"Your Diary" is coming out

[sanahtlig]

Experience makes you more efficient, but not necessarily better.  Think of quality and speed as opposing ends of a spectrum, and efficiency as a bonus that you can allocate to either one or both.  If speed is prioritized, as it often is, then the benefits of higher efficiency could end up entirely "squandered".  Speed could even be so highly prioritized that quality is actually reduced vs. an inefficient worker who takes his time to deliver a quality result.  It all comes down to the incentive structure in such cases.

Edited August 11, 2018 by sanahtlig

[Decay]

17 minutes ago, sanahtlig said:

Experience makes you more efficient, but not necessarily better.  Think of quality and speed as opposing ends of a spectrum, and efficiency as a bonus that you can allocate to either one or both.  If speed is prioritized, as it often is, then the benefits of higher efficiency could end up entirely "squandered".  Speed could even be so highly prioritized that quality is actually reduced vs. an inefficient worker who takes his time to deliver a quality result.  It all comes down to the incentive structure in such cases.

Experience can absolutely make you better, though. Proof: me. It all depends on the attitude you approach your work/hobby with. In fact, many translators I've talked to have said that they've become slower with experience, not faster. Inexperienced translators tend to work quickly, going with very direct and literal translations, and basically just translating word-by-word without thinking deeply about it. More experienced translators will tend to read into nuance, will consider context more, and will spend more time finding ways to eloquently write what is being expressed. Obviously there are outliers. Some translators will never change their translation style with experience. But that's not everyone.

The same goes for editors. When I first started Dracu-Riot, I was really sloppy. I improved rapidly as I worked though and ended up having to redo a lot of my early work. I was brand new to editing back then which accounts for the rapid pace of improvement, but I don't think I'm going to stop improving ever.

[Dergonu]

1 hour ago, Decay said:

Experience can absolutely make you better, though. Proof: me. It all depends on the attitude you approach your work/hobby with. In fact, many translators I've talked to have said that they've become slower with experience, not faster. Inexperienced translators tend to work quickly, going with very direct and literal translations, and basically just translating word-by-word without thinking deeply about it. More experienced translators will tend to read into nuance, will consider context more, and will spend more time finding ways to eloquently write what is being expressed. 

This is so true. I honestly feel like my translation speed has slowed down over time since I got into official translation, as I put a lot more effort into the work, which simply means I spend more time on each line. 

[Yuuko]

Idk why people say this is machine translation when it is not. If machine translation could give this quality it would be amazing.

[Decay]

3 hours ago, Kiriririri said:

Idk why people say this is machine translation when it is not. If machine translation could give this quality it would be amazing.

The way it reads, it's either edited machine translation (edited by someone who only somewhat knows english) or manually translated by people who are about as bad. It's not particularly relevant, though. What's important is how ridiculously awful this translation is.

[Yuuko]

2 minutes ago, Decay said:

The way it reads, it's either edited machine translation (edited by someone who only somewhat knows english) or manually translated by people who are about as bad. It's not particularly relevant, though. What's important is how ridiculously awful this translation is.

What is relevant for me is that they didn't use MTL but actually paid money for someone(s) to translate it and trusted them. It doesn't matter for the customers true but it matters for me that this was not just some quick cashgrab try.

[Plk_Lesiak]

42 minutes ago, Kiriririri said:

What is relevant for me is that they didn't use MTL but actually paid money for someone(s) to translate it and trusted them. It doesn't matter for the customers true but it matters for me that this was not just some quick cashgrab try.

Even if they hired someone and got scammed it wouldn't mean it's not an edited MTL. It damn sure looks like one and if someone really gave this mess to a well-meaning company, it probably wouldn't be below them to go all Mitch about it.

Edited August 11, 2018 by Plk_Lesiak

[Yuuko]

3 minutes ago, Plk_Lesiak said:

Even if they hired someone and got scammed it wouldn't mean it's not an edited MTL. It damn sure looks like one and if someone really gave this mess to a well-meaning company, it probably wouldn't be below them to go all Mitch about it.

Just saying this doesn't look like edited MTL imo

[Veshurik]

And what's next?

 

They didn't even delay the release. They released it and beg "improve the English TL", but I don't believe it.

Hope that Chinese TL is fine at least.

Fuck that shit. Sorry. I am just so tired about that. Usually Japanese companies should understand that quality matters, they also doing their games with that slogan. But... eh... They think that TL quality is not important? Holy crap.

