Creating Fansubs - Questions and Answers

[Raijenki]

Originally Posted by ScR3WiEuS

haha.
anyway, yeah, timing with spectrum is a whole lot easier.
just be sure to turn the settings down, or you'll get a lot of noise in your spectrum, and it'll be even harder to time.
my karatiming went from 1 hour for a song to 15 minutes thanks to spectrum mode.

I take 15~20 mins in the audio-mode, depending on song.
For normal timing, I prefer audio-mode, I'm more faster on it. There are some parts that is impossible with spectrum.

Anyway, that's just habit.

[Crapzo]

Hi,
Is any one know a software like this ?




This on is just a frontend, not function in...

[Yuudai-kun]

Edit: Nevermind, my mistake.

Thanks martino, i didn't pay attention :/

[martino]

That's a completely different program, looking at the title bar (hint hint); Karaoke Maker vs Advanced Karamaker Studio.

[VAXVA]

Originally Posted by fireshark

VAXVA, I wrote a thing to do that in Aegisub

sorry for not clarify myself proberly, hm because i was new to animesuki forum, i start off a thread asking this question, and recived a warning for duplication thread exisit so admin moved my 'post here'
----------------------------
i'm trying to it in MaestroSBT actually ' i do know subtitlecreator have the ability to do the subtitle but the quality of the graphic subtitle is abit out...
-----------------------------
about guys who mention, big left margin, i do think of it. but this will end up the subtitle all starts from the point you set, which 'div' of subtitle is not centered.

i would also like to know how fireshark do in aegissub, and let me see if maestrosbt read those tag in ssa for maestrosbt feeds
-----------------------------

Extra Question:
is there a font to display japanese character with translation/sound of hanji character? and display on top of that character? i also saw this in all the major japanese dvd subtitle.

example ntsc R2J dvd digimon movie 5

[jfs]

VAXVA: Getting the readings of the kanji above them (called furigana) isn't possible with a font, since one kanji can have multiple readings, but Aegisub (or rather Automation 4 Lua) is able to generate it. However, it isn't that easy... I might make a real guide on it some day.

[ScR3WiEuS]

Originally Posted by Raijenki

I take 15~20 mins in the audio-mode, depending on song.
For normal timing, I prefer audio-mode, I'm more faster on it. There are some parts that is impossible with spectrum.

Anyway, that's just habit.

try changing the spectrum settings ;p nothing is impossible.

[Kristen]

Sorry, this may be a stupid question, but I've tried two encoders so far and myself, and it hasn't worked.
I'm trying to encode a 720p DCIISS. Description is that it is vfr, aac, in an mp4 container. However, it's a really weird vfr. It comes up as 23.976 fps when you check it in avsp, and 25.00 when you check it in VDubMod. So, I went ahead and used the same filter chain I always do.

Spoiler for Filter Chain:

LoadPlugin("C:\Documents and Settings\hrissie\Desktop\Encoding\AviSynth_Plugins \awarpsharp.dll")
LoadPlugin("C:\Documents and Settings\Chrissie\Desktop\Encoding\AviSynth_Plugin s\Deen.dll")
LoadPlugin("C:\Documents and Settings\Chrissie\Desktop\Encoding\AviSynth_Plugin s\undot.dll")
LoadPlugin("C:\Documents and Settings\Chrissie\Desktop\Encoding\AviSynth_Plugin s\MSharpen.dll")
DirectShowSource("C:\Documents and Settings\Chrissie\Desktop\DCIISS06.mp4", fps=23.976, convertfps=true, audio=false)
ConvertToYV12()
awarpsharp()
deen()
msharpen(threshold=10,strength=90,mask=false,highq =true)
undot()

Sure, laugh at my filter chain if you want. I've been told it's crap multiple times already. I just know it works for my purposes.
I drag that into VDubMod, change the compression to lagarinth (RGB checked, null frames enabled)
Anyways, I get a lossless from that. Then I extract audio in VDubMod as a .wav. I tried it both by using the raw as the audio and the .wav as the audio.
Then I opened up Megui. I imported the video in with the following AVSP script:

Spoiler for h.264 video script:

AviSource("C:\Documents and Settings\Chrissie\Desktop\Lossless.avi")
Converttoyv12
#deinterlace
#crop
#resize
Undot() # Minimal Noise

I used the following video profile:

Spoiler for Video Profile:

--pass 2 --bitrate 250 --stats ".stats" --no-cabac --subme 6 --partitions p8x8,b8x8,i4x4 --me umh --threads auto --thread-input --progress --no-psnr --no-ssim --output "output" "input"

Then, I imported in the audio. I tried using both the raw and the extracted .wav, but it both produced the same results.

Spoiler for Audio Profile:

Uncheked "Force Decoding", keep original channels, No increase volume, improve accurace checked, HE-AAC+PS prfile, Channel Mode: Joint, unchecked "Produce MPEG2 AAC", slider at 128 kbps. Codec is winampAAC, and the extension is Raw-AAC. Delay 0 ms

Then I hit "AutoEncode". Under container I select the MKV container, no splitting. For size and bitrate, I select "File Size" and click the arrow. I then click "Select size" and input 222 MB. Queue it, start it.

