Best Compression Methods for 'Specific' Games. Q&A

[Masquerade]

Anyone knows how long it takes for oodlrec to parse cyberpunk?

It's really damn slow, was wondering if anyone knows how long it's expected to go for.

Update: takes scooter ~4 hrs to get to 4.2%

[parnexcodes]

Cyberpunk 2077 (Based on Codex) English Only

Original Size : 59GB (Excluding 3 GB Bonus Content)
Repack Size : 39.5 GB

Method : oo2reck+srep+xz (LOLZ much better but very slow)

Use oodle7 file from \bin\x64\

Compression time : 18hours
Decompression time : 2hr50m on 4core 8gb ram HDD

[Gehrman]

English Only

Original Size : 59.6GB
Repack Size : 40.1 GB >>>>> 59.6GB> 125GB > 84.5GB > 40.1GB

Method : xtool+srep:m3f:l512:m512+4x4:lzma:a1:mfbt4:32m:fb2 73:mc10000:lc8+diskspan:4.25GB:4.32GB

Use oodle7 file from \bin\x64\

Compression Time : 4Hours
Decompression Time : 70Min on 6Core 16GB Ram |RZ1600

Arc:

[External compressor:xtool]
header    = 0
packcmd   = XTool.exe precomp -mcp2077 -c128mb -t100p --dbase - - <stdin> <stdout>
unpackcmd = XTool.exe decode -t100p - - <stdin> <stdout>

[Masquerade]

Spent ages patching videos out of Cyberpunk 2077 just for bpk to do that endless extraction thing.

What a waste of time

Managed to get it to 36.9GB by patched out videos but since bpk does that thing you can only use srep leading to a 37.5GB repack (xtool + srep + lolz).

For the audio packs, xtool+srep+lzma2 (lolz is not worth it)

I am going to post my data sheet, here's all the videos with their decimal locations from the game ranging from largest size to smallest size (40mb) - all smaller ones were ignored.

