Blitz 3D Strange perspective

Blitz3D Forums/Blitz3D Programming/Blitz 3D Strange perspective

mouss38(Posted 2003) [#1]
The question remains : why Blitz 3D perspective is so strangely deformed ? This fact is visible in all of its 3D applications

Look at this screenshot :



Jim Teeuwen(Posted 2003) [#2]
Try fiddling with CameraZoom() to adjust the perspective level..


DH(Posted 2003) [#3]
if you have the camera as a child of another object (like a parent pivot or something) and you scale the parenting object, it scales the camera and distorts it as well..


Rob Farley(Posted 2003) [#4]
As Defiance said, put the camera zoom at around 1.5 and you'll find it won't have the fisheye effect.


Hotcakes(Posted 2003) [#5]
I picked up from someone in these forums somewhere that a camera zoom of 1.6 is apparently closest to that of the human eye. Or something. =]


Anthony Flack(Posted 2003) [#6]
Perhaps, but remember human vision has a ratio of about 2:1... whereas your monitor has a ratio of 3:4. so does a 1.6 zoom indicate the correct ratio for the height, or for the width?


Rottbott(Posted 2003) [#7]
No idea but I always use 1.5 myself so obviously quite a few people have reached a similar conclusion.


Hotcakes(Posted 2003) [#8]
So, ideally, using an example of a resolution of 640x480, the display should be rendered to no more than 640x320. Save that last 160 pixels for status messages and such =]


jhocking(Posted 2003) [#9]
Human vision has a ratio of 2:1? Where'ld you come up with that? It reminds me of people saying that human vision has such-and-such framerate. You can't attach definite numbers to human vision; our eyes aren't monitors. Among other things peripheral vision defies any attempt to quantify a human aspect ratio (because different people have different peripheral vision acuity, because peripheral vision wraps around instead of just seeing a flat projection, etc.)

This gets to rendered images not looking exactly like human vision: it's because it is impossible to look exactly like human vision. Monitors are fundamentally different from our eyes so the best you can do is some sort of approximation by fiddling with CameraZoom.


LAB[au](Posted 2003) [#10]
Field of view is more or less 64° for human stereoscopic vision (horizontal excluding peripheral "blurry" vision), thus according to this formula

1/Tan (64.0/2) = 1.60033

is the value I use.