Profile Picture

Looking for tips on the audio aspect of the animation process

Posted By IO1 4 Months Ago
You don't have permission to rate!

Looking for tips on the audio aspect of the animation process

Author
Message
IO1
General.png
IO1
Posted 4 Months Ago
View Quick Profile
Distinguished Member

Distinguished Member (1.1K reputation)Distinguished Member (1.1K reputation)Distinguished Member (1.1K reputation)Distinguished Member (1.1K reputation)Distinguished Member (1.1K reputation)Distinguished Member (1.1K reputation)Distinguished Member (1.1K reputation)Distinguished Member (1.1K reputation)Distinguished Member (1.1K reputation)

Group: Forum Members
Last Active: Last Month
Posts: 76, Visits: 365
I've been looking around and have manged to find good sources for voice overs but im not sure if i should add the voice overs before or after i've animated, same goes for sound effects like footsteps, guns ect. Has anyone allready got an effcent process developed that they use to accommodate the audio side of thier projects. Thanks for your help. IO
BenjaminTuttle
BenjaminTuttle
Posted 4 Months Ago
View Quick Profile
Distinguished Member

Distinguished Member (2.6K reputation)Distinguished Member (2.6K reputation)Distinguished Member (2.6K reputation)Distinguished Member (2.6K reputation)Distinguished Member (2.6K reputation)Distinguished Member (2.6K reputation)Distinguished Member (2.6K reputation)Distinguished Member (2.6K reputation)Distinguished Member (2.6K reputation)

Group: Forum Members
Last Active: Yesterday
Posts: 310, Visits: 3.9K
Voice overs in iClone, usually it would be the first I add in the character, then I work on animations to match the voice. That's the only process in iClone that I would add sound. 

Anything like footsteps I add in later, I have a weird set up on how to get it right. I add a layer of noise tone through a prop every time they make a step, think of it as a marker. I match the footstep with the tone and delete the tone when I am done. 

Ambience, other noises, and music goes into post production for full control down the road. 

iClone Certified Director and makes miserable Westerns.
animagic
animagic
Posted 4 Months Ago
View Quick Profile
Distinguished Member

Distinguished Member (19.3K reputation)Distinguished Member (19.3K reputation)Distinguished Member (19.3K reputation)Distinguished Member (19.3K reputation)Distinguished Member (19.3K reputation)Distinguished Member (19.3K reputation)Distinguished Member (19.3K reputation)Distinguished Member (19.3K reputation)Distinguished Member (19.3K reputation)

Group: Forum Members
Last Active: 2 hours ago
Posts: 11.3K, Visits: 19.6K
I work similar, but I do the timing for my voices (such as pauses between sentences) externally using Audacity. I do this per scene, where a scene is usually not more than on minute of animation.

I export the various voice tracks from Audacity to iClone for lip-syncing and then create stereo tracks with ambiance to use in post-production, where everything comes together.

I export animation as image sequences, so the audio is a separate endeavor anyways and requires a post-production process.

Some people want to do everything within iClone, including sound, but I find that extremely limiting.


https://forum.reallusion.com/uploads/images/1a09220f-ab50-42ac-ad1a-33ec.pnghttps://forum.reallusion.com/Uploads/Images/d14339d0-cd32-4b35-88f9-40a0.png


justaviking
justaviking
Posted 4 Months Ago
View Quick Profile
Distinguished Member

Distinguished Member (15.3K reputation)Distinguished Member (15.3K reputation)Distinguished Member (15.3K reputation)Distinguished Member (15.3K reputation)Distinguished Member (15.3K reputation)Distinguished Member (15.3K reputation)Distinguished Member (15.3K reputation)Distinguished Member (15.3K reputation)Distinguished Member (15.3K reputation)

Group: Forum Members
Last Active: 4 hours ago
Posts: 7.5K, Visits: 23.0K
I support what Benjamin said.

(ADDED... and I support what Animagic said, too.  I got interrupted while typing, and we cross-posted.)


SHORT ANSWER:  I animate to the voice.


LONG ANSWER...  DIALOGUE...

I typically do some crude animation* to block out the basics of the scene before I add the voices.  Sometimes that's because I have some script refinement still taking place, and seeing the scene helps me finalize the writing.  Other times, more typically, it's a simple matter of when I have access to my voice actors.  If I've finished the script, I can start animating for a week or two and then get together with my voice actor(s) and record dialogue.  And sometimes I'm simply impatient and excited to start seeing some animation on the screen.

I do not have the actors "voice-over" existing finalized animation.  As an amateur animator, who does not use motion capture, it would be hard to animate with believable and natural timing.  My voice actors are also amateurs, and I believe it would be too difficult for them to try to "act" while simultaneously trying to match the timing of pre-existing animation.

