encoding - Can i use the same huffman table to decode and encode each frame in a video?

Question

Currently working on a code where I used motion-compensated prediction using SAD/SSD. It all works (by which I mean the PSNR and BPP are hitting the desired values) but I find that the encoding and decoding take MUCH longer than I would like.

The general process: For each frame in the video (only a couple of frames), I go through and create a Huffman table. Huffman encodes and then proceed to Huffman decode. Then I finally go through some reconstruction to create the predicted frame.

My question is, Can I use the same Huffman table to encode all of the frames instead of making them per frame?

I can't imagine that modern video encoders would make a table per frame. Feels like it would take way too long and would be video-specific.

question from:https://stackoverflow.com/questions/65920237/can-i-use-the-same-huffman-table-to-decode-and-encode-each-frame-in-a-video

Answer 1

Waitting for answers