Hacking Final Fantasy Type-0 RomHacking

[GHANMI]

Of course arguing about small details such as subtitles for backgrounds/font used/honorifics/rebalancing/sub vs dub/speech patterns/patch format/English names ... won't be useful to the project.
Remember Gods Eater Burst and recent One Piece games had most of the Japanese in the background left untranslated. Terranigma and Wind Waker, translated by Nintendo of all people, have Japanese text left all over the place.
At most I'd see personally these subs as an icing on the cake, but nowhere near crucial. As far as I'm concerned, they could be added to the actual subtitles for the spoken dialogue or even omitted altogether, it's good to see they are even trying to reverse engineer a video format just for this minor detail (such a feat hasn't even been done for the MODS video format for the DS, otherwise the "easiest" console to hack games for). We should be grateful the game is in English at all.

 

[HappehLemons]

I would strongly suggest not going overboard. We've seen projects flatline because they were too picky about how the final product turned out.

90% of the people who are going to end up playing this just want it for that: to _play_ it. They're not looking for perfection. Considering Square up and abandoned them, having _something_ is better than what's available right now.

And consider the history of official translations has always been imperfect. Look at the older FF games of the SNES era. FF Tactics. Famously botched, but still were published because they were good enough.

Don't worry so much about subtitles over videos and discerning the writers intent in a language that doesn't directly translate. It should be readable and let people fill in the gaps.

Couldn't agree more with this! Really... we all just want to play it. You've done an amazing job assembling this team sky, and I appreciate all of the work that you and the team has put into everything. Doing all of this without asking for a dime is insanely passionate and you're making so many people happy with this project. All I can say is thank you so much everyone who's working on this.

 

[Ziggi_Strife]

The team alredy said they are going to make it as close to an offical realese as possible. in the end its for the to decide.

 

[SkyBladeCloud]

I would strongly suggest not going overboard. We've seen projects flatline because they were too picky about how the final product turned out.

90% of the people who are going to end up playing this just want it for that: to _play_ it. They're not looking for perfection. Considering Square up and abandoned them, having _something_ is better than what's available right now.

And consider the history of official translations has always been imperfect. Look at the older FF games of the SNES era. FF Tactics. Famously botched, but still were published because they were good enough.

Don't worry so much about subtitles over videos and discerning the writers intent in a language that doesn't directly translate. It should be readable and let people fill in the gaps.

Thanks, and I do agree, it's just that there is no other way for us to translate those "video texts". I have to say that right now, they are the least of our priorities anyway

i see..it is difficult to do if there are no video editors, so it will take some experiments..

Btw, is there any changes about the translation progress?

Yes, every Sunday, team member Adam S makes a status report telling the rest of the team how much time will takes us to finish the main scenario file. I'll let you guys know about it later when he does it at night. Last week it was:

7 translators 10 lines of actual Japanese a day each = 12 months
7 translators 20 lines of actual Japanese a day each = 6 months
7 translators 30 lines of actual Japanese a day each = 4 months
7 translators 40 lines of actual Japanese a day each = 3 months
7 translators 50 lines of actual Japanese a day each = 2.5 months

I'd say that translators have a better pace now that they are used to our working style, so I'm pretty sure the report will be even better today ^^"


Sky

 

[Swing]

Hey I've been translating I just haven't sent anything over yet. I've also been playing the game to try and understand more of the story so I'll do a better translation job since I'm 2 chapters from beating the game I think. On a separate note, I've had my VITA in it's JP account so I can't Skype as that's on my American account because I've been waiting for the release of the next hack for the latest exploitable game. It's been released but the game for the exploit costs 50 dollars American for a stupid JP PSP mini for the VITA. I was wondering if anybody knew if it would be worth it because I know it can play some SNES and N64 emulators, but it isn't a CFW, but do you guys know if once I get the VHBL or whatever it is for my VITA would I be able to downgrade so I can get the REAL CWF for it so I can play ISO's like Type-0 on my VITA??

