vagonumero13 tools (REDELBE, rdbtool, g1mtools, doa6decsave) Update 17 Aug: REDELBE 3.0

[Yeyrahk]

I am getting an error when i try to apply a mod to the game DOA6 using REDELBE 

Failed to apply patch "Layer2:PatchBGS"

Can some one help me out please

The mod I am trying to apply is [DOA6] SaafRats Nude Mod v1.0 1.0.0 and I did install this into the REDELBE folder correctly, I think the .dll for REDELBE is not working for me.

[vagonumero13]

5 hours ago, Yeyrahk said:

I am getting an error when i try to apply a mod to the game DOA6 using REDELBE 

Failed to apply patch "Layer2:PatchBGS"

Can some one help me out please

The mod I am trying to apply is [DOA6] SaafRats Nude Mod v1.0 1.0.0 and I did install this into the REDELBE folder correctly, I think the .dll for REDELBE is not working for me.

You are using old version of game.

[gatto tom]

Vago do you know where are located the soundtracks/bgm? ?

[vagonumero13]

15 minutes ago, gatto tom said:

Vago do you know where are located the soundtracks/bgm? ?

Maybe SE1_Common_BGM.srst, in RRPreview, That's a pretty big file, so I guess it must be that containing all bgm.

They seem to use the Ogg codec but with custom container.

[gatto tom]

57 minutes ago, vagonumero13 said:

Maybe SE1_Common_BGM.srst, in RRPreview, That's a pretty big file, so I guess it must be that containing all bgm.

They seem to use the Ogg codec but with custom container.

Any idea on  how to extract them from there?

[vagonumero13]

10 hours ago, gatto tom said:

Any idea on  how to extract them from there?

It would require to write a program to convert the outer header. Normal ogg files seem to have an outer "OggS" header, and then followed by "OggS" header for each audio data segment or whatever.

 

But looking at these files, in the content inside, they only seem to have the later, and not the main "outer" ogg header. So extracting them with a hex editor is useless, I already tried that and VLC didn't like it. I also tried this: https://github.com/FelixdelasPozas/OGG-Extractor/releases

And it didn't find anything.

[vagonumero13]

51 minutes ago, HyperBob said:

Can you share an extracted file ?

 

Here is a sample of the first music track in SE1_Common_BGM.

The file "1.bin" is the file as it is. With that "OfeP" signature which I don't know what it is, it may be a Team Ninja thing.  In the .srst files, each of this is preceded by a "KOVS" chunk, but the KOVS chunk just tells the size of the file.

The first ogg segment would start at offset 0x1002 (signature OggS). There are lots of this "OggS" in the file, that is normal in ogg files.

 

The file "11.bin" is just the same file, where I removed the first 0x1002 bytes so that it started at the OggS signature.

 

None of them play in VLC. Looking at a normal .ogg file, it looks like they need another main "OggS" header, which these things don't have (maybe the "OfeP" takes that role, albeit 0x1002 bytes seem like quite a large header).

1.bin 11.bin

[vagonumero13]

39 minutes ago, vagonumero13 said:

 

Here is a sample of the first music track in SE1_Common_BGM.

The file "1.bin" is the file as it is. With that "OfeP" signature which I don't know what it is, it may be a Team Ninja thing.  In the .srst files, each of this is preceded by a "KOVS" chunk, but the KOVS chunk just tells the size of the file.

The first ogg segment would start at offset 0x1002 (signature OggS). There are lots of this "OggS" in the file, that is normal in ogg files.

 

The file "11.bin" is just the same file, where I removed the first 0x1002 bytes so that it started at the OggS signature.

 

None of them play in VLC. Looking at a normal .ogg file, it looks like they need another main "OggS" header, which these things don't have (maybe the "OfeP" takes that role, albeit 0x1002 bytes seem like quite a large header).

