How many times have ever asked yourself: "Can I trust my satellite experiments' outcome?". Performing experiments on real satellite system can either be (1) costly, as the radio resource may be scarce or (2) not possible, as you can hardly change the waveforms transmitted by the satellite platform. Moreover, assessing user applications QoE can hardly be done using only simulated environments while the QoS modeling of a satellite system can often lead to non-conclusive or ambiguous results. The aim of this paper is to bring out representative solutions allowing the networking community to drive consistent experiments using open-source tools. To this end, we compare Mininet and OpenSAND satellite emulator to a real satellite access provided by CNES. We consider VoIP traffic to analyze the trade-off between reliability of the results, ease of use and reproducibility of the experiments. Pros and cons Result fidelity Recommended VoIP use cases Satellite QoE Satellite QoS Real satellite access [2] Pros: realism, transparency, advanced satellite features (e.g., performance-enhancing proxies) Cons: infrastructure access, shared access for some commercial offers, cost, complex architecture Mean Opinion Score Packets dropped Average jitter Max. one-way delay Performance evaluation, tests prior to commercial deployments Mininet [3] Pros: easy deployment on a single machine, shared user space with virtualized network Cons: user should create satellite topology from scratch using hosts, switches and Mininet's API Mean Opinion Score Packets dropped Average jitter Max. one-way delay Stress tests OpenSAND [4] Pros: fine-grained configuration (e.g., carriers, modulation types), emulation of end-to-end satellite communication system with resource allocation Cons: several machines required (3), manual management of the whole system Mean Opinion Score Packets dropped Average jitter Max. one-way delay Performance evaluation, system dimensioning