[INDEX] Demon Slayer Season 1-5 {鬼滅の刃 | Kimetsu no Yaiba} [ITA + US.BD][AV1 | 10bit][1080p][FLAC] (Dual Audio + Multi Sub)

Nvenc again for those who don’t go look at the very bottom of the mediainfo

@NekoTrix

What are you even talking about??? Does that even mater? No.

A lot of people felt the same way until they looked into it and determined it makes a difference and here’s why.

Some people assume that software AV1 encoders, like SVT-AV1, and NVENC’s AV1 hardware encoder produce identical results because both generate modern, efficient AV1 video. However, the difference in approach significantly impacts the quality per bitrate, with software generally offering better detail and fewer artifacts.

Software-based AV1 encoders utilize very complex, computationally intensive algorithms that can analyze frames deeply and are updated over time via open source. These are designed to maximize AV1’s compression efficiency, despite requiring more time and CPU power. That’s why at any given bitrate, software-encoded AV1 tends to maintain more detail, handle motion better, and avoid artifacts more effectively.

The AV1 encoder in NVENC, though impressive for real-time use, remains limited by its hardware implementation. It is optimized for speed and low latency rather than trying dozens of expensive possibilities and then picking the best. So, at the same bitrate, NVENC usually shows slightly more smoothing, mild artifacts in high-motion scenes, or reduced fine detail compared to top-tier software AV1 encoders.

In other words, NVENC AV1 excels in streaming and real-time recording, where speed is prioritized over maximum quality. But for absolute efficiency and quality per bit, especially in offline encoding or archival content, software AV1 encoders are still superior.

So does that even “mater” to you specifically @Mr_Kimiko? Probably not.

What are you even talking about??? Does that even mater? No.

It does matter to a lot of people, lol. Go read up more on this topic before posting such a shit take.

Uh oh, look what you did

I didn’t encode those. So don’t blame me.

Don’t worry, you’re only getting blamed for your uninformed take, but it’s all good, teachable moment.

Was it like this:
“Wow who does this Trix person think they are?! I’ll post this in response:”
"@NekoTrix
What are you even talking about??? Does that even mater? No."

"Now who the heck is this guy? Does he even know anything about av1 at all?, let’s find out…“
https://www.google.com/search?q=nekotrix
"Oh no…”

@LastBreeze

Not in that photo you posted. Someone is clearly blaming me for these encodes when it wasn’t me.

https://animetosho.torrentbay.st/s__cache/nyaasi/view/2042077
I think everyone knows you’re just the poster, not necessarily the encoder.

There is no world where NVenc is the preferred encoder for anything related to storing anime or movies long-term, for anyone except the most braindead, blind anime watchers who have fewer braincells than they do TB of NAS storage.

I honestly don’t know why you keep wasting your time.

Fixed-feature hardware encoders were designed for streamers and for intellectually lazy, incurious people who just want to archive their home blu-rays and don’t care about detail loss, smudging, ghosting, and blocking artifacts (among other problems). Just because you do not have the skill or patience to find the differences and problems between source and your encoding results does not mean there are no differences or problems.

Hardware encoders are fixed-feature. That means they are as good (or as bad) as they will ever be as soon as the chip fab etches the silicon wafers. They cannot be improved with driver updates or software updates.

Software encoders get regular improvements for quality and efficiency, I think SVT-AV1 mainline has been getting regular improvements almost every quarter. Not to mention the various forks which have improvements for different types of content or which perform differently based on a particular community’s preferences. Can’t do that with hardware encoders.

I don’t understand this mission you are on. You aren’t proving anything to anyone about NVenc. You will never get featured on any of the preferred-release databases, nor on curated trackers. Nobody will ever respect what you are doing.

Just admit you are too lazy or not smart enough to learn the software encoder features.

By all means, continue to use NVenc for your private collection. Nobody else is impressed.

To be fair, all we asked for was transparency, and you go out of your way to do the opposite.

Trying to hide that you’re using NVenc by not including mediainfo really shows you don’t actually believe your own claims that NVEncC is superior at high bitrates. Someone was nice enough to share rigaya’s benchmarks, and even with the outdated metrics ssim and vmaf and unusual preset choices, they clearly do show SVT-AV1 3.1.0 outperforms NVEncC 8.08 even at preset 5 on 10bit anime encodes in both ssim and vmaf in --rc 0 (CRF) and --rc 1 (VBR). The gap is probably much wider with reasonable settings as these benchmarks only list bitrates, it doesn’t actually mention what values were used for crf. Neither mode, --rc 1 and --rc 2, are suitable for high quality encoding and you really should only be comparing to --rc 0, Constant Rate Factor mode with stated values.
https://rigaya.github.io/vq_results/

meme release, just like every other INDEX release

I think the AV1 people would be plenty happy even if you used a decent x264 or x265 config. Don’t think they care at all that it has to be specifically AV1, Kimiko.

hmu I’ll give you some decent baseline x264 settings

What is this for bullshit. Bro I even complimented you in another release and also complimented you for at least including mediainfo, even while I disagree with the use of NVENC. Now I fully think you deserve to be blacklisted on every release database.

Thanks for the release and dont care too much with the toxic stuff.
Someone will always complain, noone can do anything about that.