While waiting for my "real" dialogue (which I still call "voice-over," even though it's not done "over" the animation), I might record some place-holder dialogue using the recording capability right in iClone.  It's a bit disturbing to have my insanely masculine voice coming from the female characters, but after a while one gets used to it, and it does help me at least get close to the right timing of an animation sequence.

When I record the official dialogue, using Audacity, I record as cleanly as possible, and still always apply a bit of Noise Reduction to clean it up even more.

Then I get crazy... 

I break the recorded dialogue into very short segments.  Short paragraphs, or even individual sentences.  So, yeah, I have a lot of audio files.  I usually have numbered the lines in my script, and use that to support a naming convention.  WHY SHORT SEGMENTS?  I import them into iClone (deleting any temporary dialogue and visemes first)., and now it is easy to have a character start taking a second later or half a second sooner because I can simply drag that segment.  Because of this, I love the "insert/delete frames" function on the Timeline.

I have also gotten in the habit of "replacing" the iClone audio in my video editing (NLE) software.  It's pretty easy to add the audio files into the NLE, align them with my iClone audio, and then mute/delete the audio that came from iClone.  WHY?  This gives me full control over the volume for each voice, and I can adjust the volume from one line to the next, and can fade it as a character gets farther away or turns his back to the camera.


LONG ANSWER...  Sound effects...

I will add a few key sound effects in my iClone project if I think it helps me with the creative/animation process.  For example, I might add a gunshot sound.  HOW?  I create a DUMMY object, add the sound to that dummy, and name it something like "Gunshot 1" or "Door squeak."  Like the dialogue, this can help me with the timing of the scene.  It also makes it more fun to animate when I have some sound.


LONG ANSWER...  Music, ambient, etc...

Unless I have a specific need (or emotional desire), I do not put music and other background sounds in the iClone file.  I do all that work in my NLE where I have much better tools (such as "unlimited tracks") to control the volume and stuff like that.  iClone's audio tools are adequate for animation, but not for finishing a good project.


CLOSING...

This is one area where I believe I've really evolved over the years.  With each project, I've grown more sensitive to "audio quality."  Even though I usually have insufficient background/ambient sounds a audio effects, I've gotten much better at the quality of the ones I do.

Good luck, have fun, and feel free to ask more questions.




*Crude animation - Even my final projects typically consist of "crude" animation, LOL.  But at least I have fun.




iClone 7... Character Creator... Substance Designer/Painter... Blender... Audacity...
Desktop (homebuilt) - Windows 7, i7-3770k CPU, GTX 1080 GPU (8GB), 16GB RAM, Asus P8Z77-V Pro motherboard, 500 GB SSD, terabytes of disk space, dual  monitors.
Laptop - Windows 10, MSI GS63VR STEALTH-252, 16GB RAM, GTX 1060 (6GB), 256GB SSD and 1TB HDD

Edited
4 Months Ago by justaviking
AutoDidact
AutoDidact
Posted 4 Months Ago
View Quick Profile
Distinguished Member

Distinguished Member (1.1K reputation)Distinguished Member (1.1K reputation)Distinguished Member (1.1K reputation)Distinguished Member (1.1K reputation)Distinguished Member (1.1K reputation)Distinguished Member (1.1K reputation)Distinguished Member (1.1K reputation)Distinguished Member (1.1K reputation)Distinguished Member (1.1K reputation)

Group: Forum Members
Last Active: 6 hours ago
Posts: 266, Visits: 1.3K
I think most of us are using very similar pipelines for Audio
in our animation projects.

I  have a dual OS operation ,Windows/Mac
I pre-record my voice audio (over on my mac usually) 
and drop the raw audio, on my windows server, as I have found a really good voice
 MORPHING tool called Morphvox pro.

It truly changes  recorded voices beyond simple pitching up and down.
I also use a (sadly Mac only ) program Called "Ghost reader"
it is a text to speech engine that has some of the most realistic
human voices I have found any where with realistic non-American Accents!!.


The Female Doctor in this clip below and the Triage Droid are 
both  Computer generated from
typed text files:



I use the raw audio to generate my lipsinc phonemes with Daz mimic live and
manually facial animation on top of that with the curve editor in Daz studio
as Daz has Much better facial expressions than Iclone.

However the final voice tracks are synched with the color graded,
post produced clips in Final cut pro along with  any foley and other sound effects.
 


---------------------------------------------------------------------------------------------------------------------
My name is Galactus,Devourer of worlds.
Look on my works ye mighty,..and despair!

