Doom9's Forum - View Single Post

StainlessS · 28th April 2012, 00:14

Jmac, if you want to try the current beta, probably about complete, maybe just a little playing with format
and set a preliminary ChromaWeight default to 0.333.

Any body else want to try it, and any results/conclusions would be good, otherwise, may keep this chromaweight of 0.333 (What I 'feel' is about right).

EDIT: I think I'm correct about most filters using a weighting of about 0.5 for combined chroma, dont myself think
this is a good measure, luma commonly fills it's full dynamic range, whereas chroma huddles about the middle,
as chroma habitually occupies considerably less dynamic range, it should count for less in any kind of weighting, even
if combining two chroma channels. If the dynmic range of chroma is somewhat reduced, then the difference between
two identical frames where one has been reduced/compressed has got to be likewise reduced and of lesser importance
as a measure. Think about 33% might be a better guestimate.

See here: (Will be deleted when new non beta published)

EDIT: Link Removed.

28th April 2012, 00:14	#6 \| Link
StainlessS HeartlessS Usurer Join Date: Dec 2009 Location: Over the rainbow Posts: 10,980	Jmac, if you want to try the current beta, probably about complete, maybe just a little playing with format and set a preliminary ChromaWeight default to 0.333. Any body else want to try it, and any results/conclusions would be good, otherwise, may keep this chromaweight of 0.333 (What I 'feel' is about right). EDIT: I think I'm correct about most filters using a weighting of about 0.5 for combined chroma, dont myself think this is a good measure, luma commonly fills it's full dynamic range, whereas chroma huddles about the middle, as chroma habitually occupies considerably less dynamic range, it should count for less in any kind of weighting, even if combining two chroma channels. If the dynmic range of chroma is somewhat reduced, then the difference between two identical frames where one has been reduced/compressed has got to be likewise reduced and of lesser importance as a measure. Think about 33% might be a better guestimate. See here: (Will be deleted when new non beta published) EDIT: Link Removed. __________________ I sometimes post sober. StainlessS@MediaFire ::: AND/OR ::: StainlessS@SendSpace "Some infinities are bigger than other infinities", but how many of them are infinitely bigger ??? Last edited by StainlessS; 27th May 2012 at 11:44.