How do you like your fansub typesets?

[Nikolai]

There are a few typesetting trends that I've noticed, but it's rarely discussed how effective they are. Being a typesetter, I'm wondering what styles people like, and I figure this would be a good forum to ask around about which way is the most preferred.

Among the following sets of options, which would you prefer? (of course, "it doesn't matter" is an option):

A. Romaji karaoke on top, song translation on bottom
B. Romaji karaoke on bottom, song translation on top

A. Japanese karaoke text on top or bottom next to the romaji
B. Japanese karaoke text on either side
C. No Japanese text

A. Opening credits translated/names transcribed
B. Not important

A. Different font colors for each character
B. Different font colors for on-screen/off-screen/thoughts
C. Same font color all around

Edited: Forget the following question; everyone likes A.

[And a timing question:

A. Timings that start when the voice actor starts speaking
B. Timings that start when the character opens his/her mouth]

[wsheit]

a,a,a,c,a

I hope that helps

[StarCreator]

I don't see why you didn't include a "no karaoke" in that list.

I personally have no preference as long as it doesn't look overdone or too showy. (Translating credits is a nice idea, but there's no way to really do it without completely blocking either the animation or the text, and there's also a margin of error in translating names when you don't have the production studio's assistance - thus, it's better left to the professionals to do that kind of work.)

[complich8]

I'd vote a, a or b - depends, b with conditions, b, a.

reasons:
first one: I'm used to reading my translations on the bottom, and not reading japanese at all. karaoke the japanese syllables somehow, if you so desire.

second one: top to bottom looks stylish, but can become obtrusive. side to side is less obtrusive but also less stylish. I think it depends where the credits and the video are, what's already visible and obscured, etc. I like to see the kanji put in (and karaoked too) more as a mark of dedication to doing the job all the way than anything else. (Maybe just an indicator of mad fanaticism, I dunno ).

third one: full translation of everything is excessive as a rule -- who cares who the key grip was, or whatever. Still, fanaticism isn't a bad thing.... At very least, if there's any song info at all, I want to see it translated as appropriate and nicely styled to fit in. If there's no song infos, don't just make it up, of course . If your translator can't translate it, too.

fourth one: sometimes it's hard to tell if someone's thinking or speaking, or hard to tell who's talking if there's two similar sounding voice actors (or more likely similar sounding voice actresses). Especially if you're busy reading subs and can't watch as much of the animation.

last one: yeah, time to the audio, scene time to the video. a lot of animation doesn't match up, but if someone starts moving their mouth and they're not saying anything, you're not gonna sub "flap flap flap" for their lip or anything .... ya know? You're subbing what they are actually saying.

that's my opinion on it =D

[exedore]

A. Romaji karaoke on top, song translation on bottom (though I am quite happy with no Karaoke)

C. No Japanese text (It adds clutter and some groups add it even when it's already there)

A. Opening credits translated/names transcribed (Props to all the groups who do this)

C. Same font color all around (If there's a Narrator I tend to give them an alternate colour, but otherwise same unless it's two people at once)

And a timing question:

A. Timings that start when the voice actor starts speaking (Time by wav file = start speaking...of course with old animation, they don't synch quite as well.)

[getfresh]

I vote for "d" ;p

d - typeset and timing are created formatted based on the sources style/theme.


-=Full Explanation=-
Styles vary from anime to anime, leaving some looking better plan and some looking better with FX that do more then just copy onscreen j-credits, then use a diff FX then the j-credit but it all seems to fit together theme wise. On the other hand this doesn't work at all with some other animes. The same applies for audio. Every anime with a diff postproduction team, recording engineer, mastering process, ADR engineer(automated dialogue replacement), and gear used to do the audio editing+sound design.

-=Cliff Notes=-
Treating all anime the same and creating subs with the same overall base format would be like if every peice of art had the same review no matter the content.

And I'll agree with exedore's statement for the most part that timing or dialogue isn't done based on characters mouth as queue very often at all anymore. This can only be done if the audio is at exact to near exact sync with characters mouth. But some of us(not many) do scene change over-run removal, which does require timing dialogue based on visual queues.


crap... I went off on a rant again. Oh well said my piece. Back to plaing FFX-2

[Somedude]

Personally, concerning karaokes, I'd prefer to not see a romaji karaoke and just include a Japanese karaoke, with maybe hiragana over the kanji to indicate the syllables. Reason being, is romaji is too non-standard, and I usually find myself just reading along with the Japanese characters(Or trying to. ). Plus, anyone who isn't able to read along with the Japanese characters fares little better with the romaji and is probably going mostly on memorization anyways.

