My question is about what exactly is meant by a binocular being collimated/aligned. I'm familiar with optics in general, but not with the nitty gritty of binocular design.
There are various resources online (here is one example) that describe different methods to align binoculars (sometimes also called collimation but I will avoid this term here to prevent confusion), usually without explaining in detail what alignment means. One might think (perhaps naively) that alignment means making it so that the optical axes of the left and right "telescopes" extending beyond the objectives are parallel, i.e. both "telescopes" are pointing in the same direction. But I'm not sure if this is the case for reasons I'm about to describe.
I currently have two binoculars, one is a Celestron 8x56 and the other is a new Oberwerk LW 12x60. With both binoculars, the fields of view (FoV) I get through the left and right "telescopes" of the binocular are considerably different: namely there is a horizontal shift. This is not due to parallax because even when an object at infinity (e.g. a star) is at the center of the left eye FoV, it is to the left of the center of the right eye FoV. This tells me that the optical axes of the left and right "telescopes" extending beyond the objectives are not exactly parallel, i.e. they are pointing in slightly different directions. Am I right? If so, why aren't the optical axes parallel?
Various resources on aligning binoculars say such a shift is "normal", but without explaining exactly how much shift is "normal" and why (I understand the brain is capable of merging the two images even in the presence of an undesirable shift). But why should there be such a shift at all in a perfectly aligned binocular? Is this always an unintentional consequence of the left and right components becoming misaligned over time, or is the non-zero angle between the optical axes a design parameter? The former would mean both of my binoculars (including the new Oberwerk) are significantly misaligned, while the latter possibility I also find unlikely because when looking at an object at infinity without a binocular, the "optical axes" of the eyes (each intersecting the retina at the fovea) should be parallel.
Any insights are appreciated.
Edit: Regarding how much of a "shift" I'm talking about: I have not tried to quantify it yet. I will try to estimate the shift when I have time later. But it is small enough not to cause "double vision" when looking at a star, and large enough that it is very easily seen by closing one eye and then the other when looking at a star. I thought this horizontal divergence of the optical axes might be intentional since I see it with both of my binoculars (with no appreciable vertical divergence), and at least one manufacturer (Oberwerk) makes a big deal of how well they collimate their binoculars. But I'm not sure.