safebooru 2022.05 rip + addons from yande-re, gelbooru, anime-pictures, zerochan

So, here is DOUBLE volume V2022B for interval 01.2021-05.2022 in series of composite safebooru-based rips 11.2021 - 01.2022 volume V2022A 08.2021 - 11.2021 volume V2021D 06.2021 - 08.2021 volume V2021C 03.2021 - 06.2021 volume V2021B 12.2020 - 03.2021 volume V2021A 09.2020 - 12.2020 volume V2020D 06.2020 - 09.2020 volume V2020C 02.2020 - 05.2020 volume V2020B 08.2019 - 01.2020 volume V2020A 11.2018 - 08.2019 volume V2019 aimed to feed BOORU-CHARS OPEN DATASET 2021 and 2022

This rips not intended to be "complete and maximum quality" but rather "representative the best of" to help users not to loose interesting fandom, artist or even single prominent picture and get all stuff with several clicks. Another reason to build this megalythe is neural network training over art images. There are promising results, stay tuned.

Sources used (priorities high to low when deduplicating):

• safebooru.org (ID 38xxxxx and 39xxxxx) letter S in archive/folder name
• yande.re (with some questionable images in separate Q-folders) letter Y
• gelbooru.com (a little bit NSFW in Q) G
• anime-pictures.net A
• zerochan.net Z (WTF - I have no idea how to change this letter)

218.022 images sorted and zipped according aspect ratio (dimensions 2 folders) priorities high to low :

• ~ 74700 "artbook pages" 7x10 (+/- 4%)
• ~ 33500 “wide pages” 3x4 (+/- 10%)
• ~ 35200 “squares” 1x1 (+/- 20%)
• ~ 39200 “wallpapers and computer screens” 3x2 (+/- 40%)
• ~ 34200 "high pages" 2x3 (+/- 40%) folder name contains 1x2

and also for source and (sometimes) ID range, mentioned in folder/archive name. You can browse pictures directly in archives with FastStone MaxView of something like it.

File names structure : %website% - %id% - %up_to_3_copyrights% ~ %up_to_5_characters% (%up_to_2_artists%).%ext% where

• %copyright% , %character% and %artist% may be used as filter for search on source booru
• %website% + %id% is unique and also may be used to get direct booru url

so you can extract subsets of interest with xcopy (from already unzipped images) or unzipping (from release on the fly) e.g.

for %%F in ("d:\Safebooru 2022b\*.zip") do 7z x -r -o"e:\sortarea\" "%%F" *sono*bisque*doll*
xcopy /s d:\Safebooru 2022b\*sono*bisque*doll* e:\sortarea

Transformations and filters:

• initially filtered Mpixels >= 1.2, width >= 900, height >= 900
• PNG converted to JPG (quality 94%), no animations
• downsize to 60MPix and/or maxsize 9000 px, stripes dropped or adjusted to aspect ratio 0.4 .. 2.1
• manually (yep, plenty of handjob behind this release)
  • comic and 4koma, segmented scans and overtexted covers filtered out
  • real-life photos, no-character landscapes, most of line-arts and primitive chibi thrown away
  • too explicit images (uncensored nipples or vulva, obvious adult actions etc) excluded from "questionable" downloads
  • crops done when large simple or dirty background, most artbooks de-bordered
  • occationally gamma correction, denoise and other nontrivial improvements made
• carefully deduplicatied (with AntiDupl NET up to 4% similarity) along with several past releases

Some meta-information included in tab delimited files :

• V2022B_files.TSV post info (size, resolution, MD5 etc) with concatenated copyrights / characters / artists tags (Excel capable)
• V2022B_tags.TSV all tags (incl. general and meta) one tag per line (5.124.543 rows, not fit into Excel)

Using some database you can play with SQL and xcopy (from already unzipped images, copypasting query result) anything you want, e.g.

select 'xcopy "d:\'||torr_path||'\'||file_name||'" e:\sortarea ' xc
from files f
join tags t on t.booru=f.booru and t.fid=f.fid
where t.tag like 'ukrain%' -- support our fight for freedom any way you can

NOTE-1: several hundreds too-NSFW images excluded from ZIPs during last-minute patch, that's not reflected in TSV metadata