The end result I get is video and audio that aren't synched. How do I fix that?

[TheFluff]

Originally Posted by ChrissieXD

The end result I get is video and audio that aren't synched. How do I fix that?

I'd suggest not using directshowsource for the lossless.

(also JESUS CHRIST STOP USING DEEN, I WILL KILL A KITTEN EVERY TIME I SEE YOU DO THIS)

[edogawaconan]

Originally Posted by TheFluff

I'd suggest not using directshowsource for the lossless.

(also JESUS CHRIST STOP USING DEEN, I WILL KILL A KITTEN EVERY TIME I SEE YOU DO THIS)

and freakin default values for awarpsharp

oh and stop converting colorspace

and perhaps not specifying framerate helps

and why are you still using vdubmod

and why don't you encode video first and use raw audio source directly to mux to mkv

and use mkvmerge

and add some chapters

oh also if you're going to hardsub, might be better just use mp4

etc etc

ps: which team are you working at? I'll try avoiding them

[martino]

...and UnDot since there's RemoveGrain(1).

[Kristen]

TheFluff: What do I use instead? I had someone try mkvsource, but that didn't work.

edogawaconan: That was the responce I was expecting... Really I just want audio and video to line up, I don't want filter help. They come out pretty enough for my purposes (Speedsubs), and encode fast enough so I have the h.264 ready to mux by the time the scripts are ready.
My process is lossless->Megui encode (Raw 222 MB h.264)->MKVMerge (Subs). But the Megui encode is what is giving me issues. So, I do use MKVMerge.
And no chapters. Speedsubs don't have chapters. XD. We would only release XVid anyways if h.264 wouldn't give us instant release (Just mux subs and go. ^_^)
So, yeah. I do softsub. It's just the initial encode giving me issues.
Also, could the frame rate conversion be what's giving me the issues? You mentioned it, but I wasn't sure if that was part of the "low quality" issues you were saying, or part of the desynchronization.
The group name is Chihiro. Speed subs.

[TheFluff]

Originally Posted by ChrissieXD

TheFluff: What do I use instead? I had someone try mkvsource, but that didn't work.

Uuuuuuh. The lossless is an AVI, so use avisource() maybe? @____@

(also, no filtering at all would give a lot better results AND encode a lot faster than the filtering you are doing, you should stick to muxing raws together with softsubs)

[Kristen]

Originally Posted by TheFluff

Uuuuuuh. The lossless is an AVI, so use avisource() maybe? @____@

(also, no filtering at all would give a lot better results AND encode a lot faster than the filtering you are doing, you should stick to muxing raws together with softsubs)

Oh, you mean in the final encode. I haven't tried that yet, so I will as soon as I get home.

And I don't want to release 500 MB episodes, so I need to encode them, at least in Megui. But, no, I've compared raws to the filters, and I like the filtered a lot better. XD

[TheFluff]

Originally Posted by ChrissieXD

But, no, I've compared raws to the filters, and I like the filtered a lot better. XD

enjoy your eyecancer

[Eeknay]

You're doing it wrong, you need to go for gold when using those filters.

[Nicholi]

I've always considered the usage of warpsharp more unto an art form myself :P.
Ancient pic of goodnes! Name that animu!

[TheFluff]

Originally Posted by Nicholi

I've always considered the usage of warpsharp more unto an art form myself :P.
Ancient pic of goodnes! Name that animu!

I believe the term appropriate here is "fukken saved".
(Seriously my collection has been missing that for a good long while, thank you kind sir!)

[Bot1]

i have a school rumble pic with something like that somewhere...

oh and it kinda looks like magnetic rose from memories

[dragon5152]

LoadPlugin("C:\Documents and Settings\hrissie\Desktop\Encoding\AviSynth_Plugins \awarpsharp.dll")
LoadPlugin("C:\Documents and Settings\Chrissie\Desktop\Encoding\AviSynth_Plugin s\Deen.dll")
LoadPlugin("C:\Documents and Settings\Chrissie\Desktop\Encoding\AviSynth_Plugin s\undot.dll")
LoadPlugin("C:\Documents and Settings\Chrissie\Desktop\Encoding\AviSynth_Plugin s\MSharpen.dll")
DirectShowSource("C:\Documents and Settings\Chrissie\Desktop\DCIISS06.mp4", fps=23.976, convertfps=true, audio=false)
ConvertToYV12()
awarpsharp()
deen()
msharpen(threshold=10,strength=90,mask=false,highq =true)
undot()

Allow me to fix your script based on no knowledge of the source:

loadplugin("ffmpegsource")
ffmpegsource("C:\Documents and Settings\Chrissie\Desktop\DCIISS06.mp4")
fft3dfilter(sigma=1,bt=3,plane=0,bw=32,bh=32,ow=16,oh=16) #adjust sigma, higher = more denoising, try not to go over 2
LimitedSharpenFaster(strength=50) #this step is optional, try not to go over 100 let alone 50 with more sources, this will probably have to be decreased or not used depending on source