In my setting elves can see very well in the dark. I've suggested tapetum lucidim but that would make their vision less clear so would larger pupils make it easier to see in low light areas?
1$\begingroup$ When dark adapted the human eye has an aperture of about f / 2.1. This is a very good number; you could increase the aperture to f / 1.8 (one half-stop) or maybe f / 1.4 (one full stop) but not more (unless you hire Leica or Zeiss to design the lens, and even Leica or Zeiss could not go more than f / 1); if you really want them to have much better night vision than humans then you must give them bigger eyes. In low-light photography there is no replacement for sensor size. $\endgroup$– AlexPAug 26, 2017 at 14:42
$\begingroup$ @AlexP, en.wikipedia.org/wiki/Carl_Zeiss_Planar_50mm_f/0.7 suggests that Zeiss could and did. Though it would come with certain significant drawbacks. $\endgroup$– GhanimaAug 26, 2017 at 17:57
$\begingroup$ @Ghanima: There are several photographic lenses with apertures wider than f/1, such as the Zenitar 50 mm f/0.95 which is making the news this week, but they have a structure completely different from the vertebrate eye (which is essentially one spherical lens with four elements in three groups). $\endgroup$– AlexPAug 26, 2017 at 20:59
Practical limit is the precision of the whole optical elements composing the eye.
It is well known a lens system "aberrations" (imperfections due to several reasons) increase as you move out from the center and thus with larger diaphragm (i.e. iris) aperture.
This is the reason why we can use cheap and relatively inexpensive lens systems in today's cameras (and phones): the light required is much less and the "film" (CCD retina) dimensions ar much smaller (about 3x5mm or less where a Hasselblad had 60x60mm film).
Rising "sensitivity" of the film (retina) is the way we are actually bettering out "night vision".
Having a large pupil would surely help at night, but at risk of getting a blurred image due to imperfections of various lenses composing eye.
Yes and no. It COULD increase night vision, but under certain conditions, and with other trade offs in acuity.
I am assuming that the sensitivity of the light-detecting elements of the eye remain the same. In biology, there are several ways of 'detecting light' in organisms. Therefore, it is possible to evolutionary select a better, more light-sensitive sensor for the eye. Even in the human eye, we have rods and cones - a trade off between color vision and night vision. In the fly eye, the eye is broken up into individual segments, each of which acts as one 'pixel'. Each 'pixel' has a given light sensitivity, so increasing the number of 'pixels' does not increase light sensitivity, it only increases the field of vision.
To increase the 'night vision' capability of the eye, you either increase the light sensitivity of the receptor, or increase the amount of light getting to each receptor. Your question seems to indicate that you want the sensitivity of the receptor to remain the same (there ARE ways to increase the light sensitivity of the receptor - 'eat your carrots').
So, that leaves increasing the amount of light that gets to each receptor. One way is to reduce any 'transmission loss' in the eye structure - the clarity of the lens and eye fluid.
The second is to increase the exposure time - allow the eye receptor to 'accumulate light' before it 'fires'. This, of course, slows down vision. It is the basis for time-lapse photography.
The third is to somehow amplify the light. Night vision goggles do this, but it means having a secondary receptor in the middle of the light path, that 'adds' light to strengthen the incoming signal. However, it means you have a very sensitive receptor in the first place, to pick up the available light in order to amplify it.
The fourth is to allow more light to hit each receptor. This is the tapetum lucidim - bouncing the same light around so more of it hits the eye sensor. However, it is also possible to do this by allowing a bigger 'cone' of light to hit each sensor. That is, each sensor picks up light from a wider area of the scene. This is accomplished by either focusing the same amount of light on a smaller area of sensors, or increasing the amount of light that enters the eye. Both methods involve changing the focal length - the distance between the lens and the light sensor. In the first case, you have lower resolution - fewer 'pixels' that each receive a greater proportion of the available light, or visual field. Making the pixels bigger (the light-sensitive elements) also accomplishes the same thing. The bigger the individual light sensing cell is, the more available light hits it. However, in any case, the resolution of the eye is decreased.
Increasing the amount of available light is the method that you have asked about - increasing the pupil size. The trick is, you have to focus all of this light onto the available retina - letting more light in is useless if it doesn't fall on the light sensitive areas. The increased light has to be re-focused on the available retinal surface - the focal length has to change. The retina has to be optically brought closer to the lens. Thus, increasing the size of the pupil only works if you change the real or virtual distance between the lens and the eye. Doing so, as every photographer knows, has the drawback of reducing the depth of field - only things that are a given distance away are in focus, objects that are closer or further are increasingly out of focus.
All of this assumes the existence of a suitable lens. In a pin-hole camera (that is, a camera with no lens, but just a pin hole to let the light through) the smaller the pin hole, the greater the focus. Making the hole larger (bigger pupil) only creates a more blurred, diffuse image, but does allow more light through.
Think of an array of one hundred different colored lights, ten across and ten high. You are positioned well behind a screen, and the screen is a great distance from the light array. You are looking at these lights through a window (hole) in the screen. If the screen has a hole (window) sufficiently small enough that you can only see one light at a time (the visual cone of only one light is observable by your eye), you will, of course, see only one color light at a time. You have to move your head around to see each other color. If you make the hole bigger, you will see more color lights at a time. The light cone is broad enough to allow you to see perhaps colors from four lights at a time (less resolution, the colors of all the lights blends together), but the light is brighter. Get further away, and you again will only see one light at a time (one color). Get closer, you will see more blended color lights at a time (but it will be brighter). Make the hole bigger, you have to move further away in order to see only one color light at a time.
Thus, there is a trade off between the size of the pupils, for increased night vision, and the acuity (sharpness) of vision, for the same receptor and the same distance from the pupil. Making the entire eye bigger (increasing the distance from the viewer to the screen, to increase the resolution), means that the amount of light reaching the sensor decreases, reducing night vision.
However, caveat emptor, as every pirate knows (this is why they wear an eye patch) increasing night vision has devastating consequences in full sunlight. Your elves would need eye shields, or filters, for daylight.