Synchronized acoustic-articulatory data is the basis of various applications, such as acoustic to articulatory inversion (AAI), articulatory to acoustic mapping (MAP), etc. Most of the studies in these fields directly trained various models with EMA synchronized speech, while the target input or output are stand-alone speech in real applications. However, the recording conditions of EMA-synchronized speech and stand-alone speech are different, which may make the EMA-synchronized speech different to the stand-alone speech and degrade the performance of downstream tasks. In this study, we attempt to present a general view of whether the differences affect the performance of AAI by training 3 latest AAI model with EMAsynchronize speech and testing them with both EMAsynchronized speech and stand-alone speech. The results indicate that the performance of all the 3 AAI models degrade dramatically in the sense of both Root-Mean-Square errors and Pearson’s correlation coefficients.