Quantization Method

[teamhawaii]

Anyone know the difference between non-linear and linear and how would one change such method in TMPGEnc?

I've checking my MPG2 encodes with Bitrate Viewer and found that even though I use VBR, my bitrate looks rather constant yet my quantization level is rather high. I dont know if that is a good thing or bad thing..

[tacosalad]

Bitrate Viewer graphs the quantization_factor value, rather than actually measuring quantization. (It's not feasible to measure without comparing with the source.)

An encoder adjusts the quantization_factor to maintain the desired (or required) bit rate. Higher quantization reduces the bit rate, but reduces quality. Bit Rate Viewer might identify scenes with unusually high quantization values, but you should judge for yourself whether that actually impairs the quality of that particular scene.

As for linear vs. non-linear, I'm not certain where you encountered these terms so I don't know the context. Mpeg does include linear and non-linear quantizaiton matrices, and TMPGEnc lets you choose these yourself. The mpeg standard suggest a linear matrix for intra macroblocks, and a flat (non-linear?) matrix for non-intra macroblocks.

[teamhawaii]

Ahh I see.. so is there a way to adjust the factor in TMPGEnc? I'm seeing a rather large factor in Bitrate Viewer, averaging 10.5+ for an MPG2 file I'm trying to create. Is this good or bad?

What I want to see is if it improves quality of the video. I've tried using MPEG standard and CG/Animation and the factor is the same.