heyas;
okay, so i'm following this proceedure....
1: enter dialogue text and cinematic face expression.
2: generate VO and FaceFX (robot voice)
3: lots of rigamarole to get my wav files to go with the dlg file (tutorial forthcoming!)
now, the part i'm a little uncertain on...
i preview the line... and the lipsynch is off. let's say for example, the VO is talking really fast, and the lipsynch is talking at normal speed. so, i tweak the dialogue text to shorten it up to make the lipsynch match the actual sound file better.
this results in the dlg generating a new robot VO. so i save the dlg, close the toolset, delete the temp fsb/fev the robot vo created, and restart the toolset. then i have new lipsynch to the real wav file.
aside from going in and manually editing the facefx, is there a shorter way to do this? DOES preview with new robobrad re-do the lipsynch, without me having to retype the line? (i thought that fixed a bad lipsynch one time, but no luck these latter days.) or is there a way for me to retype the line and get new lipsynch to it without creating a robot vo wav file?
am i totally doing this wrong?
Lipsynch Tweaking
Débuté par
BloodsongVengeance
, oct. 02 2010 12:45
Retour en haut
#2
Posté 02 octobre 2010 - 11:28
-Semper-
-
- Members
- 2 256 messages
just a question: you're talking about wav files. do you have recorded your dialogue lines with real voices? if that's the case you should name them like the appropriate line (e.g.: 1456015_m.wav) and just do the facefx. the generate vo is just a quick placeholder to test a dialogue without the need to record something.
#3
Posté 03 octobre 2010 - 01:28
BloodsongVengeance
-
- Members
- 590 messages
heyas;
yeah, it has custom wav files in there, and that's what the lip synch isnt matching. i did all the naming and numbering and fsb and fev and fdp junk.
okay, i think i sussed it.... you right click the line, and edit facefx. in the animation tab, instead of the animation curves... there's the text of the dialogue and phonemes. if you mess around stretching and squashing those, then post to local.... you can mush the actor's lips around to the right timing.
unless there's something else im missing?
like, oh.... once i have real wav files, do i re-do the generate vo and face stuff again? or does that mean to make robot voices again? also, when i DO that... and it outputs a bunch of txt files for each line... is ANYthing ANYwhere making use of those txt files? instead of the written dialogue in the conversation?
this whole thing is wiggy. :X
yeah, it has custom wav files in there, and that's what the lip synch isnt matching. i did all the naming and numbering and fsb and fev and fdp junk.
okay, i think i sussed it.... you right click the line, and edit facefx. in the animation tab, instead of the animation curves... there's the text of the dialogue and phonemes. if you mess around stretching and squashing those, then post to local.... you can mush the actor's lips around to the right timing.
unless there's something else im missing?
like, oh.... once i have real wav files, do i re-do the generate vo and face stuff again? or does that mean to make robot voices again? also, when i DO that... and it outputs a bunch of txt files for each line... is ANYthing ANYwhere making use of those txt files? instead of the written dialogue in the conversation?
this whole thing is wiggy. :X
#4
Posté 03 octobre 2010 - 01:17
-Semper-
-
- Members
- 2 256 messages
if you have wave files you DON'T do the generate vo because you already have the voices. just generate the facefx and it will automatically do the lipsynch.
Modifié par -Semper-, 03 octobre 2010 - 01:17 .
#5
Posté 04 octobre 2010 - 12:01
BloodsongVengeance
-
- Members
- 590 messages
ummmm...
how do i get it to assign string id numbers? i thought the 'generate vo' step did that. i've been following charsen's vo tutorial (but fudging some stuff because i'm not copying wavs back into the original bank).
http://social.biowar...1245/blog/3370/
i really don't understand what all those steps are doing, i just monkeyed around with the steps til i got what i wanted, and it worked. if i'm doing it the long way, please let me know what a shorter method might be!
thanks
how do i get it to assign string id numbers? i thought the 'generate vo' step did that. i've been following charsen's vo tutorial (but fudging some stuff because i'm not copying wavs back into the original bank).
