Neutral Eye


Recently I have heard that the human eye does not look straight forward when neutral, but is slightly at an angle (meaning looking slightly cross eyed or possibly at an obtuse angle). Does anyone know a rule of thumb on this- do the eyes look forward parallel to each other in angle, or slightly off?

thank you


Hello comic-craig,

Very interesting question I asked myself many time .

I think that eyes are parallel when looking at an infinite point ( if infinite means neutral you get your answer ) .

So, in other cases (looking at a specific point in space), they are no more parallel and more specificaly cross eyed .

Of course, I am not sure about my answer, the best way to be sure would ask to a human anatomy scientist :scream:


Out of curiosity I did a quick search that turned up no data on any average deviation angle pertaining to line-of-sight, as it relates to the Primary Position (PP) of the eyes. PP is the “neutral” position of the eyes, when not fixated on an object. Both eyes are peering straight forward with parallel sight lines. One article mentioned that PP was ideal but gave no further info. My guess is the average deviation from the ideal PP is small enough not to be of any concern to a 3d modeler/portrait artist–could be wrong. :shrug:
I would also say it may be good to pull the eyes in slightly to give the impression that the portrait subject is focused on the viewer (if that’s the intent) and not focused out to infinity—having that thousand mile stare.

Newborns can have wacky eyes. I read here that only 23% of babies are born with “straight eyes”, coming into alignment within a few months.

Do a search on:

Primary position+eyes
Listing’s Law
Listing’s plane


Thank you both- that is very helpful. By the way- I like both your portfolios- and I’m a fan of Quadart’s kudo charge (I teach a class and use it as an example of excellence in 3D).


Quadart, it seems your search confirm what I thought about that question ^^

I was afraid to say wrong idea lol

Thank you for this quick search :wink:


If you are actually looking at something then your eyes must be converging to some degree or other, no matter how far that something is from you.

What do they do in the dark, or in a featureless environment? I don’t know.


It’s a very interesting question - i was looking for an answer myself several times, but couldn’t find anything convincing. As mentioned already, the angle is variable, depending on the distance of the focused object.

However, I tend to believe, that the difference between the viewing angles of the eyes is individual even if focusing at the same distance, depending very much on health, on state of mind and possibly also on the physiognomy.

Some people have a slight esotropia (one of the eyes points inwards), especially children - perhaps the reason why we believe it looks sweet on women. On the other hand much more adults have a very slight exotropia (one of the eyes points outwards, especially when they are tired), which we actually don’t notice while talking to them, but if you look carefully at portraits you’ll notice how many people don’t focus properly (called ‘strabismus’).

So the ideal angle in a portrait (the person focusing the viewer) depends certainly on the camera distance, resp. viewing angle. Theoretically the angle is never totally parallel - the most when focusing something far away.

PS: While doing some peaceful modeling i kept thinking about the subject, especially about “neutral angle”. I have some doubts about having such a default state in reality (eyes move as well while sleeping - probably less coordinated), which could be taken in 3d as universally usable value.


I must say that I’ve also thought about this.

When I model and position eyes in 3d I always have to point them outwards, so that the character doesn’t look cross eyed. I normally put each eye of by about 2 degrees, so in total they differ 4 degrees from eachother. To clear this up, the eyes are then slightly pointed this way:

/ \

I don’t know if this is correct, it has just always been the only way to make any characters eyes look realistic.


An anatomy teacher called Michael Mentler at ConceptArt have made a ton of awesome structured studies about measurements and angles of every part of the body. He touched on this subject very well some months ago. That when we look forward we do in fact look slightly downwards. The eyeball is also slightly elevated above the eyelids, not in center of both. Here are his illustrations!


Yes, I found the same thing about twelve years ago when I was making my first really realistic head. It’s impossible to make a realistic face with eyes perfectly straight and parallell.
I have since come to believe that this is indeed how we are constructed


Thank you people for all of your input- I welcome more. InsaneTK- thanks for the illustration- I appreciate it going the extra mile- any other links to Michael Mentlerare (or appropriate links) are appreciated.

Stahlberg- I’m not sure if I’m reading you right- are you saying that you agree with what was written above- or that the way eyes are constructed is parallel?


NOT parallell


I recently built on someone else’s head model, as they had a good topology, but I wanted to see what it would take to convert it from a western to an asian face (lose the epicanthic fold, among other details). They modeled like you describe. I think it’s easier to do it the way you say, but it’s not accurate physically. Their eyeballs were too large, and I think that’s part of the problem. For an earlier anatomy study, I found it really helped to try to marry two unrelated mesh models: a skull and a head. I had to tweak both to fit each other, as skulls really define the face, but it also gave great insight to the proper positioning of ear canal, nostril and eyeball.

(This was from a couple years ago, and not meant to be high-poly artistic. Be gentle.)

This debate reminds me of the old Renaissance Masters’ paintings, with lions and other animals that have front-facing eyes.


More factoid diggings about that neutral eye thing.

Awake, alert subjects do not seem to have a neutral eye position as the eyes are constantly fusing binocular images, fixating on objects at different distances while undergoing constant saccadic movements. *BTW, you are blind during those saccadic movements, otherwise vision would be a continuous blur.
There is an actual neutral eye position, occurring in the Physiological Rest Position and the Anatomical Rest Position, and it’s not very flattering. This position occurs during deep sleep, general anesthesia or death (prior to rigor mortis). Anatomical (at death) and physiological rest positions both involve some degree or total absence of eye muscle innervation. The eyes, under both conditions, are diverged (opposed to crossed) and rotated upward (more pronounced in death). You can also see this in blind people as well.

Another interesting and overlooked nuance is the fact that the pupil is not positioned in the center of the iris, as it seems to always be
depicted in 3d realistic human character work. You can look at an iris/pupil and immediately tell which eye it is and whether the head is right-side up or not, without seeing any other part of the eye. The pupil is always (statistically) located off-center, slightly toward the nose and slightly upward. If you see someone with a centered pupil, you are seeing an anomaly, or you are looking at an android! :surprised
It’s a subtle nuance, but if beating the Uncanny Valley is to be achieved, all anatomical subtleties should be taken into serious consideration. :shrug:

And relating directly to the offset pupil is the fact that when gazing (fixating) on a distant object, lets say the moon, the eyes are diverged, not parallel. This is due to the fact that the visual axis (line from fixation object to fovea) is not in line with the optical axis, which is the line passing through the centers of the two lenses (cornea and crystalline lens). The visual axis is rotated outward (temporally) from the optical axis by about 5° and down about 1°, ending at the location of the fovea, that tiny pit located on the retina where all of our sharp vision takes place (within about a 2° cone of view). The visual axes of both eyes are parallel when looking at, say a star, moon or mountain peak. This diverges the optic axes and therefore the eyeballs. If the eye were a perfect optical system, the fovea would lay on the optical axis, in line with the lenses. I read somewhere that if the fovea were on the optic axis, visual acuity would be 2.5 times greater (less aberration to deal with). The pupil compensates for the odd fovea position (only partially) by being offset toward the nose and upward. The pupil has it’s own axis called the Pupillary Axis.
–Thought someone might be interested, beside me.

**I am not an ophthalmologist, so do your own fact checking. :slight_smile:


Hey that’s cool to know I may have been right all these years about the divergence.
I used fashion magazines when learning to sculpt and noticed the eyes did not seem parallel but diverged(you can notice this well when looking at someone staring straight forward from below).
Also I tended to make the pupil inward slightly because it never looked center to me.


This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.