Fansub Review Grading Scheme for the Fall 2013 season and beyond.
*Red Group Name: Prosubs
*Blue Group Name: Fansubs
Although I try my best to be fair for all groups for every series, this is ultimately impossible because translation is never black and white and involves making compromises such as text length vs detail, fidelty vs transparency, and for ambiguous lines, taking risk and try to get it perfect vs playing safe to avoid critical errors. These decisions are made on a case-by-case basis by the translator, and I will also be looking at these the same way. The translations that “click” refer to releases that made such decisions and word choices in a manner that really made sense to me.
The “technical” section of my review consists of mistakes that are practically unarguable. However, the final grades are heavily influenced by my values and therefore very subjective.
Single Episode Review:
My reviews are usually done on Episode 1. This means that I will not be going back half-way through the season and go back to see if they made right decisions in lines without context. Unless otherwise noted, I have not researched past works for sequels and remakes, due to time constraints.
In addition, final grades given cannot guarantee that rest of their releases would be of same quality.
Final grades only account for one episode, often v1 only. Once again, this is due to time constraint, but episode 1 (and the final episode) is when the group and the translator should be bringing their absolute A-game. In my opinion, the quality of Ep1v1 shows how seriously a group is taking that project. It would be disappointing to see translation quality degrade in subsequent episodes, but it’s pretty much hopeless if a team can’t even do well in the first episode.
This goes without saying, but I can only list errors that I can find in a single run through the video. I do cross-check all groups’ mistakes when reviews are done, but it’s still very possible for me to let an error slip. Some ambiguous lines are impossible to determine the meaning without asking the author, who I clearly have no access to.
Please let me know if you find any extra errors or falsely listed errors, but make sure to do sufficient research before contacting me.
“Difficulty” is based on my impression of how difficult it is to get an “A-grade” in my standards.
Higher difficulty shows tend to be more dialogue-driven, have more complex terminology, and/or wordplay. The script for these shows will be under close scrutiny and tend to result in many errors.
On the contrary, I will be more lenient for low-difficulty shows in terms of “technicals”, but a single translation mistake made for easy shows would have greater effect on their final grades than harder shows.
It should still be easier to get good grades in low-difficulty shows than shows with difficulty of 7 or higher.
9 to 10: Extremely hard, if not impossible to translate. Heavy wordplay and jargon usage, or archaic discourse. Generally, these are series containing many lines that are deliberately phrased ambiguously to foreshadow future events, or contain deep meanings.
7 to 8: Very difficult series to translate. Complex Japanese dialogue. Quirky speeches. Wordplays, and/or excessive series terminology.
5 to 6: Average late-night series. Requires fluency in both source and target to properly translate.
3 to 4: Easy to translate series, often daytime series or slice of life.
1 to 2: Children’s shows.
The accuracy of script is largely determined by errors found in “technicals” section for each series. However, I will also take into account of poor translations that are not severe enough to be listed in the “technicals section” for overall grades.
Technical review format:
[Severity]Time stamp – Type - “Original Translation” or “Accurate Translation” –> “Corrected Translation” (if original translation was the first field) — Comments (if necessary)
These are divided into 3 levels of severity:
- (**)”Major errors”: A line is translated with completely wrong meaning, or vital information meant to be conveyed is missing.
- (*)”Minor errors”: A line is translated partially wrong, but it still gets the general meaning across.
- “Negligible errors”: Strong personal preference of a word choice or interpretation over the translated line.
- (+) “Surprisingly well-translated”: A difficult line is translated unconventionally and works very well. (This will not directly affect Accuracy grades, but may be taken into consideration for close calls)
Types of Mistakes
Literal: Dictionary or word-for-word translation was chosen, which resulted in wrong meaning compared to the actual dialog.
- Literal: Look with white eyes.
- Meaning: Giving the cold looks.
> Literal translation here makes no sense in English. If you were to venture a guess, you would probably think it means “to roll eyes” translated poorly.
Context: Japanese is a high-context language, and they don’t say personal pronouns like “I” or “you” sometimes, but the listeners should know which ones are implied. Context errors occur when these are misinterpreted.
It can also occur for wrong word choices in context. Often occur when the translator did not understand the line.
Scriptwriting: Additional information intentionally inserted into the script, often to make more sense of a line or to make it flow better, but changes the intended message of the line.
Detail: Important information from the dialog is omitted for the sake of flow, nuance, laziness, or carelessness.
Nuance: When nuance for a line is so bad that it actually changes the intended meaning of the line.
Logic: Lines that contradict previously established facts or inherently illogical. Often occur from poor phrasing or misinterpretation by the translator.
Consistency: A quote, specific terminology, or critical keywords are translated differently in different lines.
Error: General mishearing or line is incorrect for unknown reasons.
Also refer to Misc. Errors in the Fan Translation Guide.
Subjective evaluation of script. In general, this is how well the script “click” in my opinion with the Japanese dialogue.
- How smoothly subtitle read
- How well lines are localized to make sense in English
- Whether lines are kept in a reasonable (readable) length.
- How well lines reflect the character’s tone
- Amount of detail retained
- Word choices
The overall grade given to each release is my overall impressions of translation quality. The final grade will be a combination of Accuracy and Script Quality.
In general, Accuracy will be the main gauge for “serious shows” such as detective, symbolism-heavy, or jargon-heavy shows. Script Quality will be weighed more heavily in lighter shows such as comedy.
The number of mistakes made will greatly affect low difficult shows (because you should not be making any mistake for those series).
A+: No mistake found. localized without losing details or nuance, and only preferential changes can be made. No one has ever achieved this score. I can’t even get this rating myself, because there will always be lines I’m not satisfied with in my own translations.
A to A-: Reserved only for the best of the best. Releases that need very few corrections and has 99.5%+ accuracy (No major errors and under 2 minor errors per episode). Script quality must be at least “Very Good”.
B- to B+: “Good”. 98.75%+ accuracy (less than 2 major and 3 minor errors combined per episode in a 400-line script). Script quality must be at least “Good”.
C- to C: “Average”. 97.5%+ accuracy (less than 10 major and minor errors combined per episode). Script quality must be at least “Average”.
D- to D+: “Watchable”. 95%+ accuracy(about less than 15 major and minor errors combined per episode). Avoid these if possible.
F- to F+: “Unusable”. Trollsubs or sheer incompetence.