 

[SkyBladeCloud]

Ahh cool Swing, too bad you can't get with the rest on Skype.
Also remember to save your work and do a SVN Update every now and then to make sure we work on the same version

BTW just saw your email and nope... I don't have a fb... just yet lol.

Regarding the project... I'll be trying 3 more interesting and VERY useful romhacks in the near future, I'll keep you guys informed

~Sky

 

[Swing]

haha tight ok I'm prolly just gunna switch back to my American account anyways. I doubt that the latest VITA hack is worth it unless they add stuff to it in the future that'll grant it access to CFW... thad suck...

 

[blackfire219]

I would strongly suggest not going overboard. We've seen projects flatline because they were too picky about how the final product turned out.

90% of the people who are going to end up playing this just want it for that: to _play_ it. They're not looking for perfection. Considering Square up and abandoned them, having _something_ is better than what's available right now.

And consider the history of official translations has always been imperfect. Look at the older FF games of the SNES era. FF Tactics. Famously botched, but still were published because they were good enough.

Don't worry so much about subtitles over videos and discerning the writers intent in a language that doesn't directly translate. It should be readable and let people fill in the gaps.

I also agree with that..the untranslated video text in that part is not necessary to translate..the important thing to translate is the main scenario..

 

[SkyBladeCloud]

And speaking about the main scenario... This week's report by Adam:

10/13/13 week ending report on Sunday (United States time)

This week's top translators:

1.Hakurou with 531 (with line breaks) approx 372 (actual Japanese lines)
2.HHTurtle with 386 (with line breaks) approx 266 (actual Japanese lines)
3.Adam with 359 (with line breaks) approx 251 (actual Japanese lines)
4.Corey with 229 (with line breaks) approx 160 (actual Japanese lines)

Please also note HHTurtles contributions to the terms bible page.

Main Sceneario text
Approx lines of Japanese left: 24607 of 34549
Completed this week: 3%
Total Completed: 29%

At this weeks rate translation could be done in: 6 months

...

 

[coredecepts]

There's at least one video that has title cards (a screen with just text on it explaining a part of the story), so that's one instance where we definitely need to do some video editing.

 

[Lyan53]

Simple process:

-I extract AVI's from PMF's (no quality loss)
-A translator gives us translation for those text (they are just a few so...)
-The editor edits them (here is where I'm concerned about quality loss (?))
-I recode them back to PMF
-Inserting the PMF back to the game with the translation tools make those text to appear in English "

Also, if I may ask, how would the edited video play in terms of effects over the original text?

I'm not an expert in video encoding, but for The 3rd Birthday's hack (in french), I had the opportunity to work on PMF videos.
My observation was the following, the quality loss is inevitable... :/ (but I could be wrong)


If I take a few steps of your process

-I extract AVI's from PMF's (no quality loss)

--> exactly, if you extract PMF in uncompressed AVI

-The editor edits them (here is where I'm concerned about quality loss (?))

--> You can export directly your frames in BMP for edit them

Not replace texts, add just a subtitle on each BMP with exactly the same x/y pos (if the text position is fixed on video), there should be no loss of quality on BMP with this method

-I recode them back to PMF

--> I think the problem of quality loss is here, I don't know any method to encode PMF without quality loss


I tested with the 3rd Birthday PMF Intro, I used "UMD Stream Composer" (Sony UMD Tool) with the best encoding method I know, and result:

Zoom 200%: Original --> Test --> Test (with detailled quality loss)





Loss quality:
-Higher contrast
-Some details are no longer visible
-The gradient effect is stronger
-The weight of the PMF increases by about 145% (original: 41.5mb / test: 59.8mb)

Here's link of PMF "original + test" for see result: http://www.mediafire.com/download/kf7m2iuw0q3bq36/3rd_Birthday_PMF_test.rar


I don't know any other method for obtain best results, but if you know one, so I'm interested

 

[SkyBladeCloud]

Thanks Lyan for the BMP edition tip. I thought about a small effect to cover the Japanese text, putting English words, but the BMP idea is pretty good too.
Now about the recoding quality loss, the UMD Stream composer method was the same I used to remux videos when I merged both UMD's. However, the quality loss wasn't that big as I see in the 3rd birthday. and I got no contrast change. In fact, the loss itself is unnoticeable except when going full-screen.

