We propose facilitating the error annotation task of translation quality assessment by introducing an annotation process which consists of two separate steps that are similar to the ones required in the European Standard for translation companies EN 15038: an error analysis for errors relating to acceptability (where the target text as a whole is taken into account, as well as the target text in context), and one for errors relating to adequacy (where source segments are compared to target segments). We present a fine-grained error taxonomy suitable for a diagnostic and comparative analysis of machine translated texts, post-edited texts and human translations. Categories missing in existing metrics have been added, such as lexical issues, coherence issues, and text type-specific issues.