A linear dependent rate-quantization model for scalable video enhancement layer encoding.