Zestiria: A Good (?) Release

Happy Holidays and Happy New Year, everyone! This post has been long delayed due to lack of time and lack of playing Zestiria enough, but I guess I might as well put it out there. Here are my impressions about and critique of the Zestiria release.

Good

• Good dialogue that avoids straying into juvenile rewrite territory like Hearts R.
• Great English voice acting.
• Dual audio, in case you’re not a fan of dubs.
• Sick guitar track to cover for the vocals that they couldn’t license (?) in the opening video.
• Artes are mostly tastefully and arteistically named.
• (Steam version) Great port performance. Like Valkyria Chronicles, you can easily run this on any recent Intel integrated GPU at a fairly smooth 720p/30fps or above. (Though, supposedly, the frame limiter is not so great…)
• (Steam version) Great adaptation of the battle controls to mouse and keyboard. It took some getting used to, but whoever came up with this knew what they were doing. (Disclaimer: I’m still using a controller for the most part.)

Bad

• The font is all sorts of borked. The period looks like a comma and doesn’t sit well on the baseline, the kerni ng is wrong, etc.
• Locked at 30 fps. Guys, this is almost 2016. Unless you’re programming for a handheld or something, don’t tie your game logic to the framerate. If you were planning on a potential PC or PS4 release down the road when you were developing Zestiria, why would you do this? Hell, Abyss on PS2 had 60 fps battles… They released a Berseria trailer at 60 fps, so it seems they’ve (hopefully) already learned their lesson.
• Sharing is locked down. The Steam version disables Steam broadcasting, but the PS4 version goes a step further and disables screenshots, too. I’ve also heard that people using third-party hardware/software to stream have had their videos taken down, which seems a bit heavy-handed, though that doesn’t seem to be an issue anymore.
• The Bink2 video encoding they used is of inexcusably poor quality. Either Bink2 sucks or they have no idea how to use it. Why don’t they just use whatever Netflix, Youtube, etc. use? The bitrate is stupendously high for what looks like okayish 720p source video, to the tune of 2.2 MB/s (!). To put this into perspective, imagine if you used 3 GB of bandwidth every time you streamed a ~25 minute episode of a TV show or anime at 720p. Just no. And they don’t seem to use decent upscaling so it doesn’t even look too great in-game. I’m pretty sure even the PS3 should be able to do better than this, let alone PS4 or PC.
• The few Artes that are bad are hilariously terrible (e.g. Final Player).
• Such meemees
• (Steam version) Default non-battle keyboard keybinds are terrible and several cannot be changed.
• (Steam version) Mouse support for the menus is also present but rather janky; it feels like you’re trying to emulate a controller with a mouse rather than simply using a mouse to perform a task for which it has been popular for at least 30 years (navigate a GUI). It’s better now after a few patches than it was at release, but it’s still not totally intuitive. I have to give them props for including it, though. If only this were the greatest of this release’s problems…

Ugly

Text Presentation Failure Due to Lack of Programming Support

As usual, the localization received insufficient programming support. This is evident from how poorly the text is presented in the menus and popup dialogs. There are random, inappropriate linebreaks everywhere, descriptions spilling a single word onto the second page despite not using anywhere near the entire first page, and so on. I am going to guess this is because the localization team must guess randomly at where the linebreaks need to go and don’t have time or a QA guy to go back in once the text is running on the test build to see if all the lines are formatted appropriately.

As I’ve said before, this is a REALLY STUPID way of doing things. When you type text on your computer, do you usually hit Enter to prevent the text from going off the right side of the page? No, right? You just type and let the computer handle it. It’s called “word wrap”, and it’s been around for at least 35 years. This is exactly what the menu programming team should be implementing into Tales menus from the very start. It’s not that hard. I’ve hacked it in for Hearts. Other hobbyist hackers have done the same for other fantranslation projects. Supposedly, Earthbound implemented it on the SNES back in 1995! And you have the source code, so you don’t have to splice it in. Stop creating a huge amount of unnecessary work for every single one of your localization teams (remember, there are several target languages, not just English) and just program this into the game already!

Translation Failure Due to Lack of Developer Support

The editing is mostly good, but the base translation isn’t. Again, I’m not sure why companies don’t seem to get this. You need to pay (in terms of time and/or money) if you want a decent translation, and no amount of editing is going to fix a poor base translation. It’s like getting a famous chef to plate and garnish overcooked food or adding fancy marble tiling to a shoddily-constructed house. Other people playing with the Japanese audio can easily notice when you make huge translation errors that result in the opposite of the intended meaning. Personally, I wanted to play with the English audio because I like the English voice acting and that’s how I’ve played all previous official releases. But even there I could easily notice issues.

You might recall that I pointed out that various in-battle exchanges didn’t seem to make any sense in the demo, right? Well, about 40 minutes into the game, as Sorey and Alisha are leaving the ruins, they have a conversation that made me scratch my head and go “What??”. So, of course, I referenced the Japanese version, and it was clear that the translator completely missed the entire point of the scene. It was supposed to be about Sorey being nervous about his being obsessed with what he just heard might be a children’s book, then being relieved when he finds out that human society at large doesn’t think of it as a children’s book after all. I really don’t think this was difficult to get… at least, if you were listening to the Japanese audio, or translating the story events in chronological order, or if you had played the game before. Despite that, the end result is a conversation that is somewhat non-sequitur and contains filler lines to mask that they didn’t know what the conversation was supposed to be about.

