Was it really 84? I didn’t pay a whole lot of attention to the Aero, due to lack of interest, but could’ve sworn it was something like 95~ or so, before it later went to around 110. I do remember the 115 horizontal claim being ridiculous, given how it didn’t even reach 100 degrees, but I don’t recall it being that much below advertised.
2 Likes
Yeah so. I got some bad news about this post.
According to the China tester that has been providing us most of the information that has been discussed here, including the Risa2000 numbers, they reached out to this guy and asked them how exactly they got their numbers. He said, and I quote
He didn’t test it, just guessed by feel，he“feels” that is similar to index
So these numbers are literally just made the hell up based on how this guy felt.
Alright, guys, you can bring the pitchforks and torches back out.
7 Likes
Yeah. There is always a balance of tradeoffs to make, and when Oculus’ designers chose their’s for the Rift CV1, they picked one between FOV and resolution, that happened to give the CV1 a pretty substantial resolution advantage over the contemporary HTC Vive, BUT this came at a, in my opinion, unacceptable commensurate cost in field of view; Especially in the binocular overlap sector, where degrees of angle were taken from, in order to mitigate the combined total FOV loss. -Where the Vive was like looking through WC paper rolls, the CV1 was like looking through square section tubes which fit entirely inside those rolls, and the edges of that square view were inside the FOV given by the lens - focussed at its virtual focal distance, rather than the near-eye lens edges being what cut off the visible FOV, giving the wearer this unpleasant, cramped, boxed-in feeling (many didn’t notice this as much, because due to face shape, and the lack of eye-relief adjustment, they didn’t get close enough to the lenses, which reduced their practical FOV even further, but at least prevented them from seeing the sharp edges).
Interestingly, the square-ish per-eye views did result in distinct corners, between which an unquestionable (…and much bandied-about…) “diagonal”, which due to the square shape was noteably larger than the horizontal FOV, could be measured… :7
(Another subjective effect, which has come back up here, recently, in comparisons between Pimax wide FOV headsets and other, smaller FOV ones, was that people’s taking relative, rather than absolute impressions: From everything I could tell, degree for degree, the falloff in focus away from the centre, was about the same between the Rift CV1, and the original Vive, but some users were swearing blind that the former was infinitely better in that area… and well, kinda…: almost 100 % of its view was sharp, but that’s because it didn’t show, at all, the further 20% of FOV the Vive did, which would have been just as out of focus in the CV1 as they were in the Vive; 35 degrees out, both devices were about equally blurry, but people didn’t take their impressions by degree, but by percentage, even though these were percentages of two very different sizes.)
No, thank you; I want a realistic view – not zoomed in, not zoomed out, not bulging in the middle (…which something you wrote a bit farther up sounded alarmingly like – things like that means incorrect distortion correction and/or pupil swim (some pupil swim it to be expected with conventional lenses, unless one have eyetracking adjusting the correction profile on the fly)).
If I hold my arm out so that my hand appears 50° out to my side in the real world, it should appear 50° out in the virtual world, too, and appear the same size, or we are fast slipping away from anything you could call VR. Fisheye Quake, anybody? :7 Fisheye Quake
Only time the world should look larger or smaller, is in situations where I play something like an ant, or a giant, or simulating the sight of an animal with side-facing eyes.
Let’s leave things like exaggerated stereo separation to old flatscreen 3D, where they belong.
I guess I’m just jaded to the point of defeatism, but this time around I’m finding myself, rather than calling things out, taking the disparity between advertised and reported FOV with a lazy: “Yeah, sure - of course; What else would you expect, by now?”. 
103° is around where most headsets lie, and what one might have expected to begin with… :7
I speculated once, that maybe Pimax “had to” double down on the old 200° nonsense for the 8k series, once its “nonsense-hood” became irrefutable, or maybe they could be opening themselves up to retroactive demands, but surely they’ll be careful to not put themselves in the same situation again, when people are wary from the outset…
Anyway – one way I could envision, to get inflated numbers with a tiny kernel of truth to them, is if one screwed the lenses all the way out to the sides, way past the user’s IPD, and software adjusted IPD to account for the mismatch. That would leave you with almost no stereo overlap, and waaaay out of the sweet spot of the lenses, but the edges of the FOV would be out there, at the impressive number.
Dishonest as all hell, of course – the FOV should be measured from the optically optimal point along the axes of the lenses, and nowhere else.
