the proper sense of the word) pair of binoculars, position, time and , yes, a hint to ID (which is up to me to use, confirm or deny).
If you can't stand it, get on your horse, go back to your typewriter (I use pencil and paper!) and write a complaint.
If you don't like it, don't buy it. It's that simple.
Oh, and by the way: I did plants, too, had to use a herbarium as well as field guides, oh, and a microscope. Now I have an additional electronic guide that ASSISTS me in sorting them out. Great when collecting mushrooms, but I always make the decision on species myself.

 

[james holdsworth]

Don’t take my comments as a criticism of AX adopters, I said several times I think it’s innovative but with options I’d be unlikely to use. I am reading every comment, and your user based experience is valued and appreciated.

 

[Dyrlege]

Don’t take my comments as a criticism of AX adopters, I said several times I think it’s innovative but with options I’d be unlikely to use. I am reading every comment, and your user based experience is valued and appreciated.

Oh, I don't! But I do insist on it being an AID to ID, and a valuable documenting tool. My point is that allowing an instrument to define truth is still (and hopefully will continue to be) not acceptable, it's an aid (just like the GPS) that runs a test and delivers a result that must be evaluated and verified, just like any other (!) test. And as I familiarise with it's abilities and quirks I find it to be helpful. I think the analog-digital combination is exciting, and it is not just for birds!
Late edit: I forgot to mention it takes a photo every time you push the "button" to ID an object, and it stores the photo with GPS data, timestamp, bells and whistles.

 

[Patudo]

Close to 1100 grams? You must be joking
With neckstrap and eyecaps (I don‘t use the objective cover), it‘s more like 1300 grams (adding the objective cover, you end up at 1340 grams), so a lot more than with your usual 10x32 bino.

That's a heavy weight ... and I'm not so sure the photo capabilities (pretty good given the small size of the photo unit, but average at best in absolute terms) are going to make it easier to bear. As for the ID capabilities, I think most decent birders could probably identify the more common birds at the distances that have been discussed (I can identify a grey heron in flight at a heck of a lot further than 80m).

I appreciate weight and bulk isn't a consideration for many users, including some birders. But the speed and enjoyment of using a really lightweight, well packaged binocular, that lets me be fast on the birds (chill6x6 had a great expression for it - I can't remember how it went), is a pleasure in itself. I have used the 10x56 SLC (stated weight c.1200gm) a number of times, and although the optics are unquestionably fabulous, I very seldom find myself wishing I had it. I can't help but think the AX's the photo and ID capabilities are going to have to be very good (better than they are at present) to outweigh (apologies for the pun) the somewhat unappealing combo of 10x56 weight and 10x32 fiddlyness ...