Anonymous and zero-presence fansubbing

[jpwong]

Originally Posted by Uchikatsu

Sounds stereotypish, but then we all have our own opinion on things. Anonymous or not won't stop people from using the subs. Probability won't change.

Proper Team? Do tell more.

How is quality anymore questionable than the normal fansub. Groups known for quality can still have "suck" quality sometimes, no matter of previous releases. Everyone releases at their convenience....and as for reputation I won't say any names, No one actually cares about that unless they really are determined for e-penis(I won't say names )

I think they're talking about true anonymous subbing, where you just release under the generic name "Anonymous", it could be different people every week just based on whatever internet dude feels like dumping out something in the name of Anonymous. One week something could be mind blowingly spectacular, but the next week it could be a bare bones speedsub.

As opposed to Anonymous fansubbing where it's "Anonymous" because they aren't using their normal fansub group IDs, but it's the same people doing the sub every week and you can see some consistent level of quality every week.

[acidflower]

Well, from what I gathered...quite a few anonymous subs were done by the same person. But eventually, this guy will burn out and stop. But I doubt many people will throw away fansub group names. ;p it's a lot more fun thinking up a name than going under "anonymous".

But as said before, quality with "anonymous" can be really questionable. I remember trying out an episode of dtb subbed by anonymous where half the lines were left untranslated. But then again, I guess people download anon subs at their own discretion.

[Access]

Well if it's really just one guy, it's hard. In a typical group, it seems like at least 24 man-hours go into producing a quality sub for a typical show. Groups have been around since the very beginning, division of labor is key and the more eyes you have looking at something the better. I've seen quality one-man stuff, but only at a very slow pace and the outfits tend to be short-lived.

[getfresh]

At A-E we started doing "creditless" subs back when I was still active cause we felt it was pointless. Most of us cared little about our credits and felt that the people who actually MADE the anime deserved to be named. If you trans all credits and then plaster fansubber credits everywhere it makes the opening/ending a cluster-fuck that is not very enjoyable to watch. Also, no one really reads credits anyways ^_^

[Cythraul]

One might wonder about the point of tagging a file Anonymous, when, if you really wanted to be unknown, you could just leave the file untagged. Then again, Anonymous!=anonymous: Using the [Anonymous] (or [Anon] or whatever) tag means you want to be identified with a certain group of people, so it's not really that "anonymous".

[getfresh]

Originally Posted by Access

Well if it's really just one guy, it's hard. In a typical group, it seems like at least 24 man-hours go into producing a quality sub for a typical show. Groups have been around since the very beginning, division of labor is key and the more eyes you have looking at something the better. I've seen quality one-man stuff, but only at a very slow pace and the outfits tend to be short-lived.

It takes well more than 24 hours to produce a quality sub, unless it is a children's show with only 1st grade level vocabulary.

[complich8]

Originally Posted by getfresh

It takes well more than 24 hours to produce a quality sub, unless it is a children's show with only 1st grade level vocabulary.

I don't think that's necessarily true.

If your translator's fairly fluent, if your timer's reasonably fast, if your editors don't have to completely rework everything the translator says, then the script part of it can be done reasonably quickly. I'd say an "average" anime in the hands of a good team only puts maybe 8-10 hours into those 3-5 phases.

Unless you have a team of like 6-8 qc'ers and are doing a couple release candidates for every ep, or you've got other massive redundancies (translation-checker-checker? third editor?) or you've got a particularly slow typesetter making large numbers of very complex signs, the >24 man-hour estimate is pretty high.

Or did you mean >24 clock-hours? Because that's different ... a 24 hour turnaround is hard without abusing timezones, but much of that is going to be downtime waiting for people to get to it.


back on the topic,

I don't really care for complete anonymity. If it's untagged, I generally don't notice or download it. At the same time, I don't actually care who the fansubbers are ... all I care is that they're reasonably good.

With a group tag (ie: pseudonymous releases), you get the implication that the group's reputation will extend to whatever that release is. I know that my favorite groups are going to give me something that's reasonably accurate, reasonably watchable, and reasonably consistent. I don't know that about groups that aren't in that category.

It's no guarantee that the quality will be there, and I've been let down by that in the past. But honestly, I've been let down a lot more by the random newbie-groups that pop up than by the people on my list of known-good subbers.

