AnimeSuki Forums

Register Forum Rules FAQ Members List Social Groups Search Today's Posts Mark Forums Read

Go Back   AnimeSuki Forum > Anime Related Topics > Fansub Groups

Notices

Reply
 
Thread Tools
Old 2009-01-06, 07:32   Link #841
getfresh
done
*Fansubber
 
 
Join Date: Nov 2003
Location: Yokosuka, JP
Age: 33
Quote:
Originally Posted by TheFluff View Post
by the way if you're going to play the I AM A VERY EXPERIENCED AND RESPECTED FANSUBBER, I SHOULD KNOW WHAT I'M TALKING ABOUT card, I should mention that on the few occasions I have actually seen audio shifted by video (usually done because the song has hardsubbed japanese lyrics that the entire kara is shifted to) it has always led to perceptible desyncs after like one episode regardless of raw source. "fixing" a/v synch is a really inaccurate business because the only way of doing it is experimentally by ear, and human aural perception is really bad at telling when something is in or out of synch.
dude argue it all day long if you like, but the fact is it is easier and more accurate to sync by a visual cue. If this had to do with encoding I'd bow to you in a heartbeat, but it has to do with timing/shifting of scripts. And you are also running yourself in circles with comments like "and human aural perception is really bad at telling when something is in or out of synch." Wouldn't that also mean that syncing the script to the audio is just as flaky?
getfresh is offline   Reply With Quote
Old 2009-01-06, 08:05   Link #842
pichu
Senior Member
*Fansubber
 
 
Join Date: Jul 2004
Quote:
Originally Posted by getfresh View Post
dude argue it all day long if you like, but the fact is it is easier and more accurate to sync by a visual cue. If this had to do with encoding I'd bow to you in a heartbeat, but it has to do with timing/shifting of scripts. And you are also running yourself in circles with comments like "and human aural perception is really bad at telling when something is in or out of synch." Wouldn't that also mean that syncing the script to the audio is just as flaky?
Yeah, I usually shift by \k timed by audio or by scene, as it takes seconds to do so. And, I'm quite accurate with shifting from audio (by listening, and I tested that my shifting accuracy has been under 6ms -- by ears + eyes and then by visual [eyes only]). That's because I'm experienced with karaoke timing.

As for shifting the karaoke effects, it becomes more painful to shift it aurally, as karaoke makers such as myself do enjoy putting arbitrary lead-in and lead-out timings to lines, so unless you know the exact starting point of the line, it becomes harder to do so. But, the best way is to compare against to the one that's timed correctly. This takes time to do so to get it as accurate as I want to get.

Regardless, in the case of myself and many others, we can shift karaoke aurally bested by under +/-2 frames for a 24fps or 30fps footage. And obviously, those who can't perform this kind of accuracies are inept, and you shouldn't even use them at all -- not to mention I had to step in and reshift the timings on a few occasions.

-----------------

Edit: if the frames per second is higher, the shifting will be far more accurate. (and I estimate about the same frames accuracy for a 60fps footage) But the best accuracy we can ever get is under +/- 2 frames. Why? Round-off errors. And, it won't make a difference the more you shift, the accuracy will still remain: +/- 2 frames (+/-1.5 frames to be precise which leads to my estimation and test of accuracy in under 6ms in my case for a 24fps footage). And, the errors don't add up, since we're not making a recursive relationship; to put it simple that one can understand, we're using the original to shift to the new one.


Dude, a lot of gamers can tell the difference at a 60fps and 120fps gaming, go read on the gamer reports.
__________________
A small tool I made, and Learn how to make fansub karaoke in After-Effects: AFXKRK

Last edited by pichu; 2009-01-06 at 08:31.
pichu is offline   Reply With Quote
Old 2009-01-06, 14:40   Link #843
TheFluff
Excessively jovial fellow
 
 
Join Date: Dec 2005
Location: ISDB-T
Age: 27
Quote:
Originally Posted by getfresh View Post
dude argue it all day long if you like, but the fact is it is easier and more accurate to sync by a visual cue. If this had to do with encoding I'd bow to you in a heartbeat, but it has to do with timing/shifting of scripts. And you are also running yourself in circles with comments like "and human aural perception is really bad at telling when something is in or out of synch." Wouldn't that also mean that syncing the script to the audio is just as flaky?

