The philosophy of science, which is behind materialism, limits the investigation of reality, to only that which we all can agree, using external sensory input.
For example, say we were all camping, at night, in the woods. It is dark and scary for some. A rustling in the brush is thought, by some, to be a bear. While to others, the shadow near the tree looks like a wolf. The philosophy of science, as applied to collective observations in this camping trip, will limit itself to only those things that everyone at the camp site can agree. It will factor out the bear and the wolf, simply because we all cannot agree on these things. Not everyone sees these things for consensus verification.
If all cannot agree, it is called subjective. Subjective does not mean that the consciousness of a few, did not perceive this. It only means that everyone cannot agree on these things, for it to be considered objective consensus. This is the constraint of the philosophy of science; age of enlightenment. Reality was defined as that which we all can agree. This factors out religion, personal opinion, and even philosophy. For example, we can all agree on the sun, but we cannot all agree on the chariot of Helios, even if a few see the chariot.
This philosophical assumption was very useful to narrow down what we will look at. However, it is not complete, since it does not do justice to all phenomena, especially those connected to consciousness. This is why consciousness is the final frontier of science. The philosophy specifically factored out certain brain output.
As an example, say I related a dream in a lot of detail. On the one hand, we have all had dreams, and science knows dreams occur, so this is not out of the realm of reality. However, the details of my dream is not something that can be verified, or seen by anyone else. It can be real and objective, but only to an audience of one. It can never reach the level of an objective sensory consensus, even if real and tangible in some brain output way. The philosophy of science factors out that which is not sensory consensus thereby mislabeling sound observation, as subjective.
Conceptually, way into the future, we will be able to video broadcast live dreams on a screen so all can see and verify, to satisfy the philosophy of science. But for now, this is called subjective at worse and soft science at best. The materialists philosophy and the philosophy of science were not designed to go certain places, without technology. It own limits and philosophy can cause it to deny some areas of reality, connected to output from the very tool, used to observed all material things; consciousness.
The main tool of science; consciousness, is not considered consensus objective, since we do not have a way to form an object consensus; no consensus definition. If consciousness is the main tool of science, and we cannot define consciousness in a way to make it an objective consensus observation; philosophy, how do we know if the consciousness tool, is properly calibrated, for use as the foundation of objectivity? A collective biased can exist, in the tool, that we are not aware of.
Theory, for example, can bias how we look at things and thereby created a conditioned consensus; sensory expectation. For example, at one time the earth was assumed flat. Once this premiss was taught and accepted, it would create its own reality. The consciousness tool was not properly calibrated, allowing the bias to become consensus and systematic. The lack of calibration was induced by a collective day dream, and a prestige affect. It was also induced by science not dealing with consciousness affects.
The final frontier of science; consciousness, will need to be addressed, or science can never know if the main tool of science is properly calibrated. Are we defining reality as it is, or are we day dreaming it, as we wish it to be. Science, evolves over time; past daydreams into new ones. The philosophy of science, as is, is designed to ignore this calibration, since unconscious exploration needed will result in unique subjective data that has to be pieced together, to indirectly infer objective consensus of a calibration standard.