Tail Concerto undub

[rari_teh]

Tail Concerto undub 1.0a

[DOWNLOAD]

It's been long since I've wanted to play Tail Concerto, but I've never managed to last longer than Resaca Harbor because of the grating voice acting. Finally, I managed to set aside some off-time to work on an undub so that I can finally enjoy the game playtesting the patch through, killing two birds with one stone. These are the fruits of my labour.

While this undub is fully playable, I still consider it unfinished, as it doesn't cover everything it should.

I've restored all FMVs' audio tracks to the original Japanese and subbed the videos. The subs are based on the official French subs from the PAL version, since I do not speak a lick of Japanese and the French version has a reputation of being more accurate to the original than the English dub. I couldn't, however, fit the in-game cutscenes' voiced lines in, since almost all of them are much longer in Japanese and the game only plays the lines for the exact length the English ones last. Fitting them in would require analyzing the game's code and altering the lengths of each line, which is way over my league. All I could do was mute all the lines, like the French version did. The only extant remnants of the English dub are the groans, grunts and yelps that Waffle and the Black Cats let out in battle: those are stored in VAG (Very Audio Good) format, and it seems like there aren't any existing tools that allow you to do anything with those except for extracting. Because of that, I wasn't able to swap them with the Japanese sounds, nor could I mute them.

While this was a fun learning project and I don't rule out making similar projects in the future, I won't be coming back at this one since I found the game to be a tad underwhelming after I managed to play it. If you love Tail Concerto, though, I encourage you to pick up from where I left off and turn this into a fully-featured undub, or make your own from scratch. All assets and tools I used to make this patch are available on my Github, together with an in-depth explanation of my findings.

This patch has been tested through on mednafen 1.27.0 and it's bug-free. I also gave a go at my modchipped PSOne and it doesn't seem like it has issues on real hardware either, though I didn't test it there as throughly.

More details and patching instructions can be found on the release notes in the linked archive.

Enjoy! :3

[DonPatchi]

Great thank you!

[VahnFannel]

Thank you very much for your undub! I had wanted to play this game for years but could not bear the anime scenes in English. Even if the undub is "incomplete", it is enough for me to enjoy it.

Thanks, again!

[Foxero]

I want to say that this is an interesting project of yours. ^^
I'm making a Finnish Dub of this game and it's 90% done. (I just need to find a way to replace the sfx that the characters have in gameplay)

However, I read your github site of your TC project and who ever was talking about the FMV's audio quality, then you are a silly billy.
I might of read it wrong so here's a quote from there:
                                                                                                   
"The main feat I managed to achieve here was restoring full stereo audio in all FMVs: while the original "JAPANESE" (It is not a dub) was mastered in 18.9 kHz stereo audio, the American dub decided to sacrifice the second channel to double the bitrate, mastering in 37.8 kHz mono. This means that both versions use up the same number of samples (and therefore the same filesize) to encode any given length of time, but with a terrible incompatibility: sure, one could toss the Japanese FMV soundtracks on Audacity, flatten the channels to mono and output it with an artificially doubled bitrate and it all would be ready to be injected into the North American ROM, but that would incur in a very considerable loss of quality, especially when you consider that you'd also have to do it with the opening song that uses and abuses of the stereo sound with drums alternating sides for effect."

So you're saying that higher bitrate means lower quality?
NO.

Yes, the audio is in mono but it doesn't mean that the audio is now suddenly lower quality.

Also, use the PAL versions STR files, because it has Higher Quality audio. It's still in Stereo, but now in 37800khz, which is the highest quality that the PS1 supports.

If you find a way to make the in-game cutscenes dialogue longer, then it would help me out a lot, because making this dub was pretty tough due to the timing.

[rari_teh]

So you're saying that higher bitrate means lower quality?
NO.

You thoroughly misunderstood what I said. The American dub has a higher bitrate, but is mastered in mono. The Japanese dub has a lower bitrate, but is mastered in stereo. Both the bitrate and the number of channels affect the quality of the audio, the more the better. Which out of the two dubs has better quality audio is debatable – one is binaural, but has lower fidelity; the other has higher fidelity, but the sound is flattened to one single channel. To inject the low-fidelity stereo Japanese audio into the American video that's expecting a single high-fidelity mono soundtrack, one would have to either A) fiddle with the binaries to instruct it that the audio is now low-fidelity stereo, or B) turn the low-bitrate stereo track into a high-bitrate mono track. Sadly, definition is limited by the weakest link in the chain: transcoding a low-bitrate audio file into a higher bitrate does not improve its quality, the same way that opening a JPEG on Paint and saving it as PNG doesn't magically do away with the artifacting. To get a true high-bitrate mono Japanese track, I would need to have access to the lossless (or high-bitrate) masters, which naturally were never released to the public. The result of this bad transcode would be a track that's both mono and effectively low-bitrate, because every second sample of the now technically higher bitrate would be repeated, which would be, yes, considerably lower quality than before, as in practice the bitrate stayed the same and one sound channel was lost in the flattening.

[Foxero]

You thoroughly misunderstood what I said. The American dub has a higher bitrate, but is mastered in mono. The Japanese dub has a lower bitrate, but is mastered in stereo. Both the bitrate and the number of channels affect the quality of the audio, the more the better. Which out of the two dubs has better quality audio is debatable – one is binaural, but has lower fidelity; the other has higher fidelity, but the sound is flattened to one single channel. To inject the low-fidelity stereo Japanese audio into the American video that's expecting a single high-fidelity mono soundtrack, one would have to either A) fiddle with the binaries to instruct it that the audio is now low-fidelity stereo, or B) turn the low-bitrate stereo track into a high-bitrate mono track. Sadly, definition is limited by the weakest link in the chain: transcoding a low-bitrate audio file into a higher bitrate does not improve its quality, the same way that opening a JPEG on Paint and saving it as PNG doesn't magically do away with the artifacting. To get a true high-bitrate mono Japanese track, I would need to have access to the lossless (or high-bitrate) masters, which naturally were never released to the public. The result of this bad transcode would be a track that's both mono and effectively low-bitrate, because every second sample of the now technically higher bitrate would be repeated, which would be, yes, considerably lower quality than before, as in practice the bitrate stayed the same and one sound channel was lost in the flattening.

I know the differences between the US and Japanese audio.
But take a look at the the PAL version of the game.
It has higher quality audio and is in Stereo, so if it's possible, you could try to put those into the Undub.

But if you ever happen to figure out how to change the lenght of the in-game audio, when replacing it to Japanese voices, please do let me know. ^^
It's been really tough with the timing of the original English audio, that some of my audio still gets interrupted, no matter how hard I try to make my lines short.