please tell me again shifting to the start of the song with this kind of precision is "flaky"

Quote:
Originally Posted by pichu View Post
Dude, a lot of gamers can tell the difference at a 60fps and 120fps gaming, go read on the gamer reports.
they can't, it's placebo, practically no monitors do more than 60 refreshes per second these days
__________________
| ffmpegsource
17:43:13 <~deculture> Also, TheFluff, you are so fucking slowpoke.jpg that people think we dropped the DVD's.
17:43:16 <~deculture> nice job, fag!

01:04:41 < Plorkyeran> it was annoying to typeset so it should be annoying to read
TheFluff is offline   Reply With Quote
Old 2009-01-07, 09:46   Link #844
pichu
Senior Member
*Fansubber
 
 
Join Date: Jul 2004
Quote:
Originally Posted by TheFluff View Post
please tell me again shifting to the start of the song with this kind of precision is "flaky"
I listen to random syllables until it's shifted correctly, and audio is more accurate than visual perceptions, but it's considerably slower to perform. I usually test the audio hearing in the middle of the line, because of the ssss-ounds. So how could that be 'flaky?'

Quote:
they can't, it's placebo, practically no monitors do more than 60 refreshes per second these days
umm pro-gamers prefer CRTs, but yes, there are LCDs that can do 90Hz. And, as you know, CRTs can allow higher than 120Hz refresh rates (probably before your times if you still want to continue arguing this basic issue).
__________________
A small tool I made, and Learn how to make fansub karaoke in After-Effects: AFXKRK
pichu is offline   Reply With Quote
Old 2009-01-07, 09:56   Link #845
TheFluff
Excessively jovial fellow
 
 
Join Date: Dec 2005
Location: ISDB-T
Age: 27
Quote:
Originally Posted by pichu View Post
I listen to random syllables until it's shifted correctly, and audio is more accurate than visual perceptions, but it's considerably slower to perform. I usually test the audio hearing in the middle of the line, because of the ssss-ounds. So how could that be 'flaky?'
it isn't, that's my point

Quote:
Originally Posted by pichu View Post
umm pro-gamers prefer CRTs, but yes, there are LCDs that can do 90Hz. And, as you know, CRTs can allow higher than 120Hz refresh rates (probably before your times if you still want to continue arguing this basic issue).
what is with the I AM OLD AND EXPERIENCED gimmick these days
I may be younger than you but I'm not THAT young, I have owned at least one CRT that could do 120Hz at certain resolutions

jfs informs me though that in a lot of quake-engine-based games the physics simulation depends on the framerate which might have something to do with it
__________________
| ffmpegsource
17:43:13 <~deculture> Also, TheFluff, you are so fucking slowpoke.jpg that people think we dropped the DVD's.
17:43:16 <~deculture> nice job, fag!

01:04:41 < Plorkyeran> it was annoying to typeset so it should be annoying to read
TheFluff is offline   Reply With Quote
Old 2009-01-07, 10:04   Link #846
pichu
Senior Member
*Fansubber
 
 
Join Date: Jul 2004
Quote:
Originally Posted by TheFluff View Post
it isn't, that's my point


what is with the I AM OLD AND EXPERIENCED gimmick these days
I may be younger than you but I'm not THAT young, I have owned at least one CRT that could do 120Hz at certain resolutions

jfs informs me though that in a lot of quake-engine-based games the physics simulation depends on the framerate which might have something to do with it
No, the precision in your graph is garbage, as it appears that the syllable (or whatever it is) doesn't fall on top of the audio/singing (an assumption). But I usually aid the shifting through hearing on a per-syllable basis to insure that I've shifted correctly. Now, that's to check and adjust for accuracies and precisions.

As for the latter, I'm not doubting your experience, but you're generalizing and making it to sound as if 60Hz is the limitation of any monitors produced these days, which is far from being correct.

Of course, as we moved into a digital world, nothing is analog anymore nowadays, and hence we get frame rates now. For example, physics simulations and any other digital processing all convert from analog into some digital forms, thereby limiting other feasible factors (i.e. time-dependent motions or the so-called frame rates).

