In this work we study the information provided by a detector click on the state of an initially excited two level system. By computing the time evolution of the corresponding conditioned probability beyond the rotating wave approximation, as needed for short time analysis, we show that a click in the detector is related with the decay of the source only for long times of interaction. For short times, non-rotating wave approximation effects, like self-excitations of the detector, forbid a naïve interpretation of the detector readings. These effects might appear in circuit QED experiments.PACS numbers: 11.10.-z, 29.40.-n, 42.50.-p, 85.25.-j Quantum detection theory was created to study the behavior of detectors in presence of radiation [1]. Highly satisfactory up to date, it relies on the conspicuous rotating wave approximation (RWA), which neglects the so-called counterrotating terms, irrelevant in most cases. These terms give important contributions for strong atom-field couplings and very short times as compared to the system time scale, meaning that for any effect beyond RWA to be acknowledged our measurements must be very precise and the observation times quite fast. This is particularly problematic for Quantum Optics experiments, due to the very small matter-radiation coupling and the fact that observation times must be at the femtosecond scale for most cases (nanosecond for hyperfine qubits), which is ridiculously small for current experiments (∼ µs for trapped ions [2]).However, circuit QED [3] provides a framework in which those phenomena are accessible to explore. By using superconducting qubits as artificial atoms coupled to a transmission line, one gets a system which behaves analogously as a 1-D radiation-matter interaction in Quantum Optics, while working at the microwave frequency range [4]. In this setup, parameters can be easily tuned, and the coupling between qubits and transmission line can be modulated up to ultrastrong levels. [5,6] Besides, fast qubit state readout (∼ ns) is possible using a pulsed DC-SQUID scheme [7] .In those conditions, phenomena beyond RWA have already been reported [8], [9]. In this regime Glauber's theory is no longer valid and quantum detectors should be described by a non-RWA model.A counterintuitive direct consequence of the breakdown of the RWA is that a detector in its ground state interacting with the vacuum of the field has a certain probability of getting excited and emitting a photon. There is however not a real consensus on the physical reality of this effect. Introducing counterrotating terms is interpreted by some to be a problem as the processes described by those terms seem virtual. It seems counterintuitive to accept that a detector in a ground state in the vacuum could get excited. As a matter of fact, there have been attempts of suggesting effective detector models by imposing this phenomenon to be impossible [10].We should however point out here that there is no problem whatsoever with energy conservation, as unitary evolution of the states under a ...