Keep me updated on progress guys. Once the file is split up we can work out a plan to get the lip syncing going. Hopefully we can get several people on that.
Giving Cassidy a voice.
- Thread starter Dravean
- Start date
Soda_fr
It Wandered In From the Wastes
@Dravean
PM sent. I'll will send you a sample with the new filter to get your approval before moving on with the rest of the files.
@tobecooper
Thanks for the offer, I shall put a short tutorial within few days to help you getting started.
@Killap
I'm away from home for 2 days, but I shall be able to start working on the chopping/converting/filtering on Wednesday. Depending on the number of files, it might take few days so I'm not sure I can make it for the end of the public beta, but I'll do my best.
PM sent. I'll will send you a sample with the new filter to get your approval before moving on with the rest of the files.
@tobecooper
Thanks for the offer, I shall put a short tutorial within few days to help you getting started.
@Killap
I'm away from home for 2 days, but I shall be able to start working on the chopping/converting/filtering on Wednesday. Depending on the number of files, it might take few days so I'm not sure I can make it for the end of the public beta, but I'll do my best.
I'm going to be extending the beta anyway, so no worries.Aguirre said:
Let me know how I can assist in all this. I do appreciate you all stepping in though!
Guys, this is the phoneme chart for Cassidy:
The numbers are the animation frames. The letters DO NOT represent letters, but SOUNDS or PHONEMES, i.e.: phoneme "I" doesn't read "Aye", but "Ee". And phoneme "E" is not "Ee", but rather like the "e" in the word "red" ("eh"?).
This is also the reason why there are no letters "Q" or "C" in here, because these phonemes would read the same as phonemes "K" or "S".
The talking animation doesn't have a good mouth position for "U", so it can either be assigned as position 1 or 3 - the artist will have to see what fits best in a given line.
Other than these few remarks, I'm sure Aguirre will fill you in on how to use the chart effectively. He's the master of the trade.
What else can I say? Good luck with the files, I guess
The numbers are the animation frames. The letters DO NOT represent letters, but SOUNDS or PHONEMES, i.e.: phoneme "I" doesn't read "Aye", but "Ee". And phoneme "E" is not "Ee", but rather like the "e" in the word "red" ("eh"?).
This is also the reason why there are no letters "Q" or "C" in here, because these phonemes would read the same as phonemes "K" or "S".
The talking animation doesn't have a good mouth position for "U", so it can either be assigned as position 1 or 3 - the artist will have to see what fits best in a given line.
Other than these few remarks, I'm sure Aguirre will fill you in on how to use the chart effectively. He's the master of the trade.
What else can I say? Good luck with the files, I guess
I've got everything I need to start the lip syncing. Big thanks to Ardent and Aguirre for all this.
If anyone is willing to help in this area, please pm me. It'd be great to get ~5 people, so that each person ends up working on 10 files (we have 55 files in total and I'll probably end up working on some myself).
We have written instructions (with pictures), in addition to a video tutorial. I'll send these to you once you've expressed interest and if you become uninterested after seeing them, that's perfectly fine.
Looks like a voiced Cassidy just might be appearing as an option in the upcoming final RP.
If anyone is willing to help in this area, please pm me. It'd be great to get ~5 people, so that each person ends up working on 10 files (we have 55 files in total and I'll probably end up working on some myself).
We have written instructions (with pictures), in addition to a video tutorial. I'll send these to you once you've expressed interest and if you become uninterested after seeing them, that's perfectly fine.
Looks like a voiced Cassidy just might be appearing as an option in the upcoming final RP.
Soda_fr
It Wandered In From the Wastes
Agris also deserves a big thank.
@Killap: I'll be doing the next 3 lip files (Cas0112, Cas0114, Cas0118) so that the 5 potential volunteers get exactly 10 files each.
I'll PM them to you in 2/3 days.
@Killap: I'll be doing the next 3 lip files (Cas0112, Cas0114, Cas0118) so that the 5 potential volunteers get exactly 10 files each.
I'll PM them to you in 2/3 days.
Sigh.. I can't believe how bad I am at this. This isn't going to happen if I'm the only one doing it, so volunteers are greatly needed...
EDIT: Maybe it would help to have more example sounds to go with the chart you have. Like the examples you were giving Ardent. I don't know, maybe it's just me.
EDIT: Maybe it would help to have more example sounds to go with the chart you have. Like the examples you were giving Ardent. I don't know, maybe it's just me.
Okay, let's try to analyze one of Cassidy's lines (line 137). Note, I don't have Dravean's recording of it, so it'll be done purely on the basis of my knowledge of English phonetics (not too good):
You always need to start with closed mouth. So drop the first marker at 0000 and use the default position, which in the case of Cassidy's TH is 6.
Now, I'll just take a short bit of the dialogue line and see what it gives:
First word: "After". I'm not sure whether Dravean pronounced the "R" here, but I'd assume it to be barely audible anyway, so for the first word we only need 4 phonemes (therefore, 4 markers in the lip editor): A-F-T-E.
In animation frames, that makes: 7-2-0-5
Each marker must be separated by a couple of milliseconds and obviously, they need to match the sounds as they are being pronounced. You've got to do it by ear and sight - listen to the voice and watch the mouth to see if it moves when the sounds change. It's the most difficult and the most tedious part of the whole process.
Then there's "they", so we only need three phonemes: TH-E-Y
Or: 8-5-1
These were the easy words. The next word in line: "hauled". Since "au" is pronounced like some variant of "o", and since you don't pronounce the "e", then you only need 4 phonemes: H-O-L-D
Or in animation frames: 1-7-8-0.
Next word: "away".
