There is a basic distinction between "literal meaning" and "communicative intent". A sentence like "The tea is in the cupboard" seems like a very simple statement of fact referring to a well-known thing, an piece of furniture, and a spatial relation. The communicative intent behind a person saying this is very complex, to the point of being unknowable. In some cases, communicative intent gets tangled up with cultural conventions so that the sentence "There is tea in the kitchen if you want some" literally says something bizarre, that the actual existence of tea in a certain location depends on the wishes of the addressee. In terms of intent, it is said with the intent of communicating that "There is tea in the kitchen, and you may have some if you want it". The specific constructions used in this way are culture specific.
A person who is truly bilingual in e.g. English and Chinese is able to find culturally-appropriate ways of communicating in English, or Chinese, depending on what language / culture they are "in". That person is not guaranteed of being able to "translate" from English to Chinese, or vice versa. Translation requires introspection about the circumstances and comparing different expressions to see if sentence A is appropriate for "this circumstance" – they have to have a clear understanding of what constitutes an specific "circumstance", and usually people are not conscious of all of those factors. The problem of translating is that you need a theory of why you say things one way vs. the other (you're not just saying things using your native-speaker knowledge, you're computing metalinguistically). Of course you might translate acceptably in many cases.
If you set aside the goal of translation, then you might ask "Are there any thoughts that cannot be expressed in some language, because of the structure of that language?". The problem with such a question is that the notion of "a thought" needs to be carefully studied – nobody knows what "a thought" is. A vague emotional state is not a "thought", and you probably cannot string together a sentence that encapsulates that state. If it's specific enough that you have an actual proposition, then it is already in a specific linguistic form (I claim: this is a point of controversy). There are some notions in the literature that there is an unknowable secret "language of thought" which everybody knows, and the task of the linguist (semanticist) is to find the translation from "language of thought" to English, or Chinese. I don't know if this idea is held by linguists anymore: I'm denying that there is a universal underlying "language of thought", instead, actual thoughts are in a language (but not all mental impressions constitute "thoughts").