In this case, I don’t think this is because the translator is bad at translating. I think the issue is that the translator is expected to rush-translate barebones Japanese text in a spreadsheet without any context. When you want a book translated, do you copy and paste individual lines, out of order, into a spreadsheet and hand it off to some subcontractor who doesn’t know how the lines fit into the book as a whole? Is that how you translate movies? No, typically, if you want a work to be translated well, you give it to the translator so they can experience it for themselves, and then let them translate their experience into an experience the target audience can understand. Give the translator the associated audio and/or the scenario writer’s script or notes! Bamco, please, develop a good translation workflow environment that you can more-or-less reuse across multiple future projects. Give your translators all the tools they need to be successful. Otherwise, you just plain aren’t going to get a quality end product. It’s basically impossible.

Game Mechanics Communication Failure Due to Lack of Quality Assurance

There was an apparent lack of QA, which I would have imagined would have caught many of the above issues before release, plus a few glaringly obvious ones I didn’t mention above that had to be patched out of the Steam version soon after release (oops keybinds and configuration changes don’t work right OPPPPSSSS).

I think my favorite example is still RESTRAIN!, which is this game’s iteration of Failed English-to-English Translation™. I discussed last time how I couldn’t even understand how it could be possible to come up with such a baffling, inappropriate “translation”. As it turns out, there’s even a tutorial entry that describes exactly what it is supposed to mean, so there’s no excuse. Can someone explain to me how any QA tester could see RESTRAIN! pop up on their screen or see that tutorial text box and not ask, “Hey, why does it say ‘RESTRAIN!’? What is that even supposed to mean, and does it clearly communicate to the player the information the game is actually trying to convey?”

Other issues include an obvious lack of understanding of the battle system by the translators, editors, and/or QA team members. Tying into the above major points, this could have been avoided if the translator actually had a passing understanding of what it is he’s translating from the outset (by having played the game). But I guess in the real world you can’t necessarily expect that of a translator, especially not a subcontracted one. Shouldn’t an internal Bamco team then be on hand to fix any translation errors made as a result of not understanding battle mechanics? You don’t even need to go back to the original Japanese to understand and fix these issues.

For example, lots of tutorial text appears to be utterly confused whenever it talks about hits, combos, and chains. It talks about, say, pressing a button after “3 combos” or, worse, “3 hits”. This is wrong; they mean after the third action in your chain — when you have taken three actions in a row (e.g., front step, Hidden Arte, Hidden Arte) without pausing and returning to your default ready stance. Not after “3 combos” (what, three separate combos?), a “3-hit combo” (Trihawk hits three times by itself…), “3 hits” (ditto), or anything else. You would think it would be simple enough to figure out that you should use the word “Chain” when the ability is called “Chain Blast”, but I guess not.

Another great example is this gem that I encountered fairly early on regarding Seraphic Artes. There are all sorts of issues here, with weird as hell linebreaking (why the super long 3rd line and super short 4th line?), weird use of articles/generic plurals (why “an enemy”?), completely opaque clauses whose meaning and relevance were probably unclear even to the translator/editor (“needs charging”? What does that mean? And so what? Why should we care?), and coming to an almost entirely wrong conclusion (it doesn’t “halve casting speed”, which would be a BAD thing; they probably meant to write that it halves casting time… except that would ALSO be inaccurate).

All these issues should have been plainly obvious to anyone who actually took a bit of time to understand the battle system and check that the tutorial text and Battle Action descriptions accurately describe how the game works. The only conclusion to be drawn is that no one bothered to do that. Not cool.

—

Despite my ranting above, it looks pretty good and enjoyable overall! Buy it if you have time and money and are looking for a decent JRPG to play. I just wish the industry were more concerned as a whole with having the trifecta: reasonably accurate translations, reasonably well-localized scripts, and reasonably well-presented text.

I’ll comment on two more things before bidding you enjoy the new year. One is that if you have the Steam version, you should really grab the awesome unofficial fix patch by Kaldaien and DrDaxxy. This game really deserves 60 fps. In case you’re curious, I have a laptop with a GTX 660M and these are my settings. The cutscenes get physics glitches and visual garbage at 60 fps, so I disabled that, and the resolution and quality options are purposely turned down to keep the fps stable, which is super important since the game logic is still tied to the framerate. Even if you don’t have the hardware to run 60 fps, you should still grab it, since according to the author it fixes stuttering and unnecessary frame drops.

The second is that, as I predicted, Japanese players e-lynched poor Baba-san for basically no reason. Now, I’m not that far into the game (just got Windstepping), but I’m pretty sure basically none of the complaints I’ve heard were levied against the game regarding her treatment at the hands of the party have any merit. Maybe they’ll give her the Flynn treatment in Tales of Zestiria R, or maybe the Japanese fanbase shot themselves in the foot by preventing future Zestiria content set in what seems to be an interesting world with interesting characters from coming out. Other than Berseria, anyway.

Well, that’s all for now. Hopefully Stuff will Happen in 2016!

(Thanks to HystericalGameZ and Shirrako for the video links!)