Switch Layers, Lip Sync, Audio not loud enough?

[lovebiscuit]

Folks, I have been animating with ASP 8 for a while now and finding it great - but recently have been finding that when asking a switch layer to obey an audio file it doesn't generate key frames for many of the spoken words. It generates them only for the louder sentences. I tried amplifying the audio file to assist the switch layer, but although the sound gets a lot louder, the keyframes remain strangely absent in specific sections.

It's a little puzzling to me, as this has only started happening recently - yet I can't seem to identify anything I am doing particularly differently.

Has anyone else experienced it and, more importantly, is there a way to change the sensitivity of the switch layers to the audio files or some other work around?

Thanks!

[slowtiger]

Open the audio file in Audacity or any other sound editor and normalize it. Then try again.

[Seyrse]

I haven't found a new need to look into this after I did my first animation where I did a bunch of lip sync to a song. But as I have been looking into scripts I decided to look at the sound ones. I found.

-- **************************************************
-- Recurring values
-- **************************************************

LM_BoneSound.audioLayer = 0
LM_BoneSound.magnitude = 90
LM_BoneSound.stepSize = 2

I haven't gone any further as I'm on vacation and poking around on a laptop.

It would seem that the LM_BoneSound. xxx area is where to start. Once one found out what to set you could put it into a external box and adjust.

The other thing I found from my first lip sync is once you run the script the key frames are changed . You can then TWEEK them where you want more bone movement.

Steve E.

[lovebiscuit]

Thanks folks.

I use Audacity in general, slowtiger, so it was easy to see if the normalising thing worked - unfortunately not.

Steve, your suggestion is a bit technical for me!

The sound file in question, a female actor's dialogue, seems very good quality and not unusual in any way - but I just can't seem to make this work so it seems my only option right now is to manually switch the layers.

Thanks again.

[Seyrse]

It would be important to remember the script for lip sync is just a quick and dirty cheat for true lip sync. The script just responds to loudness of the audio track. If you had one that were just drum beats. A good loud sound. You would get a good bone action. It would look weird if the real audio track was a baby laughing. There is no real correlation between the track and sync other than the change of the loudness (DB's). This is where you need the mouth setups for the shape of real sounds. I would re-read the following in the manual. Now that you have tried the script and found it lacking perhaps this review will help.

For the most accurate lip-sync results, we recommend using a program like Papagayo. Lip-syncing using Papagayo is covered in “Tutorial 5.2: Phoneme Lip-Sync” on page 105.

I wounder if you could "add" more action (fine tune to the sound) to each phonemes with the audio script? So a I sound shape would change some based on how loud it was "said".

Steve E.

[lovebiscuit]

Oh, I know, yes...

But I still can't figure out why it's suddenly stopped responding in relation to large swathes of this particular piece of voicework - it's every bit as loud and clear as other pieces of voicework by the same actor.

Generally, I use the tool as a first step and supplement with my own adjustments.

[Seyrse]

Your saying that it was working the way you wanted then suddenly stopped for a section of your audio track?

Steve E.

[lovebiscuit]

Yeah, that's more or less it.

The lead actress voicework was recorded over 3 days. She's a professional and all her material was consistent. The material was recorded with Audacity and the settings did not change throughout. Yet sometimes the switch layer sync tool just doesn't respond.

In the case of the scene I am having trouble with, it waits until half way through the scene and then starts responding. Yet there is no discernible difference - no jump or drop in volume etc.

Normalising the track, also in Audacity, hasn't helped.

It's a boring complaint, because as you say, the switch layer sync is a blunt tool to begin with! But I still wonder why it happens and it does inconvenience me...