Affiliated Groups: AC, FFF, Beatrice
File Extension: .mkv, x264 @ 1080p (1920x1080-JPBD) [Beatrice]
Audio Track 1: Japanese 2.0 FLAC
Subtitle Track 1: English Subtitles [FFF]
Subtitles were timed and edited fresh from FFF, There was Coalgirls release who utilized FFF but there was still a few errors and the dialogue scene bleeds weren't all fixed so I decided to start fresh and fix all the shit.
Yes I will be doing season 2 + OVA's once all sources have been checked/compared for dialogue ect...
https://slowpics.org/comparison/e865a0b1-c708-4d93-a04c-8e9791cc0571
Here's MTBB vs. Beatrice screenshot comparisons from S2, for your convenience and presented without comment.
Thanks for the screens @motob, that is one part of the comparison. Next is dialogue translation.
From the screens provided. Beatrice went stronger on the noise reduction and looses slight more detail/blurred in comparison.
*How do you take your screens? I do hope not from a media player or without any filters at least if a media player was used*
Edit: Also from what I read is this from your release? “arigato” as “thankies”, inserting words like “Totes” etc? (Please no lol)
And I guess you used qaac?
Yes, qaac.
I understand the frustration with the script (I love it but I understand why other people don't), but before you use the officialsubs (either Sentai or CR--yes, they are different) you should really consult a translator to determine whether that's a good idea.
I take my screenshots by converting the images to RGB in Vapoursynth using dither_type = "error_diffusion". Spline36 upscale.
Nope no active translator until the guy gets back who is doing tales for me. But I don't think he would want to do a whole series qc but meh who knows.
Localized scripts are hit and miss for people just like with Haganai
@motob I will do a video comp direct from the 720p source. You have upscaled where I always go with source image as its more accurate and no sharpening ect is applied. But thanks for answering my questions btw :)
If it turns out to just be better I will sort out a fresh encode script for 1080p.
Wait, you compare without doing upscaling? How does that even work? I can understand not liking Spline36 as a kernel, but how do you even do a comparison without upscaling first? Wouldn't you need an image processor in your brain in order to do an effective comparison? If I looked at a 720p vs 1080p encode without doing any scaling, I'd nod my head and say "yup, yup, the 720p looks sharp and the 1080p looks blurry" and not be able to draw any more conclusions than that.
When you upscale a 720p encode to 1080p, you're simulating how that encode will look full-screened on someone's 1080p screen. A raw screenshot of a 1080p encode is also how that encode will look fullscreened on a 1080p screen. So doesn't it make sense to upscale the 720p and compare the two?
The best way for quality of image checks for comparison is to simply go onto aegisub save a frame as png and compaire straight. Also I do my checks with a 4k monitor and 1080p tv.
Just zoom to line detail, banding ect
For example the process you used would turn far different results with NGU or the more commonly base used jinc (lower end) in media playback with madvr
>The best way for quality of image checks for comparison is to simply to onto aegisub save a frame as png and compaire straight.
You literally did not address any of my questions or points at all regarding upscaling.
It's certainly true that upscaling leads to different results depending on what scaler you use. But I can think of at least one thing where (1) the type of scaler doesn't really matter and (2) the comparison is very, very hard to do without scaling: background detail loss. When you're trying to see if an encode lost detail and you're not doing a direct (i.e. same-scale) comparison, you basically have to to have a pixel-perfect photographic memory when you switch back and forth between the screenshots unless the detail loss is REALLY different. I would also find it difficult to compare how well an encode dealt with banding/noise--or even something as simple as encoding artifacts--unless, again, the difference was substantial.
As a broader matter, you can account for scaler differences by understanding the nature of the scaler you're using. Like if I see that my Spline36 upscale is rather sharp and also has some ringing--well, that's just the nature of the upscaler, and I can hold that in the back of my mind when I'm doing comparisons. If my upscaled 720p images are sharper than the 1080p images, then I know not to take it all that seriously since the viewer's experience may be different.
I have Maid sama Dual-Audio with Beatrice and some subtitle source I can't remember atm but not releasing due to the one day HTT release if it ever happens(they were doing a heavy script edit) and hopefully they do Love Live season 2 as well one day lol.
hello buddy. i have a query regarding seedbox. i hope you can help me.
can i download or seed a torrent,with the help of seedbox, even if the torrent has 0 seeds.
No you need a seed to get 100% of the torrent before you can start to seed.
Peers may help you get some of the way there but to finish you need a seed. Or if you allready have the file downloaded you can just open the torrent to start seeding after the check
ok, Thnx for the reply. Just to be clear, i have 2 questions.
1. i cant download a dead torrent even with a seedbox?
2. can i seed a dead torrent with a seedbox?
1 = No you may get some of the way there depending on peers but you wont get the whole torrent without a seeder even with a seedbox.
2 = you can only seed what you have downloaded.
I love and appreciate your work but just to ask why is the file sizes relatively smaller when compared to the coalgirls release by 100-200mb? Please do not misunderstand I just wanted to know and Thanks once again.
The are in that thread you linked just look.
Though I edited them for myself with a resize and positioning. They were too small and low down on the video for my liking so fixed that up. Next to no one puts subtitles at the verry bottom of a video.
My edited Subtitle for the Move here - https://drive.google.com/open?id=1Tz5zlFoiJGSRUkE-vWjg-AKoBOxk6BVK
@AC can you tell me what is going on with nanda.to? It is down again right now, it was down most of yesterday and the day before. Confirmed with downforeveryoneorjustme.com
Comments - 31
SomaHeir
AC_ (uploader)
motbob
DR3AMKiiLL3R
Astral
AC_ (uploader)
motbob
herkz
AC_ (uploader)
herkz
motbob
AC_ (uploader)
motbob
AC_ (uploader)
motbob
AC_ (uploader)
Arisa24
AC_ (uploader)
avisingh
AC_ (uploader)
avisingh
AC_ (uploader)
Rick0Namikaze
Rick0Namikaze
AC_ (uploader)
freakicho
FrostbiteStardust
AC_ (uploader)
FrostbiteStardust
AC_ (uploader)
FrostbiteStardust