[TheFluff]

Originally Posted by getfresh

It takes well more than 24 hours to produce a quality sub, unless it is a children's show with only 1st grade level vocabulary.

you are pretty hilariously wrong, whichever way you mean it (clock hours or man hours), and I think you know it so please stop trolling kthx

edit for content:
man hours:
TL - anywhere between 2-6 hours depending on amount of dialogue and its complexity; also speed of TL
Edit pass 1: 1-2 hours
Timing: 0.5 - 2 hours
Edit pass 2: 1-2 hours
TS - 0-infinity hours depending on how anal you want to be and how evil the show is, 0.5-1 hours in the typical case though
Encode - maybe 0.5-1 hour of actual work, can reach 6-8 hours of CPU time but that doesn't count
2x QC pass: 2x 1 hour
TL-check: 1 hour
Distro: 0.5 hours

So in the worst case scenario (with a show that's not too bad on TS) we get about 17.5 man-hours. Of course in some "quality" groups (I've been in some) this is artificially stretched towards infinity by requiring re-encodes and v2's for a single mistyped word somewhere, having edit wars, having 1643656 QC'ers who all report the same errors or amuse themselves with backseat editing (resulting in internal bitching and/or more edit wars), having people who have a turnaround time of 5 days or more, or having people who seem to spend half of their effective time masturbating. If you have a staff that actually knows what the fuck they're doing, you usually don't need to do anything twice.
As for spacing all this into 24 clock hours, well, Eclipse does it on a weekly basis and has been doing it for the last one and a half years or so. It's not even that hard since a lot of the work can be run in parallel.

[getfresh]

Heh, I'm not trolling I'm stating an opinion from experience. Yes you can do a rush job of it, but that is not quality work. And if you wish to call it "hilariously wrong" that just means you have not experienced a QC/RC thats worth a damn. The down time between a QC report being submitted and the response is part of the time. A lot of groups also have a T/L, then a T/L check. There is no way you can place times on things like trans/edit/qc/rc. And I don't consider flashy typesetting to be a vital part of a quality fansub. Quality deals with the translation, editing and timing. Go ahead and disagree all you want, but I'll still trust my past experiences over your generic time tables any day.

[TheFluff]

Originally Posted by getfresh

Heh, I'm not trolling I'm stating an opinion from experience. Yes you can do a rush job of it, but that is not quality work. (...) Go ahead and disagree all you want, but I'll still trust my past experiences over your generic time tables any day.

And what I listed are my past experiences, and I trust them more than I trust some random bitter troll who probably hasn't even fansubbed anything in the last year or two.

Originally Posted by getfresh

And if you wish to call it "hilariously wrong" that just means you have not experienced a QC/RC thats worth a damn. The down time between a QC report being submitted and the response is part of the time.

Interestingly, your definition of "a QC process worth a damn" seems to be "delay for no real reason whatsoever and spend days bogged down in stiff bureaucracy". And no, the downtime between a QC report and its answer IS NOT TIME ANYONE SPENDS WORKING ON IT AND IF YOU WEREN'T ILLITERATE YOU'D REALIZE THAT THE TABLE LISTS MAN HOURS NOT CLOCK HOURS.
I've been in about six or eight groups now; the one with the BY FAR most inefficient and annoying QC process was Ayu (which incidentally is regarded as a really good quality group), which had a very "traditional" edit/QC process. QC took ages for several reasons. The "downtime" between a QC report and its answer was usually measured in days, and every time a RC was declared it had to go through a new full QC pass with all or at least several of the QC'ers (add two to four days). RC3's or RC4's and lots of reencodes were not unusual (partly because some of the QC'ers didn't want to learn how to do proper softsub QC) which was highly annoying for the encoder (me). In one case some of the editors got into an edit war that ended with one of them quitting the group; it ended with something like RC11 and a release date delayed by about a month or more.
I completely fail to see how this kind of grinding improves quality at all; at best it makes your QC'ers burn out and go fuck fansubbing because it's not only really fucking boring for both them and the person who has to approve the reports, it doesn't really help all that much either because once the script has been edited and QC'ed once it's not going to go through any more major changes; anything left at that point should just be typos at worst and anyone who knows decent English can copy-edit a script, but I guess your definition of "better quality" is "has one less typo every second episode".