1.bin 6.2 MB · 1 download 11.bin 6.2 MB · 1 download

 

Uh... I just found some thing curious. If you "xor" OfeP (4F 66 65 50) with "00 01 02 03", you get "OggS". This cannot be a coincidence. Specially because the next bytes seem to follow the logic "04 07 06 07", which, after being xored with "04 05 06 07" become "00 01 00 00", which is a typical follow in a ogg header.

 

Or in other words, the OfeP is just the main OggS header with this stupid (and childish, I'd dare to say) scrambling.

[vagonumero13]

And my theory proved correct: the first 256 bytes (0x100) of each track are scrambled with the sequence "00 01... FF".

I attach first file after being unscrambled (rename with ogg extension, and you can play on vlc, I don't attach it iwth .ogg extension, because this thing detects it as a video).

 

If I'm bored, I will create later or tomorrow some program to extract these ones from .srst files.

 

1_unscrambled.bin

[vanvandеr]

I downloaded these tracks on LL about 3 months ago.

 

doa6_soundtrack.7z

 

[vagonumero13]

41 minutes ago, vanvandеr said:

I downloaded these tracks on LL about 3 months ago.

 

doa6_soundtrack.7z 68.71 MB · 1 download

 

 

That looks like someone just recorded the sound output and then converted to .mp3. The srst extraction will allow to have the original audio files.

 

[vagonumero13]

18 minutes ago, HyperBob said:

Ok, I was thinking about AFS conttainer or ADX files like old DOA Games but this is not that.

Does ogg can have infinite loop like ADX ?

Apparently oggs file can have tags, that some players may interpret for looping, but not sure if that's standard or if the game engine would enforce it.

Also, these ogg files are inside other container with "KOVS" signature, which is also preceeded with a another header with bytes 09 D4 F4 15. If they implemented loop support, they my have done it their own way using the extra data of those containers of them (only theory).

[vagonumero13]

Added srsttool 0.1, a tool to extract the ogg files from .srst.

At the moment only extract, it may evolve into injecting in the future.

[wingzg]

Thanks for the srsttool.  I'm looking for the character voices but I don't think they're any of the srst files.  Is it possible to extract the srsa ones? 

[vagonumero13]

1 hour ago, wingzg said:

Thanks for the srsttool.  I'm looking for the character voices but I don't think they're any of the srst files.  Is it possible to extract the srsa ones? 

 

I will add srsa support tomorrow. I used to think srsa only had subtitules "pointers" (as subtitles  text are in encrypted files in numeric folder), but I took a look again and I found they have ogg audio data too (with same ridicolous scrambling). I guess that since they are voice, they are in very low bitrates and possibly in mono-channel, thus the low file sizes.

 

It will be easy to do, the srsa shares same structures as the srst.

[wingzg]

44 minutes ago, vagonumero13 said:

 

I will add srsa support tomorrow. I used to think srsa only had subtitules "pointers" (as subtitles  text are in encrypted files in numeric folder), but I took a look again and I found they have ogg audio data too (with same ridicolous scrambling). I guess that since they are voice, they are in very low bitrates and possibly in mono-channel, thus the low file sizes.

 

It will be easy to do, the srsa shares same structures as the srst.

Awesome, thanks.  Looking forward to it.

[vagonumero13]

It looks like some srsa files do actually have tracks that use the adx codec instead of ogg. So, those will be extracted as adx, you will need additional programs to play those.

 

 

[vagonumero13]

2 hours ago, HyperBob said:

Cinepak will do the job, or adx plugin for foobar.

 

 

Well, at the end those tracks aren't adx. I got confused because they start with "80 00", like adx files. The files are not recognized by vlc (aka libavcodec).

Also, the license screeen doesn't mention criware, which would happen if adx was used. Well, they use some things without crediting, like the Crypto++ library, but in addition to that, I don't see in the exe the typical strings you find in games that use Criware libraries.

 

Since it is only used for voices, they may have implemented some simple codec on their own, maybe a variation of adpcm or something alike. Something with decoding speed in mind.