2 Likes
Conjunction junction, what’s your function? 
According to Road to VR, yes. And various other sources measured its FOV being really small, too.
The Crystal has many similarities to the Aero, and Pimax does seem to be positioning the Crystal to compete against it. So it may be useful to look more at what the Aero did since Pimax may be following a similar pattern.
The Aero also didn’t have all of its features at release (and still doesn’t) and had an FOV much smaller than advertised (more so than the Crystal). I actually wrote a post here back then about how I thought it was strange because, while all manufacturers exaggerate their FOV numbers, Varjo was far out of line with the others and nobody seemed to be fussed about that.
It’s interesting to see Pimax doing what appears to be the same thing Varjo did, and people are very fussed when Pimax does it. I think this demonstrates the power of the vastly different marketing image between these two companies.
For myself, coming from an 8KX, I’m not interested in the Crystal anyway. I can’t go back to narrow FOV. A few degrees one way or another isn’t going to make the difference. Even if it had its full advertised FOV, it would still be around the same as all the other narrow FOV headsets. I wonder if there might be a psychological thing here where people are hoping they won’t be sacrificing wide FOV so much with the Crystal somehow, and so smaller than expected numbers are compounding what was already a false hope.
2 Likes
I dare hypothesise that if few people made a big fuss about Varjo’s misleading marketing, that is because their pricing places them way, way, way beyond registering on almost everybody’s horizon of attention… besides, I guess all the dead pixels stole the limelight.
2 Likes
Varjo advertises with 115 Hor/134 diag while it’s in fact rendering 111 hor/121 diag. The Crystal seems to be rendering 103 horizontal while Pimax advertised it as 125 if I’m correct? So that’s quite a bit worse.
Not defending Varjo at all btw, I HATE it when companies think they need to mislead their customers in order to sell their products. I much prefer a company like Bytedance who actually advertises correct numbers. But I think the fuss here is not just the incorrect Pimax FoV claim, but that they’ve been inflating their numbers by this much. I don’t think there’s any other headset out there with such inflated & misleading claims.
That being said, you do have somewhat of a point, because while the Aero renders 111 horizontal, the subjectively measured value is oftentimes even way worse, since the standard gasket hides part of the FoV. I needed to do a mod to get to closer to the rendered FoV. While with the aero at least for now it SEEMS there’s no need to do such a mod.
Either way, just like you, I can’t get back to such narrow FoV myself neither. I liked the Aero for a while, but after switching to the XTAL it was clear to me I can’t ever go back, even though the visual fidelity of the Aero is quite a bit better.
4 Likes
I don’t know why nobody mentioned it here. Risa2000 is using the rendered 3D space on the panel to calculate the FOV, but the rendered 3D space can be easily changed by the software. If you prefer, you can show 360 degree of a landscape into the panel, but that doesn’t mean the headset has 360 degree FOV.
Yes, it’s a hard numbers that Risa2000 can give you, but why everyone here think that’s a number represents the actually FOV you see ?
Someone above mentioned binocular FOV, which is a good example. In a binocular, you are watching a picture which have a much larger FOV than the actual one. You feel that the vision is 45 degrees, but that’s actually 8x or 10x magnified. The real FOV you are seeing is more like 5 degrees. Did anyone complain about a massively exaggerated FOV to the companies that make binoculars? Guess not. Because that’s expected.
So ask yourself this question: do you expect a picture remains the same FOV after passing through convex lens? Shouldn’t you expect a zoom-in picture with ashperical lens?
If you never expected asperical lens will cause the panels to appear larger than they actually are. Now you do. And now you know that’s most other VR headsets do as well, they display a distorted smaller picture with larger FOV, after passing the lens, the picture will be zoomed in and displayed correctly. So don’t be surprised that somehow the real FOV was different from the rendered FOV.
On the other hand, Pimax can easily change the FOV rendered on the panels with a simple update. If you want 125, they can give you 125 or even 145 degrees, the Risa2000 number will be good, but of course it will cause more serious distortion problems. No one will really want that.
Risa’s tool gives you the Maximum possible FoV one could see. Not to be confused with perceived FoV which will vary dependent on a variety of factors related to the user.
You cannot perceive more than is rendered.
With your rendered example sure you can render a really distorted fish eye pic with lots of FoV but you need the optics to make it look right instead of a funhouse mirror image.