By contrast some other groups I've been in (especially ones that softsub) take a completely different approach:
First of all, the script and the encoding are entirely separated until the script is DONE. It goes through the usual TL -> edit -> time -> edit/qc -> final qc -> TL check/approval chain, but it doesn't touch the video until it's approved. THEN an RC is created and uploaded, and someone (usually the encoder) gives it a watch through and if it passes that, it's released. The encoder is also responsible for QC'ing signs since those are hardsubbed. This avoids all the enormous RC2/3/4/etc timesinks since you only very rarely need to create another RC (basically limited to encoding failures); if the encoder spots any obvious typos he can correct them immediately.

tl;dr: more QC/edit passes don't make a script better. It's much like DivX5 encoding was back in the day; anything beyond the 4th pass was pointless because it didn't add anything really. If you haven't managed to figure out a good way to express something after 5 people (TL + editor + timer + 2 QC'ers) has had a good look at it and maybe a chance to discuss it some, you're not going to find a better way even if you spend a week letting the script rot on your FTP.

Originally Posted by getfresh

There is no way you can place times on things like trans/edit/qc/rc.

And why the fuck not, exactly? Do they take place in some special division the universe that lacks the fourth dimension?
Remember I'm listing man hours, not clock hours.

Originally Posted by getfresh

And I don't consider flashy typesetting to be a vital part of a quality fansub.

why don't you go complain on the internet about it

Originally Posted by getfresh

Quality deals with the translation, editing and timing.

Yes, but SPENDING DAYS DOING ABSOLUTELY NOTHING BUT WAITING IS STILL NOT GOING TO IMPROVE IT, NO MATTER HOW MUCH YOU WISH IT.


(By the way, I've found that the more people you involve in a project, the more of the time you will end up waiting for someone to come do their thing and the slower things will be. Having more than 4-5 people involved with the script creates far more overhead and disagreements than it's worth.)

[getfresh]

I'm not the one getting nasty or catching feelings. I haven't cursed at you or insulted you. I really hate to tell you but all you are doing is flipping out and making a jackass out of yourself for no reason. Even though I haven't done anything near a troll in this discussion, I am enjoying your over reaction quite a bit.

BTW, using caps doesn't make your opinion any more valid ^^

[danayelx]

Originally Posted by getfresh

I'm not the one getting nasty or catching feelings. I haven't cursed at you or insulted you. I really hate to tell you but all you are doing is flipping out and making a jackass out of yourself for no reason. Even though I haven't done anything near a troll in this discussion, I am enjoying your over reaction quite a bit.

BTW, using caps doesn't make your opinion any more valid ^^

Fluffy just needs a nice buttplugg :P

[getfresh]

Originally Posted by complich8

I don't think that's necessarily true.

If your translator's fairly fluent, if your timer's reasonably fast, if your editors don't have to completely rework everything the translator says, then the script part of it can be done reasonably quickly. I'd say an "average" anime in the hands of a good team only puts maybe 8-10 hours into those 3-5 phases.

Sorry I didn't see you post at first since someone posted a massive one right after. You misunderstood what I meant as well but I'll explain what I meant to you since you are level headed enough to listen.

I didn't mean that the subs produced in under 24 hours or on a schedule are poor, shit, trash, sub-par, what have you. What I meant is for something to be truly "Quality" it needs as much time as it takes to get it done. It's like the different between something mass-produced and something hand crafted by an artisan(not saying fansubbers are artisans). To be Quality or above average is to place extra care and attention on the creation of. Comp, we have both wasted a whole lot of time subbing, and I'm sure you will agree a different amount of work goes into a weekly release and a "special" release, of quality, or distinct worth.

And yes, the staff matter quite a bit, but a group releasing with only 24hr of work and is known as one of the well know groups for ability, is fast because that team has been working together for some time. When a team is just formed it takes awhile for them to get in the groove and know each others habits. Even then you are reaching a level that is above the average "one hit wonder" subbers, but it is still not what I consider that groups and/or the seasoned subbers "quality" release.