http://social.biowar...1245/blog/3370/
i really don't understand what all those steps are doing, i just monkeyed around with the steps til i got what i wanted, and it worked. if i'm doing it the long way, please let me know what a shorter method might be!
thanks
#6
Posté 04 octobre 2010 - 03:48
-Semper-
-
- Members
- 2 256 messages
BloodsongVengeance wrote...
how do i get it to assign string id numbers?
the line automatically get's an id number after creation which is displayed in the bottom left corner. now you only need to name you wave file the correct way (id_m.wav in pcm 24khz and 16b mono) as described and saved there:
my documents/bioware/dragon age/addins/module_name/module/override/toolsetexport
afterwards you only have to generate facefx and the toolset will find your custom voice over files and will do correct lip movement. in your case there are no other steps needed.
#7
Posté 07 octobre 2010 - 10:53
SilentCid
-
- Members
- 338 messages
BloodsongVengeance wrote...
ummmm...
how do i get it to assign string id numbers? i thought the 'generate vo' step did that. i've been following charsen's vo tutorial (but fudging some stuff because i'm not copying wavs back into the original bank).
http://social.biowar...1245/blog/3370/
i really don't understand what all those steps are doing, i just monkeyed around with the steps til i got what i wanted, and it worked. if i'm doing it the long way, please let me know what a shorter method might be!
thanks
Follow my VO video tutorial it shows how to lip-sync everything. Plus assigning the right string ID numbers.
#8
Posté 10 octobre 2010 - 03:10
BloodsongVengeance
-
- Members
- 590 messages
heyas;
okay guys, i've tried this six ways til tuesday.... it is NOT WORKING as you seem to be saying it should.
can you please explain....
WHAT is an fev, and what is it for? what, if anything, in the toolset uses it? what is in it?
WHAT is an fsb, and what is it for? what, if anything, in the toolset uses it? (i get that wav files are in it.)
WHAT is an fdp, and what is it for? what, if anything, in the toolset uses it?
if you make an fdp from scratch... how the bloody blue blazes do you get wav files into it????? (which is probably not neccessary, but i'm lost and clueless on what fmod is supposed to do, or what to do when it doesn't automatically do this magic thing it does.)
if the toolset can use raw wav files in its override instead of an fdp... does a module need an fdp? what is the ultimate purpose of the fdp?
WHAT uses the .wav.ph and the .txt files that the local VO and local fFX generate? anything? what are they for??? actually... where do they come from, because while i was dicking around expiermenting, these never got created.
Sid: you do the same thing challseus does. you generate robot vo wav files, then put in the real vo wav files. can you explain WHY you do that step? can you explain WHY you do 'generate vo local' *again* after you put your own wavs in? i thought the 'generate vo local' meant 'make fake robot voices.' does it mean something different?
Semper: i can't make it work the way you're telling me, by just having my wav files there and generating face fx. it just keeps failing at everything. it fails the fFX because it can't generate vo; it cant generate vo because it tells me it's missing a 'fmt ' chunk, or that my wavs are not 24000 hz, and all kinds of RPU error 1's.
am i missing some other step or something? can you list the steps you use?
i agree with you that there are some steps in there that are probably not neccessary. however, only by doing all the steps do i get anything to work right. i try to skip one... im screwt. :/
okay guys, i've tried this six ways til tuesday.... it is NOT WORKING as you seem to be saying it should.
can you please explain....
WHAT is an fev, and what is it for? what, if anything, in the toolset uses it? what is in it?
WHAT is an fsb, and what is it for? what, if anything, in the toolset uses it? (i get that wav files are in it.)
WHAT is an fdp, and what is it for? what, if anything, in the toolset uses it?
if you make an fdp from scratch... how the bloody blue blazes do you get wav files into it????? (which is probably not neccessary, but i'm lost and clueless on what fmod is supposed to do, or what to do when it doesn't automatically do this magic thing it does.)
if the toolset can use raw wav files in its override instead of an fdp... does a module need an fdp? what is the ultimate purpose of the fdp?