[sanahtlig]

I'll defer to those with experience in both fan translation and official translation.  But from personal experience (in life in general, not just translation), effort matters more than experience, and financial incentives often don't reward effort (per unit of production).  Experience certainly allows you to get more done with less effort (per unit of production), if you choose.

I'm also not convinced this is a machine translation, though 3 screenshots really aren't much to go on.  The difference matters because false accusations without any supporting evidence obstruct useful discussion and cause unnecessary polarization.  This is the difference between a deliberate scam and simple incompetence; those who pitch any bad TL as 'machine translation' are making the equivalent of a moral appeal, which is a lamentable persuasive tactic often seen in politics.

Edited August 11, 2018 by sanahtlig

[Narcosis]

Plain experience doesn't make you better. Constant practice does.

If you'll focus only on efficiency (and that's what industry requires), you will stop improving at a certain point, your main task becoming to deliver content instead. You won't have time for improvements or experimenting. Many professionals get stuck at this - they reach a certain level of proficiency and stop evolving afterwards. Becoming complacent leads to leniency. Leniency leads to routine and routine results in mistakes.

Professional TL's aren't better at all. This industry is very hermetic - the same people who used to do fan TL's in the past are currently professionals working for large publishing companies. If anything, publishers at the moment are prodding the grounds checking how far they can get away with producing as lousy releases as possible an still charge the same amount of money. It's irrelevant whether one of the translators or editors does a decent job, if the rest - especially those in charge - are a bunch of incompetent people, with sights set on the prospect of quick gain with least effort included... and it shows.

[Plk_Lesiak]

12 minutes ago, sanahtlig said:

I'm also not convinced this is a machine translation, though 3 screenshots really isn't much to go on.  The difference matters because false accusations without any supporting evidence obstruct useful discussion and cause unnecessary polarization.  This is the difference between a deliberate scam and simple incompetence; those who pitch any bad TL as 'machine translation' are making the equivalent of a moral appeal, which is a lamentable persuasive tactic often seen in politics.

1

All the screenshots I've seen gave me the Tricolor Lovestory vibe (that is MTL edited into something vaguely coherent), I might be wrong about that, as I'm not an expert, but I definitely can't really agree that's crucial in this case. I don't really care whether this is a scam or a genuine fuckup, or even whether Hobibox is responsible for this directly or were scammed by a con-artist. The effect is the same (unreadable translation) and the only possible fix is reworking the translation from the ground up. It's also not some indie team trying to make a name for themselves on the VN market, it's a huge freakin' media company and I don't give a shit about their feelings or being "fair" to them, as long as they try to sell crap like this and give half-assed responses on how they are "concerned" about the backlash.

Edited August 11, 2018 by Plk_Lesiak

[Darklord Rooke]

56 minutes ago, sanahtlig said:

I'm also not convinced this is a machine translation, though 3 screenshots really aren't much to go on. 

We'll have to convince some poor sap with loads of free time intrepid representative of the community to fork out 20 US dollaroos and do some research.

Any volunteers? Anyone at all? Come on, it looks like a really enjoyable VN ...  

[sanahtlig]

1 hour ago, Plk_Lesiak said:

All the screenshots I've seen gave me the Tricolor Lovestory vibe (that is MTL edited into something vaguely coherent), I might be wrong about that, as I'm not an expert

Here's some telltale signs of a machine translation.

1. Narration is more coherent than dialogue.  MTL does best with structured prose, whereas complex vocabulary and grammar tend to trip up human learners.  MTL's weak areas are colloquial language, character-specific language quirks, and humor.
2. (low-effort editing) Frequent and severe context errors.  Alternative definitions for a word seem to be chosen at random.  There's very little coherence.
3. (high-effort editing) Propagation of errors.  Major mistranslations occur that even those with rudimentary knowledge of JP wouldn't make, and these are propagated through surrounding text.  The mistranslations become the new context.  There may be coherence but the story itself has been seemingly rewritten.  If done well, might not be apparent to those without JP knowledge.

[Fred the Barber]

4 hours ago, sanahtlig said:

effort matters more than experience

I'll echo Decay and Derg and say that I completely disagree with this. I was pretty bad at editing when I got started. Thanks to a lot of time spent doing it, and a lot of advice from someone good at it who was looking over my shoulder for a few early months while I worked on Majo Koi (Darbury), I can confidently say I've improved massively over the two years and change that I've been doing it. (Note that I also think I still have a long way yet to go, and that I expect to improve further with more experience.)