Also, we all know that "shit happens". Ever encoded something and it takes a long time to encode cause of some insane kara's/filters needed/etc.. and then once done and it goes to check... Whoops theres a font/script/encoding/act of god, issue with the encode. Or everyone misses a minor error till the very final encode because it was one of those easy to miss things due to other "most noticeable" concerns at the same position, but when you toss it in distro one of the distro guys notices and you re-encode. The fact of the matter is, every delay is time working. Someone is there waiting for someone else to show up or to check something or whatever.

For a normal standard reason letting some small thing go isn't a big deal. We've all done it many times. But for a "quality" release, you check and re-check, discuss, what have you.

That is what I meant.

[complich8]

I think there's a fundamental disconnect here between what me and TheFluff are saying and what you're saying.

I'm saying that regardless of how much time and effort and care goes into it, if you have a competent and efficient team, more than a man-day of work into 23 minutes of a show is practically criminal.

Also, I don't consider encode time to be man-hours of work. It's time that the encoder can't game on his encoding machine, maybe, but ... if he doesn't have to sit there and think about the subject, it's not work time, it's just time.

I'm going to agree with TheFluff's estimates, though. They're pretty much spot-on for my experience.

I've worked with extreme perfectionist teams. I've worked with laid-back "just make it watchable" teams. I've worked with teams of newbies where I was the only person on the team with a modicum of real competency in the English language, a sense of linguistic style, an understanding of the nuances of timing and encoding, and a sense of pride in a job well-done. And even in those situations, demanding more than other people were capable of giving and having to teach them how to do their jobs to meet my expectations, I've seldom seen that much more work time on an episode. Maybe on the first two or three, but not by the middle of the series.

[getfresh]

We'll just have to agree to disagree and leave it at that ^^

[Access]

Originally Posted by complich8

I'm saying that regardless of how much time and effort and care goes into it, if you have a competent and efficient team, more than a man-day of work into 23 minutes of a show is practically criminal.

Also, I don't consider encode time to be man-hours of work. It's time that the encoder can't game on his encoding machine, maybe, but ... if he doesn't have to sit there and think about the subject, it's not work time, it's just time.

Criminal? It happens all the time in my experience. By man-hours I do mean what you said, time spent by a person actively doing work (encoding time, file transfer time, 'computer' or CPU time aren't included). If you look at 'everytime' spent on a quality project, people often overlook things, like the raw finder who spends an hour or two trying to find, download, compare raws, in search of perfection.

While it might be impossible to agree on a 'reasonable' figure as everyone has different ideas about how things 'should' be done, certainly you've heard the old saying: 'time flies when you are having fun'. When it's viewed as a hobby, you can't measure competency and efficiency in man-hours, many people will, consciously or not -- slow down when doing something they enjoy, and speed up when doing something they dislike. That's just human nature. It's like that RPG or video game that you could complete in 20 hours if you do it straight from beginning to end, or you could play it slow, get every single item and do every single sidequest, spending 70 hours or more. If it's an enjoyable game, most people will choose the latter.

The ideal case is not based on numbers, it's someone who works 'on pace' with the rest of the group. You don't want to be the only super-fast, impatient one in a slow, patient, laid-back team, nor do you want to be the one the rest of the team is always waiting on in a faster-working team.

[getfresh]

Hey access I'm not trying to play devils advocate here or anything, but this is just something neither side is going to agree on so lets just drop it.

[complich8]

Hmm ... guess I hadn't thought about rawhunter times. And really, encoder time depends a lot on how seriously the encoder's taking it -- it can be a lot of experimentation if the encoder's a bit filter-happy .

I guess you're right in that it really does depend on how you're approaching it. When I was fansubbing, it always felt better being high-output than low-output, because my enjoyment wasn't derived from the act itself as much as the completion of the act -- the sense of pride in an accomplishment. I don't think I would have been happy pushing out something I didn't put a lot of effort into, but I definitely enjoyed it more when the group produced more.

I guess that's really a fundamental difference in where the fun is. If the fun is in the act itself, in the moment, then that's great, doing it is its own reward, and by all means, take as much time as you want doing it.

If your motivations are similar to what mine were, then the act isn't all that great -- I mean, I should hope I can find some more enjoyable way to spend my time than fixing someone's grammar, you know? But because my motivation was somewhat extrinsic, when the projects died for reasons beyond my control, the motivation died with them.

