Today I made a quick video, and the lip syncing was probably not perfect to anyone who really pays attention, but it was good enough I could publish a video and focus on content without wasting time on mouth poses (vsemes is the word I think).

My microphone (Shure SM7B) is probably too good for my noisy apartment, so I have to record with my microphones interface 'amp' option turned down, then amplify using audio software (Audacity and / or AVS Audio Editor) and it comes out to sounding near studio quality now that I moved my PC away from my desk and put a 2 inch thick gymnast mat in front of it.

When I upload an audio file, the mouth poses just don't seem to sync as well. 

From my past experiences, audio files were so bad I used Text to Speech and adjusted my audio file to match, but the recorded voice option is close enough I could eliminate the audio editing.

The only problem is this means I have to record between 3:00 AM and 4:45 AM, and I usually go to bed about midnight.

Just curious if this is unique to my environment or if anyone else has experienced this?

Thanks

Creator of the free website:
PixelDatabase.Net
A Free Online Text Based Image Editor
75% of original size (was 670x19) - Click to enlarge

---

Alienware Aurora R16, Win 11, i9-149000KF, 3.20GHz CPU, 64GB RAM, RTX 4090 (24GB), Samsung 870 Pro 8TB, Gen3 MVNe M-2 SSD, 4TBx2, 39" Alienware Widescreen Monitor
Mike "ex-genius" Kelley

Yeah, the purpose of that video is to show you can use Papagayo with iClone.  Indeed, since I made that the new Python release for iClone makes it easier (because now we can delete and move visemes on the track -- so just add the command to delete them all first.  If this turns out something you want to use, let me know and I'll alter my script for that purpose).

I also have a number of tutorials on just using Papagayo (for the other animation program, but they are then applicable for iClone).  It's really easy, though, and I think the results would be great in iClone (as I said in the video, I use Liveface so I don't really need this, but it was kind of a thing on my "to-do" list long before we got facial mocap available.  I still might use it for some of my pre-recorded audio tracks I have).

---

Alienware Aurora R16, Win 11, i9-149000KF, 3.20GHz CPU, 64GB RAM, RTX 4090 (24GB), Samsung 870 Pro 8TB, Gen3 MVNe M-2 SSD, 4TBx2, 39" Alienware Widescreen Monitor
Mike "ex-genius" Kelley

Just a theory - The difference you notice may be due to the sound quality of the audio. 
But in this case, less is more perhaps....
If the recorded audio has higher treble/volume peaks that are less compressed, it may be the case that such peaks more easily trigger phonemes mapped on the timeline by the conversion process.

Actually more is more probably. When I record directly from IClone, I have to crank up my AMP just like I do for Skype conversations.
For Skype it doesn't matter if people hear a little background noise, but for a video it just ruins it. 
But now that I think about what you said, by using the raw video amplified IClone picks up the syllable changes better probably.

I only did 1 test, I was planning on doing more testing if I ever get caught up. 
I am going to try using an amplified version to do the vsemes, and then I can use a edited version for produced animation.

Kellytoons option looks great if I can get it working, but any steps that can be saved is always best.

I will try some more tests after work. 



Creator of the free website:
PixelDatabase.Net
A Free Online Text Based Image Editor
75% of original size (was 670x19) - Click to enlarge

I wish Reallusion would add something like PapaGayo, 

what a great editor for vsemes compared to a timeline.what a great editor for vsemes compared to a timeline.

RAG DOLL COLLISION ANIMATIONS FOR ICLONE 8 & 7
---------------------------------------------------------------------------------------------------------------------

Ghost Origins

My latest Feature length film created with Iclone.
75% of original size (was 670x19) - Click to enlarge

My Sci- Fi Graphic Novel on Amazon: https://a.co/d/9k3cwoY

My IMDB listing
Anabran Abdhul Quarnain - IMDb

@Kellytoons, is that code file published on GitHub or anywhere? 

If not, could you post the source code here, or email me please?

I am ready to Python tonight (is Python also a verb, or would be slither like a lawyer)?

Thanks

Creator of the free website:
PixelDatabase.Net
A Free Online Text Based Image Editor
75% of original size (was 670x19) - Click to enlarge

In that video on YouTube there's a link to the code (it's Dropbox so it lives there permanently).  Hmmm, tried to copy it and it didn't work.  Let me know if you can't find it there (it's in the comments section where someone else asks for it).

---

Alienware Aurora R16, Win 11, i9-149000KF, 3.20GHz CPU, 64GB RAM, RTX 4090 (24GB), Samsung 870 Pro 8TB, Gen3 MVNe M-2 SSD, 4TBx2, 39" Alienware Widescreen Monitor
Mike "ex-genius" Kelley

Pretty awesome stuff, although currently beyond my scope ;)
I will surely check this out at a later time.
Thanks guys!

---