[SFV] Hmm it's a very strange thing in the patches before this it did not happen, and in any case the materials are now unusable, I noticed that there are problems with Gildor's Umodel too, I hope it is not there some kind of half-tampering like Denuvo or similar.
Edit:
Ok after several tests something has definitely changed, first the material now they have different bytes here a sample 
https://www.dropbox.com/s/5l5sqnw6ijwj5 ... l.rar?dl=0I unpacked the game with this bat: for %%s in (*.pak) do quickbms -Y sf5.bms %%s out 
with an old script 
Code: Select all
# Unreal Engine 4 - Unreal Tournament 4 (script 0.4)
# Modified by FluffyQuack for better SF5 support
# script for QuickBMS http://quickbms.aluigi.org
# set your AES_KEY here as text or C string ("\x11\x22...")
set AES_KEY binary "_aS4mfZK8M5s5KWC2Lz2VsFnGKI7azgl"
goto -0x2c
idstring "\xe1\x12\x6f\x5a" # 0x5a6f12e1
get VERSION long
get OFFSET longlong
get SIZE longlong
getdstring HASH 20
goto OFFSET
#Acquire base name
get NAMESZ signed_long
math NAMESZ -= 9
getdstring DUMMY 9 #Skip the "..\..\..\" part of the filepath
getdstring BASENAME NAMESZ
get FILES long
for i = 0 < FILES
    callfunction GET_NAME 1
    savepos TMP_OFF
    get OFFSET longlong
    get ZSIZE longlong
    get SIZE longlong
    get ZIP long
    if VERSION <= 1
        get TSTAMP longlong
    endif
    getdstring HASH 20
    math CHUNKS = 0
    if VERSION >= 3
        if ZIP != 0
            get CHUNKS long
            for x = 0 < CHUNKS
                get CHUNK_OFFSET longlong
                get CHUNK_END_OFFSET longlong
                putarray 0 x CHUNK_OFFSET
                putarray 1 x CHUNK_END_OFFSET
            next x
        endif
        get ENCRYPTED byte
        get CHUNK_SIZE long
    endif
    comtype copy    # for AES
    if ZIP & 1
        comtype zlib
    elif ZIP & 2
        comtype gzip
    endif
    if CHUNKS > 0
        putvarchr MEMORY_FILE SIZE 0
        log MEMORY_FILE 0 0
        append
        for x = 0 < CHUNKS
            getarray CHUNK_OFFSET 0 x
            getarray CHUNK_END_OFFSET 1 x
            math CHUNK_ZSIZE = CHUNK_END_OFFSET
            math CHUNK_ZSIZE - CHUNK_OFFSET
            if ENCRYPTED != 0
                encryption aes AES_KEY "" 0 32
                math CHUNK_ZSIZE x 32
            endif
            clog MEMORY_FILE CHUNK_OFFSET CHUNK_ZSIZE CHUNK_SIZE
        next x
        append
        encryption "" ""
        log NAME 0 SIZE MEMORY_FILE
    else
        # the file offset points to an entry containing
        # the "same" OFFSET ZSIZE SIZE ZIP HASH ZERO fields,
        # just an additional backup... so let's skip them
        savepos BASE_OFF
        math BASE_OFF - TMP_OFF
        math OFFSET += BASE_OFF
        if ENCRYPTED != 0
            encryption aes AES_KEY "" 0 32
            math ZSIZE x 32
        endif
        clog NAME OFFSET ZSIZE SIZE
        encryption "" ""
    endif
next i
startfunction GET_NAME
    get NAMESZ signed_long
    if NAMESZ >= 0
        getdstring NAME NAMESZ
    else
        math NAMESZ n NAMESZ
        math NAMESZ * 2
        getdstring NAME NAMESZ
        set NAME unicode NAME
    endif
   set TMP BASENAME
   string TMP += NAME
   set NAME TMP
endfunction
The hierarchy is kept perfectly but the strangeness of the materials remains, I try the same bat file with the unreal engine 4 script (sfv key) but are created dat files in numerical sequence like 0000658 0000659 etc etc
I continue to study if I find a solution at least for the materials.