I went back over the Majo Koi text I'd already edited some three or four times over the course of a couple years, and each time I did, I found all kinds of things that "new me" found problematic. It's been a few months now since I've last looked over it, and I probably won't look at it again, but I expect I've improved further and would make yet more changes if faced with it again.

I also have gotten faster over the years, because my brain knows both what to do and why to do it more instinctively. But the much more dramatic difference is in the quality of work I do now, moreso than the quantity of work.

Also, I don't think "effort" and "experience" are the only axes to consider here. There's another axis of "ability", which is probably the most important thing to consider. Ability often is honed through experience, and for editing and translating it's also honed through things like reading more and studying. However, no amount of effort makes up for lack of ability. A bad translator spending an hour on a line is still probably going to produce a bad translation, where a good translator could produce a good translation in a minute (while also considering more things, like the broader context, the appropriate tone, etc.). I don't think people like to hear this, generally, and maybe it comes off as elitist, but that doesn't make it any less true. In Ratatouille terms, a good translator can come from anywhere, but not everybody can be a good translator.

[SaintOfVoid]

7 hours ago, Kiriririri said:

What is relevant for me is that they didn't use MTL but actually paid money for someone(s) to translate it and trusted them. It doesn't matter for the customers true but it matters for me that this was not just some quick cashgrab try.

distinctly reminded at how mikandi burned their ks-money in a similar way

[Nier]

https://old.reddit.com/r/visualnovels/comments/967edo/your_diary_translation_quality/

[Decay]

8 hours ago, sanahtlig said:

I'll defer to those with experience in both fan translation and official translation.  But from personal experience (in life in general, not just translation), effort matters more than experience, and financial incentives often don't reward effort (per unit of production).  Experience certainly allows you to get more done with less effort (per unit of production), if you choose.

I'm also not convinced this is a machine translation, though 3 screenshots really aren't much to go on.  The difference matters because false accusations without any supporting evidence obstruct useful discussion and cause unnecessary polarization.  This is the difference between a deliberate scam and simple incompetence; those who pitch any bad TL as 'machine translation' are making the equivalent of a moral appeal, which is a lamentable persuasive tactic often seen in politics.

Does your personal experience come from another creative field like localization? At the end of the day, it is a creative process and the way people grow in it is fundamentally different from the way people grow in less creative fields. Office workers tend to grow faster over time, while artists, writers, translators, etc, tend to become better. To a certain point, at least. There are apexes to be reached in any individual's growth.

[Yuuko]

If we talk about artists they can get faster with experience. However do they want to be faster? I feel like that many artists want to challenge themselves and aim for the best quality everytime even though very few other people can even spot the difference in the end product and would be totally happy with something done a lot faster. Imo a good artist can find a balance between speed and the effort (quality) put into work.

[Darklord Rooke]

BTW, I hope everybody upgraded their PC for this game. The recommended requirements are pretty heavy

[Hetzer123]

.

Edited October 2, 2019 by Hetzer123

[Nier]

39 minutes ago, Hetzer123 said:

I am curious about it. If the Japanese developers can't distinguish the difference between good and bad translation, then how do they know which companies to license their VNs or finding skilled translators and editors.

Well THAT'S the problem and why you are seeing so many crappy "official" translations.

[Toranth]

1 hour ago, Nier said:

Well THAT'S the problem and why you are seeing so many crappy "official" translations.

What are you talking about?!  The producer's cousin's son got a 9 in 10th grade English, so obviously he's qualified to do the official translation!

 

(I'll bet some of you think that's made up...)

[sanahtlig]

7 hours ago, Hetzer123 said:

I am curious about it. If the Japanese developers can't distinguish the difference between good and bad translation, then how do they know which companies to license their VNs or finding skilled translators and editors.

Reputation, data (e.g., previous successful releases), and certification by trusted experts.  The means are there, but the will to use them and reject seemingly attractive offers is necessary.  When your company is on the verge of bankruptcy, you tend to grab at short-term solutions and dismiss longer term considerations.

[phantomJS]

Hobibox released a +18 patch, but (get this) they are charging for it:

NSFW!

I thought most +18 patches are free (at least on steam) .......