A level IV parallel universe is a universe that may have different laws of nature than our own, but still work out to be mathematically true. I'm thinking about a level IV parallel universe where, instead of adding color with light (like the display on your computer, its default color with no light added is black), there could be color added to subtract light (like a painting, if you apply paint to the paper, it takes away color from the default white color). Would it be possible to have an alternate universe with the subtractive color system? (Ex: the stars would take away light from a "white color" universe)
The frustrating answer with Level IV worlds is that it's always "why not?" What you have described is so removed from physics that the answer is "sure." There's even a famous thought experiment where the sky is actually green, but everyone uses the word "blue" to describe it.
What you describe is color perception. Colors are an incredibly complicated system. They go from the spectrum of EM radiation into 3 colors, working through component processes, and eventually come out the other end as a qualia. There's a lot of slop room there! With the freedom of a Level IV alternate universe, that's easy! Arguably it can be achieved in THIS world, with a sufficient application of LSD.
Now as you add in more effects that are like those in our universe, you can run into consistency issues. If you assume light is modelable with our equations for electromagnetism and if you assume color perception in this Level IV universe can only possibly match that of our own and if you assume energy is a conserved concept and if you assume quantum mechanical rules that suggest that the addition of light must be a linear function, then you might start getting into inconsistencies dealing with the energy of photons.
But sure. It's Level IV. Virtually anything is possible.
We already add color, or things that absorb and reflect given light frequencies. The added color subtracts light, by absorbing certain frequencies and transmitting or reflecting others. You point out the case of paint: we have that. Leaves do that: they absorb a lot of light but mostly reflect green, so they look green to us. One could have colored light of a single frequency, but most cases of color entail absorption of certain spectra from multi spectrum light, with the non absorbed spectra reflected and so producing the color we see. Color is the subtraction of light.
I can see a few different ways to interpret this.
One is that the universe naturally produces a certain distribution of background wavelengths of some phenomenon which can be usefully analogized to our inverse's light, and stars and other "light sources" actually emit something which suppresses production of or removes particular wavelengths from the background radiation environment. Something like this can sort of happen in our universe, where light waves can destructively interfere with each other. It's rather difficult to get that to happen coherently on a large scale, though, and you always end up with additive interference somewhere--there is no generic "dark beam"--and I would not like trying to figure out how to design a universe where that would work out. I can see, however, two alternate possibilities for how such a system might be made to function, without relying on quantum interference.
The first is that the universe constantly generates some background radiation spontaneously, out of free space, which is directly perceived as light. Apparent "light sources", like stars, would then emit some second form of radiation whose presence suppresses the background production of light, but which is not directly perceivable. "Laser beams" in such a universe would in fact be visible off-axis in empty space, since they would produce a corridor in which the omni-directional production of perceptible radiation is suppressed, and "light sources" would produce visible haloes, rather than emitting colors only directly from their surfaces. I imagine it would take some time for the inhabitants of such a universe to figure out that their "light" is in fact two separate, but interacting, phenomena.
The second option merely requires some background radiation field, whether it is continuously produced from free space or is simply primordial, like our microwave background. They key is that the energy of this background field is of opposite sign to that emitted by other light sources. Positive and negative "light" would only interact via the normal interference processes in free space, but the color-subtractive effect would come into play in their interactions with matter (i.e., retinas, cameras, and everything else). Where a single photon/luxon/whatever of "light", whether positive or negative, may be capable of triggering an electronic-or-analogous transition, thus causing detection, the simultaneous absorption of luxons of opposite sign would result in no effect--so a light source that emits luxons of a particular color/wavelength would prevent anyone who can "see" it from seeing luxons of the same wavelength from the opposite-signed background radiation. In this sort of universe, there would still only be a single type of particle/field which makes up "light", and lasers, as in our universe, would not be visible off-axis.
Another interpretation is that rather than producing "light" at all, objects like stars merely absorb the background light, and images are typically formed by the dark patches on a bright background rather than light patches on a dark background. That's easy to do: either provide a steady-state radiation field sourced from the vacuum, or a primordial background radiation field similar to our own cosmic background many billions of years ago (when it was warmer, and in the visible rather than microwave range), and tweak things so that "stars" tend to be mostly-translucent (rather than dense and opaque). Depending on what other features you want for this universe, there's a wide array of options for explaining why large objects like stars interact with "light" so relatively little, and what happens to the energy they absorb. The functional equivalent of stars in such a universe might be more analogous to our black holes (point-like cold sinks, with a hot background, rather than a cold background with point-like heat sources), or you might come up with some more exotic method of keeping the temperature down. (For comparison, "stars" in Greg Egan's Orthogonal universe incidentally produce light, but their light is not used to supply energy to planetary biospheres--rather, "plants" can spontaneously produce all the energy they need, with heat as a byproduct of inefficiencies, and starlight merely serves as a source of low entropy.)