Edited
4 Months Ago by AutoDidact
thedirector1974
thedirector1974
Posted 4 Months Ago
View Quick Profile
Distinguished Member

Distinguished Member (4.4K reputation)Distinguished Member (4.4K reputation)Distinguished Member (4.4K reputation)Distinguished Member (4.4K reputation)Distinguished Member (4.4K reputation)Distinguished Member (4.4K reputation)Distinguished Member (4.4K reputation)Distinguished Member (4.4K reputation)Distinguished Member (4.4K reputation)

Group: Forum Members
Last Active: Yesterday
Posts: 687, Visits: 4.0K
If you have voice actors, the usual way is to write the script and then let the voice actors record their lines. When you got the recorded lines, you should cut every sentence into a piece of it's own. So you have the possibility to choose your timing. After that you load each sentence in the scene and then you can animate everything. Every other sound effect will be added later, when you eidit you animtion with your editing software.
IO1
IO1
Posted 4 Months Ago
View Quick Profile
Distinguished Member

Distinguished Member (1.1K reputation)Distinguished Member (1.1K reputation)Distinguished Member (1.1K reputation)Distinguished Member (1.1K reputation)Distinguished Member (1.1K reputation)Distinguished Member (1.1K reputation)Distinguished Member (1.1K reputation)Distinguished Member (1.1K reputation)Distinguished Member (1.1K reputation)

Group: Forum Members
Last Active: Last Month
Posts: 76, Visits: 365
Thanks for so many amazing answers guys. Now i feel like i have enough info to takle this part of my project with confidence.
Tarampa Studios
Tarampa Studios
Posted 4 Months Ago
View Quick Profile
Distinguished Member

Distinguished Member (3.7K reputation)Distinguished Member (3.7K reputation)Distinguished Member (3.7K reputation)Distinguished Member (3.7K reputation)Distinguished Member (3.7K reputation)Distinguished Member (3.7K reputation)Distinguished Member (3.7K reputation)Distinguished Member (3.7K reputation)Distinguished Member (3.7K reputation)

Group: Forum Members
Last Active: 2 Weeks Ago
Posts: 473, Visits: 4.6K
I think it also depends how professional you aim to be, and if you wish to establish good habits from the start...

eg I admire directors who choreograph their scenes to music, so explosions and other sound effects fit more naturally and enhance the viewing experience..

So I start with a rough blocking of character actions to music, and with dialogue in the subtitles (because for the projects I sell it is a legal requirement to have internationally compliant sub-titles, and so I can see when dialogue or voice tracks need to be trimmed... youtube is also being pressured for all uploads to become compliant for the deaf community and I personally prefer to ensure the quality of my subtitles does not compromise the visual and sound experience... youtube auto subtitling is atrocious and most of the cheapest subtitling services are also terribly inaccurate, even when done for major studios!!!).... then I can also tweak the dialog so it fits the timing of the action and sound track... 

Note: using text to speech for the first draft means I can also record draft expressions but I always mute the computer voice, leaving expressions and subtitles...

Then I send a silent render of the scene with draft expressions and subtitles to the actors so they can see the emotional requirements and timing better so they can record their voices more effectively... but I make sure they know that lip syncs are only very rough and will be perfected later to suit their specific nuances... or perhaps facially mo-capped but thats another story, because I prefer another method over mo cap.... 

Knowing the wordcount for your voice tracks also helps me budget payments for voice tracks more accurately in accordance with standard industry rates for non-union actors... and as an indie I try to work with actors directly so I dont have the additional expenses of their agents... and also so I can quote and get agreement on cost of their voices, laughs and other utterances written into their talent release form BEFORE they record so there's no messy negotiations or arguments later. 

If you dont get those agreements locked down you can be disqualified from film festivals and makes it much harder to sell your work through many commercial distributors later... 
without the agreements locked down, you cant get paid as big an advance coz the distributor can charge legal expenses for locking down agreements ... and if voice actors argue with them you may be forced to do allthe extra work to replace them with a more compliant... or more famous... actor... but thats another story too...

Actors are also more likely to sign their agreements and return them up front (when they send in their sound tracks) than trying to chase them later. 

So it is my company policy to use local talent as much as possible, so I can also meet them at a local radio station or uni sound studio to record their voices professionally myself. I have credit acknowledement greements with both locations so the use of their pro sound recording facilities doesnt cost me anything. Maybe you have that option in your local area too?

Otherwise if the voice talent live too far away (my last series included voices of real scientists on location everywhere from antarctica to the flanks of live tropical volcanoes - which also means they could send me one minute of ambience recorded on location at no extra cost)....

Then upon recieving 3 takes of their voices, I cut-n-paste the best of each take in my editor... taking care to tweak the micro seconds between each word and clean up the audio...