I'd prefer that the translation of the names be taken care of by the studios.

I'd prefer that the fonts stay the same for dialogue, and perhaps change for narrative or secondary speech when one dialogue overlaps another.

I like it when the subs start and end with a characters' line of dialogue. However, if the line is really drawn out, I prefer that the subs reflect this however the timer/translator/editor decide to do it.

Now, the most important thing... ALL fonts need to be readable from 10 feet away on my 25 inch TV. This is a personal preference, but I find that if I can't view it like this, than I am likely going to be squinting even on the computer monitor.

[7thMethuselah]

Originally Posted by Nikolai

There are a few typesetting trends that I've noticed, but it's rarely discussed how effective they are. Being a typesetter, I'm wondering what styles people like, and I figure this would be a good forum to ask around about which way is the most preferred.

Among the following sets of options, which would you prefer? (of course, "it doesn't matter" is an option):

A. Romaji karaoke on top, song translation on bottom
B. Romaji karaoke on bottom, song translation on top

A. Japanese karaoke text on top or bottom next to the romaji
B. Japanese karaoke text on either side
C. No Japanese text

A. Opening credits translated/names transcribed
B. Not important

A. Different font colors for each character
B. Different font colors for on-screen/off-screen/thoughts
C. Same font color all around

And a timing question:

A. Timings that start when the voice actor starts speaking
B. Timings that start when the character opens his/her mouth

About the karaoke (i love it )

a a b c (i HATE different font colors all the time )

about timing : i have no idea what you're talking about so i don't care

[exedore]

Originally Posted by Somedude

Now, the most important thing... ALL fonts need to be readable from 10 feet away on my 25 inch TV. This is a personal preference, but I find that if I can't view it like this, than I am likely going to be squinting even on the computer monitor.

YES. Big, readable subs are a must. While you don't want text eating up half the screen, it's not hard to walk to the other end of your room and see if they're readable on your monitor. If they're not, you need to make the subs bigger. Even at clubs, where things are on a huge projection screen, small subs are a pain to read as they don't scale well (small lines = lots of details = lost in compression). If you want something to proportion against, dig out your old VHS subs (or download the Central Anime encodes of Galactic Heroes) and keep those in mind.

Likewise, please don't put your subs right at the top and bottom of the screen (I haven't downloaded anything that you've done specifically, IIRC, so this is generic). I know a number of people who like to convert their digis to VCD, as well as people with home theatre PCs who hook things up to their TV, plus there's clubs and conventions showing these things. The problem is, of course, that TVs (in the NTSC part of the world) have this thing called Overscan, where a border around the picture is hidden so that you don't see various things that aren't supposed to be seen (I can't get more technical, read Video Demystified for details).

The short version of the prior pargraph is: Most TVs (and some projectors, including ones used at cons) cut off about 5-10% of the picture. If you put your subs way at the edge, they're going to get cut off. The easiest way to fix this is to go into SubStation and tell it that you want at least a 30 PIXEL VERTICAL MARGIN, but a 40 or 45 pixel vertical margin is better.

[K_R]

If your home theatre setup involves a computer, you should be able to either set the amount of overscan or use a player that can zoom in and out (eg. zoomplayer). If you're converting to VCD or some other format, you can add borders.

[scotty81]

A,
A or B ( I like it on the side or under the romaji ),
A ( definitly, you have to say who you are so theres 's no abuse of your work ),
C ( it's easier to follow, you could get mixed up in the colors, I like yellow subs, easy to read even when you're not close to the screen )

A ( I don't uderstand japanese , subs are just a help to understand, you can understand most of it by the way it's said )

[GipFace]

For the first and last questions, A. The rest are subjective.

Um ... for font selection, stay away from Comic Sans MS for anything, please. The font MarkingPen is also beginning to be used a lot, which means it'll probably be the next font "not to use".

exedore writes:
I know a number of people who like to convert their [digisubs] to VCD, as well as people with home theatre PCs who hook things up to their TV ... The problem is, of course, that TVs ... have this thing called Overscan, where a border around the picture is hidden so that you don't see various things that aren't supposed to be seen

VCD Solution:
1. Use tmpgenc's project wizard to start a VCD conversion.
2. Under the filter setting menu, check the box marked "clip frame".
3. Hit the button called "arrange setting".
4. Select "center (custom size)" under the pulldown menu.
5. Change the size to some lame messed-up resolution so you can actually see the overscan (i.e. 318x216)
6. Proceed to re-encode and get yourself a cup of hot chocolate.
7. Buy a TV-out card so you won't have to do this again.