Maybe different jobs in the process have different characteristic motivations though. Like, I know a lot of encoders who just love the excuse to play with new technology and finding ways to do interesting things to video, and a lot of typesetters who love nothing more than designing creative karaoke effects. But I can't think of a timer I've ever known who, after a month or two being a timer, said "damn, I can't wait to sit down behind substation alpha and sequester myself with a couple hundred lines of script and a wav file."

[getfresh]

Well I think a reason you'd see it from completion being your high point too is cause you have to stick in in the oven and just wait for the outcome which kinda sucks. People making changes in real time can have a bit more enjoyment in doing it. I know as well how timers bitch a bunch about it being the most boring of jobs, yet they can't wait to do. When I still timed actively I enjoyed it sometimes and hated it others, all depended on the anime. If they had super high pitched voices I hated it to death. like you said, its diff everywhere.

[Nicholi]

Quick means shitty job and anonymous means newbs on the job. Of course these are absolute truths. Now that we're done with that...

It seems to me, if anything, the "real" anonymous/zero-presence groups (ones that are not just large groups hiding because the show has been licensed) likely have only a handful of members and do not plan to move towards any sort of large group dynamic, and be anonymous at the same time. Rather than fall into the usual pattern of make group then constantly recruit to fansub more series to become fansub royalty. I suppose you could start growing a secret cool kids anonymous/zero-presence group. But it sounds pretty stupid, when it would be even easier to make a generic open group to flaunt your exclusivity. So most likely they are formed of just a few individuals (or maybe even just one) who probably know each other rather well and can get the job done without issues. No worries about some new guy which has it in for the old dogs in the group. I only distinguish the "real" anonymous groups from the others because the others have essentially the power of that big group backing them. Tons of extra manpower as well as distro power, which comes up later. While you usually might say "zomg they don't have 30 people qc'ing the subs, it must suck", I'd just say it only takes 1 sensible person who can read to beat 30 idiots. I find it pretty sad when a project team starts going into double digits for people just reading the subs, when you have only 1 person examining the actual video. So really numbers or even anonymity doesn't instantly give any indicators as to the quality. Just look at the output, instead of trying to assume/guesstimate what the people are doing to make the output.

So... while steering clear of the "big group" mindset, they are at the same time not forming their own little communities to propel them forward. No forum/irc room/whatnot in which all the fans (see leeches) can gather and babble about whatnot. Which is of course what we've seen since the beginning of digisubbing, little bubbles of anime communities forming everywhere. In one sense it sort of forces the fans of a show to just talk about the show rather than the group/etc because they can't get any info on em. But comparing groups will never stop, whether they be zero-presence or not. Just makes it harder to get into direct touch with them.

The only thing I think that is really holding back our friendly anonymi is that they will have a poor distro. Very unlikely that there would be dedicated seeds or any sort of bandwidth power to keep the releases up. Obviously no xdcc bots/direct downloads/ftps/etc. This is just what happens even with groups that have chosen this small group mindset. Things are released as they are finished, after that it is up to the fans to continue distributing it.

But of course the real bonus to being anonymous is you only deal with the work. No distractions unless you go looking for them. You get to reap all the benefits and can easily neglect all the bad feedback. No dealing directly with flames/trolls/other rants unless you really want to. At the height of being anonymous/zero-presence it really means your work speaks for itself. So if quality is a concern of said group it will be obvious in the output, without a site/banners/forum/etc to say "Zoomj we does the high quarlity now". Just as well it means if you want to speedsub, or sub just to sub because everyone else is a slacker you can do it. You can do whatever you want and all the stuckup hosers get to do is cry themselves to sleep. Or rather flame about you everywhere they can and waste their time . Whilst you are working on the next episode and can't wait to hear their screams of displeasure at its release :P.

Back to the real weird quality discussion h'yah. In the case of the "real" anonymous groups it is hard to say. As who can say if it is ever the same anonymous for every release. Depending on circumstances it would have to be judged case by case, which might even mean episode by episode. However, zero-presence groups make it much simpler. They actually have a name/tag, but just don't want to deal with you . Which puts them in the same basket as any other open group which springs up. But I think anonymous is mostly here to fill the holes the larger groups can't worry about at that moment.

Are there even any anonymous groups which are working on a show which doesn't overlap with some other open group?