Then render the audio-only voice tracks separately for each actor, load them into iClone characters and perfect each lip sync which usually also means repairing expressions...

Then I recompile it all in my editor, delete the iclone rendered sound track and restore all the separate tracks for voices, music, ambience and sfx so they can be polished better during final pass of sound engineering in a way that is also compliant with the needs of your distributor for cinema/tv/web distribution etc.

Im still on that steep learning curve myself, so for the highest paid jobs I send the final project to a pro sound engineer or lend it to my local uni for teachers to coach students on final stages of sound engineering.

At that stage even a famous sound pro can be quite cheap - eg final sound engineering for each 10 minutes of film, where you have taken care of most tracks yourself only costs about a tank full of diesel for a 4wd... in my experience at least. 

Perhaps if you want a checklist for final sound engineering it would be great to hear everyone's suggestions. eg I would start the checklist with:

*removing any hiss etc from sound that had to be recorded on location.
* avoiding or repairing echoes
* sound mastering to 0dB 
* trimming/cleaning the micro silences between each word
polishing music - especially at each scene transition...

but there are quite a few more that I am only just starting to learn myself.

Either way, good luck!! 






Anita, aka "Bleetz" the Toon Wrangler @ Tarampa Studios
SYS: Win 8.1,  64-bit, i7-4770 CPU @3.4GHz,  
32GB RAM, Intel(R)Graphics 4600 & nVIDIA GTX1060 TI 



Edited
4 Months Ago by Tarampa Studios
animagic
animagic
Posted 4 Months Ago
View Quick Profile
Distinguished Member

Distinguished Member (19.3K reputation)Distinguished Member (19.3K reputation)Distinguished Member (19.3K reputation)Distinguished Member (19.3K reputation)Distinguished Member (19.3K reputation)Distinguished Member (19.3K reputation)Distinguished Member (19.3K reputation)Distinguished Member (19.3K reputation)Distinguished Member (19.3K reputation)

Group: Forum Members
Last Active: 2 hours ago
Posts: 11.3K, Visits: 19.6K
Tarampa Studios' contribution jogged my memory.

For my last completed film, I also used TTS voices first and then showed the result to the actors, which I recorded in a local public access studio. It was not free but cheap enough. So sometimes it helps indeed for the talent to see a cut of the movie. I also like recording them face-to-face rather than doing it online.

So that is what I do in general; I do the initial version of the movie using TTS. It gives me an idea whether the dialog sounds right or needs rewording, for example.

Even though I work from a script for longer movies, I start creating scenes before everything is completely written down, just to stay inspired. This is one advantage of being independent and having no particular time constraints. No-one is waiting for the movie to be finished.

I recognized the male voice from Autodidact's clip. It is from Acapela, which is also available online at https://acapela-box.com/AcaBox/index.php. So maybe Apple licensed those voices. Their British voices are quite good.


https://forum.reallusion.com/uploads/images/1a09220f-ab50-42ac-ad1a-33ec.pnghttps://forum.reallusion.com/Uploads/Images/d14339d0-cd32-4b35-88f9-40a0.png


AutoDidact
AutoDidact
Posted 4 Months Ago
View Quick Profile
Distinguished Member

Distinguished Member (1.1K reputation)Distinguished Member (1.1K reputation)Distinguished Member (1.1K reputation)Distinguished Member (1.1K reputation)Distinguished Member (1.1K reputation)Distinguished Member (1.1K reputation)Distinguished Member (1.1K reputation)Distinguished Member (1.1K reputation)Distinguished Member (1.1K reputation)

Group: Forum Members
Last Active: 6 hours ago
Posts: 266, Visits: 1.3K
I recognized the male voice from Autodidact's clip.
 It is from Acapela, which is also available online at https://acapela-box.com/AcaB So maybe Apple licensed those voices. 
Their British voices are quite good.."


Wow, Good ears  sir!! 
You are correct , the  med-droid voice is "Graham" from Acapela.
and the British female Doctor  is "Lucy"

I have them all: the British, German,Arabic and Spanish accents

The, Mac only, Ghost reader software has a really good TTS engine
although some words have to be spelled phoneticly to get the right pronunciation

Voice acting is typically the "show stopper" for us one man operations
trying to make films with multiple actors of both genders.

But with Morphvox, Ghost reader,and my own acquired skills at North american
and European accents (learned from youtube tutes), I am making it work for My current project
now 64 minutes long.



---------------------------------------------------------------------------------------------------------------------
My name is Galactus,Devourer of worlds.
Look on my works ye mighty,..and despair!

Edited
4 Months Ago by AutoDidact



Reading This Topic