This is a suboptimal approach, since indeed the two codes (Block and Convolutional codes) are working in tandem to lessen error probability.