Edit: To reply your "they can't" comment, I remember reading a statistics report a while ago that average human eyes can preceive 1/80 seconds of the object flickering. And remember that's an average or median, so pro-gamers obviously can perceive higher, as that's what they're excelled at. And, if you want to continue, we can further the research in neurological statistics of the transmission rate between eyes and brain.
__________________
A small tool I made, and Learn how to make fansub karaoke in After-Effects: AFXKRK

Last edited by pichu; 2009-01-07 at 10:28.
pichu is offline   Reply With Quote
Old 2009-01-07, 11:29   Link #847
TGEN
Far out, man!
*Fansubber
 
Join Date: Jul 2008
Age: 30
Quote:
Originally Posted by TheFluff View Post
they can't, it's placebo, practically no monitors do more than 60 refreshes per second these days
I can notice the difference between 60fps and 100fps on a 100Hz CRT. Not that I can tell individual frames apart, but I percieve the 100fps image to be smoother than 60fps; this is especially noticeable in high-paced first-person shooters when you turn and strafe a lot. On 60Hz TFTs, this isn't relevant of course.
TGEN is offline   Reply With Quote
Old 2009-01-07, 13:23   Link #848
TheFluff
Excessively jovial fellow
 
 
Join Date: Dec 2005
Location: ISDB-T
Age: 27
Quote:
Originally Posted by pichu View Post
No, the precision in your graph is garbage, as it appears that the syllable (or whatever it is) doesn't fall on top of the audio/singing (an assumption).
wayyyyyy to miss the point there brosef
the selected area doesn't mark a "syllable" or anything else for that matter (what kind of meaningful audio clip lasts 10 milliseconds?), it is only there to demonstrate the scale of the image
the point is that it is possible to exactly determine the start of the song to within +/- 10ms (ASS doesn't allow higher precision unless you mod vsfilter), so that if you have a shifting line at the start of the song you can shift it to the exact start and be assured that the karaoke/lyrics sync up perfectly

of course all of this exxxxtreme precision is kind of pointless since your karaoke timing is likely to be a lot more off than +/- 10ms per syllable but the point was that it isn't inaccurate like getfresh claims it is, in fact it is more accurate than shifting to a video cue
__________________
| ffmpegsource
17:43:13 <~deculture> Also, TheFluff, you are so fucking slowpoke.jpg that people think we dropped the DVD's.
17:43:16 <~deculture> nice job, fag!

01:04:41 < Plorkyeran> it was annoying to typeset so it should be annoying to read
TheFluff is offline   Reply With Quote
Old 2009-01-07, 18:27   Link #849
getfresh
done
*Fansubber
 
 
Join Date: Nov 2003
Location: Yokosuka, JP
Age: 33
Quote:
Originally Posted by TheFluff View Post
wayyyyyy to miss the point there brosef
the selected area doesn't mark a "syllable" or anything else for that matter (what kind of meaningful audio clip lasts 10 milliseconds?), it is only there to demonstrate the scale of the image
the point is that it is possible to exactly determine the start of the song to within +/- 10ms (ASS doesn't allow higher precision unless you mod vsfilter), so that if you have a shifting line at the start of the song you can shift it to the exact start and be assured that the karaoke/lyrics sync up perfectly

of course all of this exxxxtreme precision is kind of pointless since your karaoke timing is likely to be a lot more off than +/- 10ms per syllable but the point was that it isn't inaccurate like getfresh claims it is, in fact it is more accurate than shifting to a video cue
it is inaccurate because there isn't always a nice "silence" before the starts for starters. There is always some cue you can use as a point though, except in extremely rare cases where the op/ed change a lot or there are no cues that are drastic enough to be trusted. Such as cases where the ending fades in and out and has a solid background the whole duration. Fluff, I really hate to tell you, but when they add the audio to these shows they do it based off of the visual cues to begin with, so I'd say that they are pretty damn solid cue points. The only time shifting using audio is at all required is when you are shifting a song as an insert or stuff like that. Why the hell would I bother with that when I'm possibly going to have to shift a credits/title script as well. Also, after shifting you need to check the scene timing of the shifted script anyways. So why not just start in video mode since you end up in it.
getfresh is offline   Reply With Quote
Old 2009-01-07, 19:50   Link #850
pichu
Senior Member
*Fansubber
 
