Duck and Cover

Hi All !
The PROTOTYPE of editor You can find at http://homepages.atnet.ru/drony/LIPEditor.zip
Coments?!?!?!? You are WELCOME at anchorite2001@yandex.ru

Is it supposed to work? Right now all I see is Myron no matter the ACM/LIP I choose and I don't see the animation. I'm guessing it's just a preview?

I like the idea of a LIP editor though as there is no such tool yet and making them by hand would be a HUGE pain in the ass...

Hi!
Sory ma bad English.
it's ONLY A PROTOTYPE.
MYRON and WASTELAND it's DEFAULT FACE.
Try load and change markers on markers view.

Editor is updated.

AH! This new prototype is MUCH better - since it makes the mnouth move.

For the first prototype it didn't have much use since you could edit the phonems by hand in a text editor with the same result... but now that you actually SEE what it does right away, it's great!

(I noticed you went from version 1.0 to 0.9a :D )

Simply I have overlooked to correct figures in dialogue :(
Forgive ;)

Any more headway with this project? I downloaded the current file from the site indicated but it gives me everybody's friendly neighborhood 0xC0000005 error.

If you've decided to drop this project perhaps you'd publish the source code so we can take up where you left off?

I continue work on this project.
Please mail me about this error more info and info about You hardware and OS.

Turns out you need the .net runtimes installed to use your program. For anybody else having trouble getting it to run, try
this.

Of course I find out it works only *after* pouring over hex dumps for hours.

Anyway keep up the good work Anchorite.

Anchorite I'm curious. I've got most of the .lip format figured out but there's still some stuff that I don't understand. Why are there so many different bytes that represent the same phoneme? And how does the engine detect what mood (good, bad, neutral) to show when a person is speaking?

I found your description of the .lip file on the teamx page and took a look at it (with the help of babelfish). It is very detailed. It doesn't explain how the engine picks good/bad/neutral so I would guess that is determined somewhere else. A note, if you would like to add it, the formula to determine how many long each phoneme is to be displayed based on the 'MARKER-SAMPLE-0' value is

MARKER-SAMPLE / (frequency of ACM file in Hz * 2) == time in seconds

I would be happy to translate your work to english if you like.

>It doesn't explain how the engine picks good/bad/neutral
I think it make script.

>A note, if you would like to add it, the formula to determine how many long each phoneme is to be displayed based on the 'MARKER-SAMPLE-0' value is MARKER-SAMPLE / (frequency of ACM file in Hz * 2) == time in seconds.
More in detail please
I not understand English very well

>I would be happy to translate your work to english if you like
It would be quite good.

I put together a bad translation with a russian-english dictionary.

60 * сиконд == 1 минута

MARKER-SAMPLE == продолжительность Сэмпл в ACM-файле, к которому привязана фонема

22050Hz == частотный ACM-файле


MARKER-SAMPLE / (частотный ACM-файле * 2) == продолжительность сиконд

я IRC-чат ты irc.iskra.ru внутренний #fallout?

ты сказать когда

Do you wanna tell about time, when phoneme must be shown?

But i think more simple way exist for it - check position in sound buffer and show necessary phoneme then it need.
It may be done at another thread or programm may handle message from timer.

пример:

Файл aeld4.lip


17 фонема целый

1. Тип 0 время код 0x0000 фонема 0x00
2. Тип 1 время код 0x00F0 фонема 0x27
3. Тип 1 время код 0x109B фонема 0x09
...

1. (начале) показывать для фонема 0x00
2. задержка для доля секунды (0x00F0 / (22050 * 2)) тогда показывать для фонема 0x27
3. задержка для доля секунды (0x109B / (22050 * 2)) тогда показывать для фонема 0x09
...


вы понимать?


I don't think using the sound buffer is good. What if you want the picture to move longer than the sound plays? A timer is a better idea I think.

Yes. I understand.

>I don't think using the sound buffer is good
Why? In DirectSound it's very good.
In DirectSound I can set notification events on the buffer.
When the buffer plays and the notification position is reached - event goes to nonsignaled state and I can handle it.
But it this version of editor it's not implemented. Yet

Well before I saw this thread about you making a lipfile editor I started making one of my own. I got to thinking that you could use the format for other things beside just showing a person talk if you were tricky about it. In that case you might want to show a frame from an FRM file at a time when no sound file is playing. But if you lock the editor into using the sound buffer as a timeline you can only show frames until the sound stops.

I'm curious, will you publish the sourcecode for your program when you are finished?

Editor is updated.
The basic editing functions are added.

The editor is updated.
Functions for work with templates are added.