WHAT uses the .wav.ph and the .txt files that the local VO and local fFX generate? anything? what are they for??? actually... where do they come from, because while i was dicking around expiermenting, these never got created.
Sid: you do the same thing challseus does. you generate robot vo wav files, then put in the real vo wav files. can you explain WHY you do that step? can you explain WHY you do 'generate vo local' *again* after you put your own wavs in? i thought the 'generate vo local' meant 'make fake robot voices.' does it mean something different?
Semper: i can't make it work the way you're telling me, by just having my wav files there and generating face fx. it just keeps failing at everything. it fails the fFX because it can't generate vo; it cant generate vo because it tells me it's missing a 'fmt ' chunk, or that my wavs are not 24000 hz, and all kinds of RPU error 1's.
am i missing some other step or something? can you list the steps you use?
i agree with you that there are some steps in there that are probably not neccessary. however, only by doing all the steps do i get anything to work right. i try to skip one... im screwt. :/
#9
Posté 24 octobre 2010 - 09:59
BloodsongVengeance
-
- Members
- 590 messages
DONT MAKE ME START A NEW THREAD!
okay, this whole thing is driving me bugnuts!!
today, i followed my old, multi-stepped process for doing the lipsynch and VO. yes, all the generate VO and fFX. and throw wav files in the temp directory, and the whole make an FDP/FSB/FEV file.
and i put the fsb/fev into the add-in module core override. and i hit preview line, and behold! it used my wav file!
of course, i couldn't see if the lipsynching was right, because the npc is a dwarf and her camera was looking at the nekkie human stand-in stomach. fine. i added a camera to the stage i was using as a test camera pointing at the human face.
the lipsynching wasn't synched at the beginning, so i hit: preview line with new robobrad.
as soon as i did that, it would NOT STOP USING ROBOT VOICE.
i closed the toolset, i moved my fsb/fev to the packages/core/override, i deleted all the fxe/fxm files out of the other override. i deleted the temp wav.ph files out of the temp placeholder-vo directory. i put the raw numbered and _m'ed wav files in the override. i generated another local vo. i did everything i could THINK of, but the smegging thing is STILL USING ROBOT VOICE instead of my wav file! and i dont know HOW! i CANT MAKE IT STOP!
OMG ITS DRIVING ME NUTS!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
can't anyone explain to me how all these flipping files work and how i can KILL ROBOT VOICE?????!?!
oh, i forgot, these are the steps i figured out and used: they worked before, i swear!!!
https://docs.google....6k5g5_5g5dkrgdv
okay, this whole thing is driving me bugnuts!!
today, i followed my old, multi-stepped process for doing the lipsynch and VO. yes, all the generate VO and fFX. and throw wav files in the temp directory, and the whole make an FDP/FSB/FEV file.
and i put the fsb/fev into the add-in module core override. and i hit preview line, and behold! it used my wav file!
of course, i couldn't see if the lipsynching was right, because the npc is a dwarf and her camera was looking at the nekkie human stand-in stomach. fine. i added a camera to the stage i was using as a test camera pointing at the human face.
the lipsynching wasn't synched at the beginning, so i hit: preview line with new robobrad.
as soon as i did that, it would NOT STOP USING ROBOT VOICE.
i closed the toolset, i moved my fsb/fev to the packages/core/override, i deleted all the fxe/fxm files out of the other override. i deleted the temp wav.ph files out of the temp placeholder-vo directory. i put the raw numbered and _m'ed wav files in the override. i generated another local vo. i did everything i could THINK of, but the smegging thing is STILL USING ROBOT VOICE instead of my wav file! and i dont know HOW! i CANT MAKE IT STOP!
OMG ITS DRIVING ME NUTS!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
can't anyone explain to me how all these flipping files work and how i can KILL ROBOT VOICE?????!?!
oh, i forgot, these are the steps i figured out and used: they worked before, i swear!!!
https://docs.google....6k5g5_5g5dkrgdv
Modifié par BloodsongVengeance, 24 octobre 2010 - 10:01 .
