LateWhiteRabbit wrote:You completely misunderstood everything I said. Eye level refers to the PHOTOGRAPHER, NOT the model.
You're absolutely right I
did misunderstand what you said because this following line didn't compute properly:
LateWhiteRabbit wrote:The horizon line in any photograph will be identical to the eyeline of the photographer if they were looking through the lens.
I'm not a photographer, so "eye-line" has a completely different meaning to me. You were referring to what the photographer sees through his camera. I was referring to the horizon line "I" see in the
resulting image. Your frame of reference and mine were not the same, so misunderstanding, or rather, mis-
translation was the result.
LateWhiteRabbit wrote:The eye line, or eye level is the consideration for your VIEWPOINT CHARACTER.
This line, above, is the biggest misunderstanding between us.
In my stories, the Viewer; the player, is Not the viewpoint character. My viewpoint character is an actual character that appears in the story just like all the others. My players are merely Observers, NOT participants. This is why I need models that are NOT looking directly at the player, and I need backgrounds and characters with a chest level horizon line.
LateWhiteRabbit wrote:VN sprites often look wrong because they are not in perspective with the backgrounds.
On this we both agree.
-- Matching perspective to the doll is something I try to do, but one can only make do with the resources one has -- or makes.
LateWhiteRabbit wrote:For instance, this:
The photograph actually has distortion (i.e. curvilinear perspective, how our eyes really see). Notice how the edge of the table curves slightly, when we know in reality it is straight. Yet the doll is flat, head-on to the viewer.
Exactly.
-- Using Photoshop's Lens Correction to straighten most of those lines, plus judicious trimming at the top can make a doll fit better. However, this doesn't actually change the horizon line or where the doll should be placed.
This is why I look for photos taken by professionals. They have less distortion because they are Not taken at the photographer's full standing height, but lower -- at chest level, the same perspective that the dolls are drawn.
LateWhiteRabbit wrote:Knowing how tall or how far off the ground your main character - the viewpoint character - is, is key to knowing where the place the horizon line and where the other characters should fall on that horizon line...
On this I will have to disagree because book illustrations, movies, and anime do NOT take the viewer's height into consideration. They lay the horizon line across the center of the image, at Chest-level on the actor/character (unless they are doing a head and shoulders close-up.) Any story that did would look Wrong to the viewer because they are Not used to seeing illustrations done in this manner -- no matter how realistic it actually is.
LateWhiteRabbit wrote:
Despite being a representation of reality, this would look Unnatural as an illustration. I would never use this image in any VN I created.
LateWhiteRabbit wrote:You can't just draw backgrounds with different horizon lines (or use photos with different horizon lines) and the same straight ahead sprite dolls. To have proper perspective in a VN, every background must have the same perspective and horizon line...
On this, we definitely agree. However, one can only work with what is available to us.
LateWhiteRabbit wrote:The only other alternative is to draw new sprites for use with each background.
Or be a photographer, which I am not.
LateWhiteRabbit wrote:Anyway, it is good this stuff is being thought about, we just had a misunderstanding on what "eye line" means.
Honestly, I think we were addressing the same issue, just approaching it from different 'perspectives'. (Pun fully intended.)