 

The EV_XXX.srsa files are fine, those use .ogg format. Those contain the character quotes at begin/end battles.

It is the CH_XXX.ssa which use this unknown format. I suspect they are the voices in the fight, but no way to confirm it yet.

 

I attach a sample of a file of this unknown format just in case there is someone that can recognize it. It comes from CHA_AYA.srsa.

unknown.bin

[gatto tom]

Thank you for your hard work! I didn't even notice that you already created a tool for that! It will be very useful. 

 

Do you think that EV_XXX.srsa files already have the upcoming intro and win quotes of NiCO and Diego?

[vagonumero13]

1 hour ago, gatto tom said:

Thank you for your hard work! I didn't even notice that you already created a tool for that! It will be very useful. 

 

Do you think that EV_XXX.srsa files already have the upcoming intro and win quotes of NiCO and Diego?

 

The Nico contains these files (never mind the dummy.ogg, other EV_XXX seem to have one of these empty ogg files too, not sure why).

EV_NIC.rar

[vagonumero13]

3 hours ago, HyperBob said:

I tried almost every format of ADPCM and PCM with goldwave but no result...

I tried only on 8000 Hz.

 

The only thing I could say is that the sound becomes more and more acute

The weird thing about that format is that I don't see any number that looks like a frequency. And looking at various files, I don't see anything either that could be the nmber of channels. That file in concrete has a "01"  that could look like it is the number of channels, but that seem to be a coincidence because there are other files where that number reaches all types of values in a short range, including 00.

 

Edit: I was actually missing some things at beginning of header (e.g. the uploaded unknown file was not complete). It looks like frequency is 48000 hz in these files (seem like a bit high for such small files...). I also found some similarities to this here: https://zenhax.com/viewtopic.php?t=3641

 

In concrete, some magic numbers coincide. Too bad the guy there didn't managed to decode it ?

Also found this: https://hcs64.com/mboard/forum.php?showthread=50551

 

Everyone seem to agree it is the ADPCM, I guess I will try several things myself too.

[vagonumero13]

Here i attach two files. fv_aya_appeal_01.bin is the same as before, with the complete header now.

I guess that the "haaaa" file contains Ayane saying "haaaa", but who knows.

 

The info of the 48000 Hz may have been wrong, but the number in the position matching what that guy of that forum said is always surrounding 48000, but not always exactly 48000....

 

Edit: I think the 46 in "80 00 46" may be the so called "block align", as there are some patterns at each 0x46 bytes. In the case of the "haaa" file, the sound raw data starts at offset 0x90.

 

fv_aya_appeal_01.bin fv_aya_battle_fun_01.bin fv_aya_doa5_battle_haaaa.bin

[gatto tom]

2 hours ago, vagonumero13 said:

 

The Nico contains these files (never mind the dummy.ogg, other EV_XXX seem to have one of these empty ogg files too, not sure why).

EV_NIC.rar 535.8 kB · 3 downloads

Sadly there isn't anything new. So EV_XXX have also japanese voices? because the file name has US at the end, so i thought there were only the eng voices

 

 

[vagonumero13]

3 minutes ago, gatto tom said:

Sadly there isn't anything new. So EV_XXX have also japanese voices? because the file name has US at the end, so i thought there were only the eng voices

 

 

Well, this was EV_NIC.srsa, that only has japanese voice. The one you are talking about is "SE1_Common_EV_NIC_US", which is the english equivalent. It has same files with same names, just with an "us" in them. 2 entry and 3 win quotes.

[vagonumero13]

Good news with the adpcm thingy. I was able to succesfully manually craft a wav-adpcm with the ayane "haaa" file (attached sample) by copying the data into a wav-adpcm created by ffmpeg and then tweaking the header.

 

File attached (there are still some things wrong in header like bitrate, data size, fiel size and num of samples, but it plays fine in vlc, I will correct those when I make these programatically).

 

adpcm_test - copia.zip