generic rescale AA from 945 lanczos taps 4

lol

lol

Second time you post this.
Please explain us why “generic rescale AA from 945 lanczos taps 4” is a bad thing.
Why not using the process you think is the best? Then posting the result with image comparison, so we can see it would be better compared to Jensen’s release.

a lot of people have come into jensen comment sections (and discord) and posted image comparisons showing why his fake rescales and chroma rape are bad but he just sticks his fingers in his ears and says that they are wrong and the way beatrice encodes is right. there is no helping this dude especially because he doesn’t understand english. stop sucking his dick.

Since 2012 (invention of insaneAA) we (Beatrice) have been told what we are doing is wrong. But since then no one has been able to achieve better final quality. Better means not adding sharpness and not removing the effects of chromatic aberrations along with non-existent ringing.
Better means visually closer to the source, but without artifacts.
The statements “here, with an increase of 800%, things are worse” also do not deserve attention, as you understand. I laughed for quite a long time at the “worst quality evidence” published by Smol, which looked like this:

I don’t intend to seriously argue about coding approaches with haters. If you are so sure that you are right, jast do better than ours. Empty statements are worthless.
So I just laugh at comments like that. After so many years you get used to it.
By the way, insaneAA algorithms determine the chroma planes untouched from the source, so I basically don’t understand what problems you’re talking about with it.

@golgoth-13 because it’s incorrect? descale exists to revert the upscaling process. it is mathematically incorrect once the image has been further altered since the upscale, which is the case for frieren

even if it was possible to descale frieren, the jpbd is lowpassed and beatrice-raws did not even attempt to counter that using web sources or alternative discs. rendering any attempts to revert the upscale entirely incorrect

Why not using the process you think is the best? Then posting the result with image comparison, so we can see it would be better compared to Jensen’s release.

So what? Exactly what competent encoders have been demonstrating for years?

Why didn’t you fix the dirty lines?

Thanks

@smol
Firstly: Do your own coding, and let the comparison happen (Sneedex is there for that);
Secondly: Or ignore it… if you don’t like the coding, just ignore it (it’s not simple 🤷🏽‍♂️), we’re in 2024, the war for better coding doesn’t exist but.

Damn try making sense

im seeing a whole lot of yapping and a whole lack of image comparisons going on in this thread

Me: Rewrite this so it makes sense.
Chat GPT: Firstly, do your own coding and let the comparisons happen naturally; Sneedex is available for that purpose.
Secondly, if you don’t like the coding, you can simply ignore it. It’s not always easy, but remember, we’re in 2024 and the competition for better coding isn’t as intense as it once was.

:joy:

@LastBreeze Thank you, my English is horrible 😐

web sources or alternative discs

There are currently no alternative publications. Perhaps in a year there will be a choice of other BDs. There will be a reason to redo the encoding.
And I consider inflating the web source to the size of a bd even because it is “lowpassed” a useless exercise. You are watching bd or web. There is no need to turn one into another. Besides, 9Volt has already done this. There is a choice. https://slow.pics/c/bvnuYncb

Amazing work!!!

Thanks for the complete Frieren BD’s!

Thank you!

Thanks