basegame_5_video.archive

938983424 - 1683451019 - 00000000000001ea.kb2

3865538560 - 4011937979 - 000000000000012f.kb2

2924453888 - 3061543279 - 0000000000000176.kb2

6979723264 - 7100791887 - 000000000000014c.kb2

7572389888 - 7688552327 - 000000000000013a.kb2

7688552448 - 7804532159 - 00000000000000dd.kb2

3727880192 - 3843607171 - 000000000000001a.kb2

7390752768 - 7504962763 - 000000000000009a.kb2

4080529408 - 4190263323 - 00000000000001f5.kb2

3061547008 - 3157061311 - 0000000000000025.kb2

6622117888 - 6717121191 - 000000000000012c.kb2

6840016896 - 6918851327 - 000000000000005e.kb2

4536360960 - 4627345415 - 0000000000000153.kb2

7869325312 - 7959684047 - 000000000000009c.kb2

8103661568 - 8193355355 - 0000000000000001.kb2

8013971456 - 8103657939 - 0000000000000220.kb2

3157061632 - 3243022439 - 00000000000000ea.kb2

4990582784 - 5068423239 - 000000000000008a.kb2

332349440 - 406449271 - 00000000000000a5.kb2

1815740416 - 1889706651 - 00000000000000e3.kb2

7317245952 - 7390749839 - 00000000000001f6.kb2

261668864 - 332348067 - 00000000000001c8.kb2

3477536768 - 3546653107 - 0000000000000035.kb2

4190265344 - 4259151603 - 0000000000000187.kb2

4011941888 - 4080526847 - 0000000000000006.kb2

7504965632 - 7572386623 - 00000000000000a4.kb2

3410862080 - 3477533547 - 00000000000000e0.kb2

4439044096 - 4505519859 - 000000000000008c.kb2

6717124608 - 6783207623 - 000000000000021f.kb2

6167703552 - 6232610759 - 000000000000019f.kb2

7804534784 - 7869321931 - 000000000000017a.kb2

5819109376 - 5881040687 - 000000000000004f.kb2

58392576 - 119832895 - 00000000000000ec.kb2

7256686592 - 7317243491 - 00000000000001c4.kb2

4685377536 - 4745411375 - 0000000000000017.kb2

3600072704 - 3659929931 - 00000000000000b3.kb2

6397444096 - 6456784187 - 0000000000000119.kb2

4096 - 58388843 - 0000000000000169.kb2

4627349504 - 4685375843 - 00000000000001c0.kb2

1901895680 - 1959742587 - 00000000000000f1.kb2

1697021952 - 1754661895 - 00000000000001d3.kb2

3354673152 - 3410859227 - 00000000000001b4.kb2

4868861952 - 4924448247 - 00000000000000db.kb2

7959687168 - 8013970607 - 000000000000013f.kb2

5758160896 - 5812178115 - 00000000000000f8.kb2

3546656768 - 3600069203 - 00000000000001f7.kb2

820768768 - 874167455 - 0000000000000022.kb2

5707190272 - 5758158895 - 0000000000000020.kb2

4259151872 - 4309797919 - 0000000000000095.kb2

6571778048 - 6622114639 - 00000000000001af.kb2

6500806656 - 6551024559 - 00000000000000b2.kb2

4333166592 - 4382905099 - 0000000000000067.kb2

5314310144 - 5363969939 - 0000000000000105.kb2

3281571840 - 3330435919 - 0000000000000051.kb2

8464351232 - 8510832983 - 000000000000020a.kb2

5183967232 - 5229091915 - 00000000000001d5.kb2

406450176 - 450814239 - 0000000000000192.kb2

6456786944 - 6500805147 - 0000000000000130.kb2

8744194048 - 8788206363 - 000000000000007a.kb2

5271851008 - 5314310055 - 00000000000000e6.kb2

[KaktoR]

@Masquerade

0000000000000017.kb2

Compressed 1 file, 60,033,840 => 50,479,677 bytes. Ratio 84.09%
Compression time: cpu 0.02 sec/real 4.31 sec = 0%. Speed 13.93 mB/s
All OK

Extracted 1 file, 50,479,677 => 60,033,840 bytes. Ratio 84.09%
Extraction time: cpu 0.05 sec/real 4.75 sec = 1%. Speed 12.64 mB/s
All OK

Edit: After extracting all files from the archive and test binkpack again on all at once, you get some error warnings like "MotionData_warn". So I guess you have to separate all the "faulty" files (I guess they contain alpha channel or something). Hopefully someday profrager will update this tool and add at least a switch for this kind of files to skip them...

[Masquerade]

@kaktor I get no such "motion data warn" on the files I've patched.

And the file you tried works also for me, but on the credits video, example:

00000000000001ea.kb2

It's stuck

[KaktoR]

Yeah then it's something like a bug inside bpk or something. I compared two videos from this (0000000000000017.kb2 + 00000000000001ea.kb2) with RADTools but they seem to be identical (except for the duration and w/h ofcourse).

I'm corious: Did you test it with the exe file from binkpack?

[Masquerade]

Originally Posted by KaktoR

I'm corious: Did you test it with the exe file from binkpack?

I haven't used bink_pack.exe since many months ago since it would always fail to pack the files properly. I could try again now.

[Snapppr6]

Originally Posted by Masquerade

I haven't used bink_pack.exe since many months ago since it would always fail to pack the files properly. I could try again now.

can iget .exe version of binkpack and binkunpack.exe and thx your repack site look amzing

[78372]

1. Audio files (x64\audio\*) 16.6Gb>15.6GB : lzt:49
2. Other rpf files except x64d.rpf 42GB> 16.2GB : xtool:mzlib+preflate:c256mb+srep:m2f:l1024+4x4:lzm a:64mb:normal:bt4:273:mc1000:lc8
3. x64d.rpf 1.68GB>1.31GB : bpk
4. Everything else 100MB>70.1MB : -mx

Total: 60.4GB > 33.3GB

[rishikrap]

What is the best compression method for Rocket League (Epic Games official Version)
Also, if anyone could link me to a tutorial which teaches how to compressit, that'd be a help.
U have UltraArc, just don't know how that works and what options to select.

[Masquerade]

Originally Posted by rishikrap

What is the best compression method for Rocket League (Epic Games official Version)

uelr+srep+lolz
(providing the epic game version is same as steam version)

[rishikrap]

Originally Posted by Masquerade

uelr+srep+lolz
(providing the epic game version is same as steam version)

It's different to some extent even though the folders are same, when I put the steam repack from fitgirl in the Epic Games folder, it had to redownload about 3.8gb.(Could be because of the crack too)
I used bpk+cp_2077(Xtools 2020)+xzlib+srep_new+lolz_mtx to achieve 48% compession i.e. from 12.9gb to 6.3gb.
Hope this helps anyone else in the future.
I think this is the absolute best conversion achievable since Fitgirl's old steam version was converted from 11.1gb to 5.5 i.e. also ~48%
Also any moderator looking at this you could add this to the Specific Game Index.

[KaktoR]

Your chain does not make any sense, just saying.

[Masquerade]

Originally Posted by rishikrap

bpk+cp_2077(Xtools 2020)+xzlib+srep_new+lolz_mtx

uhhhh ok

1. There's no need for bpk since there's no bink videos in this game

2. Why are you using the cyberpunk 2077 precompressor on not cyberpunk 2077??????

3. xzlib - we're getting somewhere, but this game is made in unreal engine 3 and is compressed with lzo and not zlib (zlib is ue4) - which is why you need to use UELR [Unreal Engine LZO Recompressor]

4. srep, great!

5. lolz_mtx - this has been proven to be slower IIRC and not worth whatever result you get. If you want a fast compression, use lzma2.