For computer -> TV, there shouldnt be any overscan at all, otherwise you wouldn't be able to see the start button on most people's desktop configurations. I don't see anything clipped on my Radeon's TV-Out.

[Badaro]

A or B - Either way is fine for me.
A or C - I don't like text on the side.
B - It's a nice thought, but tends to get in the way of the video.
C - IMO, different fonts/colors should only be used for additional information/translation notes.
A - I don't pay attention the the character's mouths, I pay attention to the sound.

[]s Badaro

[exedore]

Originally Posted by GipFace

For the first and last questions, A. The rest are subjective.

Um ... for font selection, stay away from Comic Sans MS for anything, please. The font MarkingPen is also beginning to be used a lot, which means it'll probably be the next font "not to use".

For computer -> TV, there shouldnt be any overscan at all, otherwise you wouldn't be able to see the start button on most people's desktop configurations. I don't see anything clipped on my Radeon's TV-Out.

If you're doing regular TV out from the graphics card then yes, you have some control. However, if you're using something like a Sigma Designs Netstream (or their lower end model whose name I've forgotten), there is no overscan control. The same goes for the increasing number of people buying MPEG-4 enabled standalones.

I like simple sans-serif fonts myself. They're easy to read and look clean.

[Wormlord1442]

AABBA

Originally Posted by Somedude

Plus, anyone who isn't able to read along with the Japanese characters fares little better with the romaji.

That makes no absolutely no sense. There's people who doesn't want to learn japasese writing but still want to sing along. And for people that has just started learning the kanas, it may help them to memorize them.

[Nikolai]

Originally Posted by StarCreator

I don't see why you didn't include a "no karaoke" in that list.

Well, because I like karaoke and I'm going to be watching my own.

Originally Posted by Somedude

Now, the most important thing... ALL fonts need to be readable from 10 feet away on my 25 inch TV.

Originally Posted by exedore

Likewise, please don't put your subs right at the top and bottom of the screen

Yep, I both watch anime on TV and encode SVCDs, so I always check my typesets to make sure they're readable and on-screen (when possible; Kindaichi's karaoke is lower than usual because of the Japanese karaoke that it needs to make room for).

Originally Posted by GipFace

Um ... for font selection, stay away from Comic Sans MS for anything, please.

Eww...Comic Sans. I'm staying away from it.

There seems to be a clear preference for karaoke on the top and translations on the bottom, which is surprising. I know most groups do it that way, but what's the reason? I myself prefer it the opposite way, because looking down at subtitles is more natural, and reading romaji to sing along requires more concentration than English, so I can pay attention to the karaoke and the opening video simultaneously more easily.

[ChibiDusk]

A. Romaji karaoke on top, song translation on bottom

A. Japanese karaoke text on top next to the romaji

C. Doesn't matter

A. Different font colors for each character [should not vary TOO much] [bottle fairy is good example]
B. Different font colors for on-screen/off-screen/thoughts

A. Timings that start when the voice actor starts speaking

[Somedude]

Originally Posted by Wormlord1442

AABBA




That makes no absolutely no sense. There's people who doesn't want to learn japasese writing but still want to sing along. And for people that has just started learning the kanas, it may help them to memorize them.

Well, the problem is, that romaji isn't exactly a good indication of pronunciation. I contend that most people that want to sing along are likely to just memorize the sounds rather than try to read along with romaji. Plus, I never said it was a popular opinion, just that ideally, that is the way I'd prefer to see things.

[LordBrian]

Romaji, when furigana is absent, is an excellent way to help learn kanji. That's mainly what I use romaji in karaoke for, hoping the entire time that the subbers didn't make any important mistakes, thus screwing up my lessons. ;)

[Wormlord1442]

Originally Posted by Somedude

Well, the problem is, that romaji isn't exactly a good indication of pronunciation.

Can you show me an example of why it isn't? and can you think of a better indication?

Originally Posted by Somedude

I contend that most people that want to sing along are likely to just memorize the sounds rather than try to read along with romaji.

I disagree, except the cases where the text appears just as it's being sung (such as A-Kraze's Last Exile karaoke).

Originally Posted by Somedude

Plus, I never said it was a popular opinion

I didn't complain about your preference, but with the "fares little better" part, that seems to imply that people who aren't learning japanese can't read romaji either.

Originally Posted by Nikolai

There seems to be a clear preference for karaoke on the top and translations on the bottom, which is surprising. I know most groups do it that way, but what's the reason?

Maybe it's because this way consistently concentrates all translations on the bottom. Or just that we're used to it.