 
Join Date: Jul 2004
Quote:
Originally Posted by TheFluff View Post
wayyyyyy to miss the point there brosef
the selected area doesn't mark a "syllable" or anything else for that matter (what kind of meaningful audio clip lasts 10 milliseconds?), it is only there to demonstrate the scale of the image
the point is that it is possible to exactly determine the start of the song to within +/- 10ms (ASS doesn't allow higher precision unless you mod vsfilter), so that if you have a shifting line at the start of the song you can shift it to the exact start and be assured that the karaoke/lyrics sync up perfectly

of course all of this exxxxtreme precision is kind of pointless since your karaoke timing is likely to be a lot more off than +/- 10ms per syllable but the point was that it isn't inaccurate like getfresh claims it is, in fact it is more accurate than shifting to a video cue


What I don't understand is that why do you keep on circumventing the whole discussion around frames per seconds, frame rates, and videos--as they are irrelevant. Here are known facts to subbers:
  1. In Karaoke timing, video scenes are almost never used. (i.e., no one even bothers scene-timing) Therefore, everything is timed through hearing and audio spectrum.
  2. In Karaoke shifting, I normally approach with hearing and/or audio spectrum too, which leads to a precision of +/- 20ms difference from the original version. (1/50 is acceptable in my opinion; it's possible to achieve less than 10ms accuracies through extensive trials and errors, but read on to the third point) Of course, a quick and dirty way is to specify a known line in the original version and shift the karaoke to the correct scene, but this can be off by about +/- 1.5 frame--far less accurate than shifting by mere audio.
  3. In making karaoke effects, lead-in and lead-out effects are often used. They can be as long as 5 frames in a 24fps footage, and so having an extremely accurate time-shift (i.e. under 40ms difference) is irrelevant in most cases.
  4. You just made the whole discussion to sound as if using ears is incompetent comparing to visual only approach. It makes me wonder if you're legally deaf, as you want to avoid using ears as much as possible because of the lack of trust in your own sense. It's either that, or you don't have much timing experiences to know which is better and which isn't.
__________________
A small tool I made, and Learn how to make fansub karaoke in After-Effects: AFXKRK
pichu is offline   Reply With Quote
Old 2009-01-08, 10:07   Link #851
grunty
Senior Member
 
 
Join Date: Oct 2006
Wasn't it getfresh who shifted to video and was against shifting to audio?
grunty is offline   Reply With Quote
Old 2009-01-08, 11:54   Link #852
getfresh
done
*Fansubber
 
 
Join Date: Nov 2003
Location: Yokosuka, JP
Age: 33
Quote:
Originally Posted by grunty View Post
Wasn't it getfresh who shifted to video and was against shifting to audio?
yeah it was me... somehow pichu confused the shit out of himself along the way.

And I'm not so much against shifting by audio as I am against making the tedious shit that repeats every epi less of an annoyance. it is not required 90% so why mess with finding a cue in a waveform when it is much faster to do it with a key frame. If you are making kara, of course you need to time it by ear, but once it is timed you don't need to sync of audio at all anymore. Just the video since the audio should not be changing location for the op/ed every epi anyways as long as the capper and your encoder know what they are doing.
getfresh is offline   Reply With Quote
Old 2009-01-08, 13:32   Link #853
pichu
Senior Member
*Fansubber
 
 
Join Date: Jul 2004
HUH? How can I be confused? The most accurate way to time and shift is by audio. Period. I don't see why would you want to shift it through video, and I can do it by audio just as fast as you can do it through video. And, trust me, I only need to attempt the first time to get THE AUDIO SHIFTING better than scene-time shifting since my ears are obviously a more superior tool than my eyes. Oh and screw waveform... I shift by listening and estimating... and whenever I did that way, I estimated my shift and an accuracy of less than 60ms can be done in one pass. I just feel sorry for all of you: you guys are all legally deaf. Sad but true.
__________________
A small tool I made, and Learn how to make fansub karaoke in After-Effects: AFXKRK
pichu is offline   Reply With Quote
Old 2009-01-09, 12:29   Link #854
Tofusensei
Senior Member
 
 
Join Date: May 2003
Location: Tokyo, Japan
Send a message via AIM to Tofusensei
Sorry to intrude, but isn't picking a synch point in the video and using it every week usually a better idea? At least it's been in my experience. (Unless the animators are snorting the good stuff).