The "a"'s are probably pronounced more like "e" ("red", not "ee"), so we don't use "A" phoneme for them, but rather "E" phoneme, which incidentally is animation frame 5. There's also "W", for which we need to type in frame 3.
Finally, there's the final "Y" sound (like in "Hey"). In away, normally you would be closer to the "I" phoneme, but since that's the same animation frame as the earlier "E", we need to use something different, so let's go with frame 1 for "Y".
That gives us: 4-3-5-1.
I'll skip to the word "one" now, since that's interesting for its pronunciation: in phonetic transcription this would be written as /wan/, so you need to use phonemes W-A-N, or animation frames: 3-7-0.
Some other remarks or clues:
When these three elements are fulfilled, you'll achieve great results. And with practice, you will also learn to hear and recognize sounds (phonemes) that you need to use easily.
You always need to start with closed mouth. So drop the first marker at 0000 and use the default position, which in the case of Cassidy's TH is 6.
Now, I'll just take a short bit of the dialogue line and see what it gives:
First word: "After". I'm not sure whether Dravean pronounced the "R" here, but I'd assume it to be barely audible anyway, so for the first word we only need 4 phonemes (therefore, 4 markers in the lip editor): A-F-T-E.
In animation frames, that makes: 7-2-0-5
Each marker must be separated by a couple of milliseconds and obviously, they need to match the sounds as they are being pronounced. You've got to do it by ear and sight - listen to the voice and watch the mouth to see if it moves when the sounds change. It's the most difficult and the most tedious part of the whole process.
Then there's "they", so we only need three phonemes: TH-E-Y
Or: 8-5-1
These were the easy words. The next word in line: "hauled". Since "au" is pronounced like some variant of "o", and since you don't pronounce the "e", then you only need 4 phonemes: H-O-L-D
Or in animation frames: 1-7-8-0.
Next word: "away".
The "a"'s are probably pronounced more like "e" ("red", not "ee"), so we don't use "A" phoneme for them, but rather "E" phoneme, which incidentally is animation frame 5. There's also "W", for which we need to type in frame 3.
Finally, there's the final "Y" sound (like in "Hey"). In away, normally you would be closer to the "I" phoneme, but since that's the same animation frame as the earlier "E", we need to use something different, so let's go with frame 1 for "Y".
That gives us: 4-3-5-1.
I'll skip to the word "one" now, since that's interesting for its pronunciation: in phonetic transcription this would be written as /wan/, so you need to use phonemes W-A-N, or animation frames: 3-7-0.
Some other remarks or clues:
- When there's a pause between sentences, make sure to return to the default position, i.e. frame 6.
- Also, make sure to end the file with the default position. It should be introduced when the speech finishes (usually 4000-6000 units after the penultimate phoneme).
- Don't read lines, listen to them. There are plenty of words in English that have letters that you don't even pronounce (knife = N-A-Y-F), or words that are pronounced much different than the letters would suggest (one = W-A-N; eye = A-Y) or both (bough = B-A-W; caught = K-O-T); Once you learn that phonemes != letters, you'll do fine.
- Imagine spoken language as a frog jumping across the river on stepping stones. Getting across the river is a word. Whenever the frog is in the air, it's a vowel (open mouth: A, E, I, O, U), and whenever it touches a stepping stone, it's a consonant (any other sound - most often involves partial or total closure of mouth).
Vowels usually last longer than consonants (the frog jumps high up), so people tend to notice them more. If your "Oh"'s, "Ah"'s, "Eh"'s, "Ee"'s and "Ooh"'s open up at the same moments as they can be heard in the recording - you're halfway there.
The other important bit are consonants with mouth completely closed, such as B, P and M. People will notice these often as well, as this is the clearest example of how your frog jumps (open-closed-open-closed OR air-stone-air-stone). If you make sure to punctuate your M's, B's and P's in the right places, the animation will look faithful, even if you make a glitch or two in more obscure sounds, such as K, H, G, L, T etc.
- The final detail is to make sure that the animation starts precisely at the moment when the voice can be heard and ends when the voice dies down.
When these three elements are fulfilled, you'll achieve great results. And with practice, you will also learn to hear and recognize sounds (phonemes) that you need to use easily.
I'm glad you work for us. 
tobecooper
First time out of the vault
Josan12 said:
Have you tried YU/YO and TU? They are so small that you can sometimes mask them with other words.
@Josan12: Hey, first off, sorry for taking so long with replying to your question - it's been a hectic week for me.
Well, this head has a particular problem of not having a frame that would look like "oo". So I guess you will have to make do with putting frame 0 for T and either frame 1 or 3 for U (OO), and similarly with you - frame 1 for "Y" and frame 3 for U (you can't use 1 again, obviously).
If that doesn't look good, try using frames 7 or 8 for the U sound, instead of frames 1 or 3. You just have to watch the animation and decide which works best, I'm afraid.
Hope this helps, cheers!
Josan12 said:
Well, this head has a particular problem of not having a frame that would look like "oo". So I guess you will have to make do with putting frame 0 for T and either frame 1 or 3 for U (OO), and similarly with you - frame 1 for "Y" and frame 3 for U (you can't use 1 again, obviously).
If that doesn't look good, try using frames 7 or 8 for the U sound, instead of frames 1 or 3. You just have to watch the animation and decide which works best, I'm afraid.
Hope this helps, cheers!
Thanks to the work of tobecooper and Josan12, we're getting through this. But if anyone else wants to lend a hand, it would be most appreciated!
Ardent said:
Thanks Ardent/tobecooper. You're both suggesting the same thing which is to use U, and that's mostly what i've done. As you say, Ardent, the TH doesn't really have a 'pursed lips' frame which is needed for an 'ooo' sound. Oh well.
As Killap says, any volunteers would be welcomed.