Risa’s tool simply obtains what the software requests to be rendered which represents a maximum potential of what can be seen. You cannot see more than what is there. But you can see less.
4 Likes
Using the binocular example, the maximum actual FOV you can see is 5 degrees, but you are actually seeing 45 degrees in the binocular without the funhouse image effect.
Everything is relative. Yeah, I actually see that slight funhouse effect myself, but interestingly everyone else in the roadshow and CES didn’t notice that. Don’t you find it strange?
Binoculars you have 2 values as you have 2 lenses. the end result is not 5°. It is the result of the 2 lenses combined x2.
The FoV calculator for example you linked earlier only takes into consideration Aspect ratio iirc. Which is very incomplete without other factors being worked in.
2 Likes
I don’t know enough to explain it properly to you, but I do know enough to know that you are completely misunderstanding.
In that case your FOV is 45 deg. And if you zoom out, or zoom in further, it will still be 45.
Hopefully someone can come in to explain further, perhaps @risa2000 ? Anyway, yes software can effect the fov to some level like you could use software IPD offset to get some more wide FOV at the expensve of stereo overlap (it also seems to reduce vertical FOV but I don’t know why or how). That said you’re starting to butcher the image at that point and ultimately you’re forcing something that isn’t really intended. Even with that, I think it got up to like 108deg horizontal max but wasn’t a good image.
I guess an easy way to consider your position on this is why are you apparently the only person to come to this conclusion? I’m not trying to be rude or condescending, but you do keep trying to force this point even though no one agrees.
Because it does. It literally does. Why don’t you believe literally everyone saying that it does? It’s not some grand conspiracy.
1 Like
I don’t know why you still argue with this. The end result in a binocular is much more than 5x2 = 10 degrees, right?
This has nothing to do with the FOV calculation I posted earlier. It’s a completely different subject. I admit that I did make a mistake on that calculation, so did those guys in that university. ( I shouldn’t have trust them). Yet, this is about the zoom-in effect of lens vs the Risa2000 numbers.
So no one agree that in a binocular the FOV is greatly exaggerated? What you see is 45 degree, but in reality that’s only 5 degrees. Anyone who’s disagree on this, please voice out and let us know.
Speaking out alone doesn’t mean it’s not true.
Oh, man. You gotta understand what that number actually represents, then you make your own judgement about that. Don’t just follow everyone’s voice.
Nope because your only factoring the first set of lenses and then not factoring in Stereo overlap. If you have for example simple tubes that let you only see straight ahead. Let’s use a simple number of 45°/eye Combined you have 45° as both eyes see the exact same just off set left and right.
The issue is your not understanding the info being presented. Risa’s Tool unlike the FoV calculator you linked has a ton more data like Stereo over laps, Left & right fov information etc…
Tbh I am not likely able to explain it on a level you might understand. And that is nothing to do with your ability to understand but more with my ability to explain it in a way that would make things clear.
I just understand the info to know the FoV calculator is not meant for use with hmds. Now you could likely compare Risa’s Tool with Doc-ok’s papers on FoV measurements and you might better understand.
If your nit familiar with Doc-ok highly recommend you read up.
1 Like
Since a HmD is finite from a physical point of view due to its design (only the view is infinite), you only have to move the virtual viewport about 2cm forward to subjectively get about 5 degrees more. I would like to know how many people actually realize that an object is floating that 2 cm closer to their face.
At the short distance there is still no noticeable influence on the room geometry. Everyone who wears glasses has to live with discrepancies that are many times higher.
Yes, only xxx degrees are rendered, and that should also be included in datasheets in my opinion, but an eye is not a measuring device.
1 Like
That is an uneducated presumption. Just because one understands something to be true doesn’t mean they can explain it to someone else giving them the same enlightenment to understand it themselves. 
OK then. I think other people here will be the judge. I think what I said is quite easy to understand, right? Hope someone else here to explain why Risa2000 number is so firm and be the actual final FOV , like you said.
Educated or not, we do need to measure FOV with actual optical instrument. The result won’t be affected by which college degree the operator holds.
Read uo on Doc-ok as he us a foremost recognized expert in the field.
Risa’s numbers are firm as they are max possible by the info used to render. There is no hidden magic. This is what the headset requests you cannot see more than what is there but you can see less.
This should help you get started. Doc-ok was the expert whom confirmed during the pimax original ks that it was indeed rendering wide fov.
He also has written numerous papers on measuring distortions, mura etc…
4 Likes