-Tofu
Tofusensei is offline   Reply With Quote
Old 2009-01-09, 13:01   Link #855
getfresh
done
*Fansubber
 
 
Join Date: Nov 2003
Location: Yokosuka, JP
Age: 33
Quote:
Originally Posted by Tofusensei View Post
(Unless the animators are snorting the good stuff).
I better be careful. God knows I've been smoking the good stuff~

Wait... I'm not an animator! SAFE!
getfresh is offline   Reply With Quote
Old 2009-01-31, 14:21   Link #856
.Hatake
Junior Member
 
 
Join Date: Jul 2008
Hi, I have started to do some karaoke for Soul Eater and I don't know how to do this effect :



I have tested SweetKaraoké Online but doesn't work..

Can anyone help me please ?

My line for now it's :
Code:
Dialogue: 0,0:00:44.53,0:00:46.72,OpKara,,0000,0000,0000,template syllable,{/k}I'm {\k34}{\k40}Fa{\k99}lling{\k0}
__________________


.Hatake is offline   Reply With Quote
Old 2009-01-31, 20:20   Link #857
zhurai-tsuki
...
*Fansubber
 
Join Date: Oct 2007
Location: California
Dialogue: 0,0:00:44.53,0:00:46.72,OpKara,,0000,0000,0000,tem plate syllable,{/k}I'm {\k34}{\k40}Fa{\k99}lling{\k0}
^ you don't call the line template syllable


make a new line:
(same style)
effect: template syl

then read up on karaoke templator
zhurai-tsuki is offline   Reply With Quote
Old 2009-01-31, 20:38   Link #858
.Hatake
Junior Member
 
 
Join Date: Jul 2008
Okay thanks now do you know how to do effect ? Or someone have .ass with some karaoke and effect ?

Because it's hard to do cool effect when you start but it's so boring a simple karaoke without effect
__________________


.Hatake is offline   Reply With Quote
Old 2009-02-01, 03:00   Link #859
zhurai-tsuki
...
*Fansubber
 
Join Date: Oct 2007
Location: California
just simple color change (start-mid: to pink/whatever color) (mid-end: back to white/transparent

and also start-mid: font scale x and y to 120 (or something) and then back to 100 on mid-end

for me that's all i can see atm >.>
zhurai-tsuki is offline   Reply With Quote
Old 2009-03-03, 08:53   Link #860
Kya88
Junior Member
 
Join Date: Mar 2008
It's been a while!

Hope you can help to modify this line for make the effect i need, I tried to do it on my own for a long time, but i guess i'm not smart enough.

Quote:
!retime("syl2end",0,-400+($si-1)*50)!{\fry0\t(0,0,\fry90)\an5\be1\pos($scenter,$smiddle\fad(0,300)}
As you can understand i want the syllabes rotate while fading, but i can't find the right timing...
i tried
Quote:
!retime("syl2end",0,-400+($si-1)*50)!{\fry0\t(!$send-300!,$send,\fry90)\an5\be1\pos($scenter,$smiddle\fad(0,300)}
[/QUOTE]
But it doesn't work the way i like... fad tag is quite particular i think, i can't find the timing for make it work at the same way...

I want the syllabes starts rotate 300ms before the ending while fading gradually, but i can't only make its star rotate after syl till the end of line e make them fade 300ms before the end. I want this to be contemporaneous...

I think my english is not good at all, hope you understood me, otherwise tell me what isn't clear and i'll try explain it in better way.
Hope someone wanna try to improve this line with me
Kya88 is offline   Reply With Quote
Reply

Tags
fansubbing, karaoke, software, subbing

Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT -5. The time now is 11:48.


Powered by vBulletin® Version 3.8.7
Copyright ©2000 - 2014, vBulletin Solutions, Inc.
We use Silk.