Trying to dial a camera view space taller
nomad-ads_8ecd56922e
Posts: 1,873
Okay, I'm running into a rather exasperating problem.
I have a camera I made and positioned in a scene, and I want to make the framed-area of the view in it taller without physically moving the camara farther away. It's wierd that you can set a camera to, say, 1800 x 1400 in size/shape, but cannot make it 1400 x 1800 without it rejiggering things so that the top and bottom area of what was visible in the camera at 1800 x 1400 doesn't enlarge accordingly. It's like there's some mechanism under the hood of Daz Studio that says "I'm going to visually CAMERA ZOOM IN to make sure the top and bottom of what the camera sees DOES NOT CHANGE, and in the process, I'll lop the sides of the image off accordingly!"
There's gotta be a way to get AROUND this issue, short of physically flipping the danged camera on its side, and then rotating the image in post production in a picture editor program. It's like, if the LENSE is capable of seeing 1800 pixels ACROSS of the image, you'd figure the lense would be capable of seeing 1800 pixels top to bottom instead, if I wanted to, since either way, the only limit would be where the film stock is oriented inside the camera, not the physics of how the lense behaves. 0o
The lense, after all, is ROUND, not... like....rectangular.... :D
I'm guessing there's some hidden feature or toggle in the camera's settings that will turn this infuriating "I'll just zoom the view IN for you!" behavior OFF. o0
Ideally, what I want to do is have the SIDES of the view never change outward or inward, but to shift the top and bottom larger or smaller in relation to the sides of the image seen in scene, but it keeps lopping off the sides of that area instead, or it widens the frame outward instead... which is clearly what I DON'T want.
Daz 3D is part of 
Connect
DAZ Productions, Inc.
224 S 200 W, Suite #250
Salt Lake City, UT 84101
Licensing Agreement | Terms of Service | Privacy Policy | EULA
© 2024 Daz Productions Inc. All Rights Reserved.
Comments
I always set my desired camera dimensions beforehand, so I can't say I fully understand. But isn't the frame width parameter for situations like this?
try change your frame width to value*1800/1400. It is under camera setting.
Okay, I guess I need to explain visually what I'm trying to do here.
When I have a camera selected, such that I'm looking at what the placed-out camera "sees" in scene, I see the shape of the frame of the camera-view in Viewport, representanted by a white rectangle-lines thingy. What I WANT to do is make the upper and lower part of that rectangle-lines thingy expand outward vertically within the Viewport, without changing where the sides of that rectangular-lines thingy appears inside the Viewport. I.e., I want to see more of the windows over there, and more of that big pipe thingy down there, without lopping off any of that catwalk along the right or any of that guardrail area over there on the left.
Typically, what I TRY to do is go into the Dimensions and plug in a custom width and height there under Pixel Size on the right, until that white-rectangle thingy encloses whatever part of the viewport image I want to appear inside my render. You'd figure that, with Constrain Proportions changed to OFF over there, I could simply plug whatever number I wanted into the H: field over there, and the white-rectangle thingy would become taller, without shifting the position of the sides of that rectangle around in relation to what's visible inside the Viewport there, but instead its like the Viewport space "zooms farther in" on whatever was visible in the Viewport, such that part of that catwalk on the right and those guardrails on the left get shoved out of view, and that big pipe thingy down there and the upper part of the ceiling up there get pushed out of the top of the Viewport as well. All I want to do is expand the area of the white-rectangle thingy upward and downward within the existing view, to encorporate more of what's visible in the Viewport there.
Effectively, every time I try to enlarge the white-rectangle thingy vertically, to include more of the room into it (i.e. more of the ceiling and the floor), it lops more of the SIDES of what I WANT to see in-camera OFF, so that the view of the ceiling and the floor never changes within the white-rectangle thingy. o0
Lenses are round, but not the sensors on cameras. Are you actually setting the aspect ratio to 1800x1400? There's no real reason to do that, as setting it to, for this example, 18x14 would yield exactly the same results, so long as you set the pixel size to the dimensions you want.
Well, what I was trying to say is.... the "physics" of the (simulated) lenses on the (simulated) camera inside Daz should be perfectly capabable of projecting a 1400 by 1800 image onto an arbitrarily-shaped (simulated) film-stock inside the camera as it would an 1800 x 1400 image, assuming you had a customized camera housing that let you insert such a nonstandard-shaped film stock into it, or for that matter, one that let you rotate the filmstock 90 degrees in relation to the viewfinder.... so why is something in Daz so steadfastly trying to shift my white-rectangle-space around in Viewport in a way that refeuses to recognize I might (virtually speaking here) have an "oddball" shaped film stock I want to use?
Anyway, I was turning Constrain Proportions to Off with the assumption it would automatically adjust the Aspect Ratio W: and H: fields dynamically (that is, not try to force the ratio of the rectangle to fit the numbers plugged into the Aspect Ratio fields), and simply, and that it would automatically rejigger the Aspect Ratio to be equivalent to whatever width-and-height numbers I plugged into Pixel Size width and height.
And no, I wasn't plugging any numbers into the Aspect Ratio feild (i.e. where it says 3.0 and 2.0 in the first screengrab), I'm trying to make a render that is 1800 x 1400 in pixels (by plugging those values into the Pixel Size fields), because I want to make a render set (made from a bunch of diffrerent cameras placed out in a scene) all have the same size for what amounts to a web-comic.
I'm still not following you 100%, but it sounds like you're talking about the image distorting when you resize the camera? IIRC, you need to adjust the focal length to flatten your image.
Suppose you start off with a square image, its aspect ratio is 1:1 and it encloses a certain part of your screen. (Note that it doesn't matter if you have constrain proportions on or off, or if you change the aspect ratio numbers or the pixel numbers, what matters is the ratio you end up with between the height and width)
Now you change that ratio to 2:1 - a rectangle twice as wide as it is high. What should it do? Should it preserve the amount of the scene captured vertically, and add a chunk to each side (kinda like zooming out), or should it preserve the amount that's captured horizontally and chop chunks off the top and bottom (kinda like zooming in) to get the right shape? Similarly, if you make it 1:2, do you slice from the sides or add to the top and bottom?
There's no right or wrong answer to that question - both approaches achieve the stated goal.
What Daz Studio actually does is always preserve the amount of vertical space selected. If you make the frame wider, it does that by adding chunks to the sides. If you try to make it taller, it just chips away at the sides to make it narrower instead - the same ratio between height and width is achieved. That's probably what most people want most of the time - we go between normal screen and wide screen, not between wide screen and tall screen. I don't believe you can switch it to behave another way - other than, as you say, rotating your camera by 90 degrees.
It has nothing to do with image distortion, it has everything to do with what objects in viewport I want to include in the render. I have objects that are positioned above or below where the rendered camera-view wants to be, and I want to rejigger the camera-view so it now includes those objects.... without physically moving the camara farther away in the set. It frustrates me because, if there wer objects that same distance to the left or the right of that camera-view area, it's trivially easy to get them into that camera-view area, because the camera-view area will widen routinely for me, bringing those objects into the rendered area. The system refuses to do that for stuff above or below that view-area.