I'm currently working implementing a ASM compression algorithm in the game, so I can put original non-remuxed video files in the merged ISO. IDK if it's gonna work (it might need more memory than available for decompression), but it would be useless if we loose quality after the edition.

...

EDIT....

Huh--- PMF's are actually MPEG2 encoded video streams... So in theory, we could *simply* extract the stream blocks containing the text to edit... Do the edition, re encode it, and then insert the edited blocks where they belong in the original stream. This should avoid quality loss, except in the edited blocks :/....

Sounds difficult, huh? Well, adding compression to the game is even harder, so if it works... Why not edit individual PMF stream blocks? xD

~Sky

 

[Skies of Arcadia]

Ah If proofreading entails rewriting translated script so that it makes sense in the game, then I'd be very willing to help Core. If that's ok with you guys I mean.

 

[Ziggi_Strife]

Hmmm 6 months. the suspense is killing me but il manage xD Its good to see that we have 7 active translators but i think we need a magic 10. But there could be managment problems then just like Sky said.

 

[MrShyCity]

Hmmm 6 months. the suspense is killing me but il manage xD Its good to see that we have 7 active translators but i think we need a magic 10. But there could be managment problems then just like Sky said.

7 translator is good amount
10 that's a lot of people to work with + working with that much of people would be pain in ass
I still see where your getting at

 

[coredecepts]

There's a fair amount of ramp-up involved in new people joining the team. Learning to use the various tracking and record-keeping documents we have, learning to use our source control system, and just getting used to working with each other.

We're at a good place right now, and it all feels very manageable. We don't feel the need to add more members to the team at this time, but maybe we'll feel differently about it in a few weeks, and we can reevaluate then.

 

[blackfire219]

do u also have all of the other leaked UMD tools of Sony aside from UMD stream composer?

 

[SkyBladeCloud]

Yeah, but they won't help us keep the video quality :/

Btw, status report on the new romhack:

-Yes, I'm still implementing ASM compression in the game. The bottom line is to be able to use original untouched video files in the merged ISO, compressing some of the game's files using the LZO algorithm.

-Said algorithm is one of the (if not THE) fastest decompression algorithms ever. And uses almost no memory, so game performance won't change (even if it changed, it would still be faster than playing the game from UMD). It is a loosless data compression algo.

-Due to the PSP's RAM page size, I compress only files that I can decompress with 0 bytes of memory usage, using overlapped decompression (decompressing to the same buffer where the compressed data is stored).

-The initial study shows that, while the LZO compression ratio is not the best (having a good decompression speed has that downside), it is enough to hold all untouched game files in less than 2GB.

-It will take tame to finish implementing the hack, specially because I have to rewrite part of the ISO merging hack (in order to read a file, the game needs its ISO sector, and size, now it will also need the compressed size). But since the translation itself is gonna take even more time, this wont stall the project at all.

-Fun fact: LZO-compressing the game's files, AND remuxing it's videos at the same times, makes the whole game take 1.7 GB. So yeah, Type-0 could have been released in one single UMD. The only difference would be when playing the game in a TV (and even in fullscreen the video quality loss is minimal, no one would have complained lol).

~Sky

 

[toroal]

Every day, I turn on the PC and open that page, just to see news or something about this project, I really want to make something to help, but I don't know what.
So I will continue to encourage the team xD
Guys fight on!

 

[coredecepts]

Everyone who wants to help are welcome to spread the word about the project once we finish. We'll likely be too tired to campaign!