Topic - MPEG-1 question

Reply to this topic

Topic Starter hv_a - Posted on Tuesday, January 22 2008 01:06:13 PM

I was debugging some MPEG-1 clips and hit upon the case wherein a row is not equal a slice. The way I am detecting this is that the new slice start code is a huge # compared to the previous one (not an increment by 1 which is usually the case).

In this scenario, I thought we should just go ahead processing just the macroblock() function as per the spec (page 40). But, I see that the decoder still does the "quantizer_scale_code" and "extra_bit_slice" (ie, it processes the slice elements instead of going directly to the macroblock elements)!! Any thoughts on why this is so and where is this documented?

Best Regards,