Mimic Pro for Carrara

Joseph_T

Does anyone know if there is a time limit or file size limit when uploading a 16bit wav file in Carrara with Mimic Pro?

Regards

Dartanbeck

I don't think so. Here's a link to the User's Manual, if you want to check it out.

In good practices, however, I would try to break long recordings down to shorter files. Most of the time this stuff doesn't have to ALL be lip synced, keeping some of the dialog (or whatever) being done 'off camera' as the scene switches to another view. But there are always those times when we want to push those limits and just go for it, right? I think it can handle quite a bit, but it may take some time if it's really long. 

Joseph_T

Thanks, it's just that I uploaded a 3 minute wav file and got an error, but it accepted 2:30.

Dartanbeck

Again... that's waaaay long for a normal lip sync session. You can do it - so it seems. It might be difficult to keep from boring your audience over two and a half minutes of a non-changing view. And if it does change, it makes more sense to only load in the length you need.

But you can do it!

Here's more on this idea:

I'm certainly not trying to bash using long files like that or creating long camera shots. But I am trying to warn you away from doing it. Go ahead and try this one and you'll likely see what I mean here. Beyond just using a plugin's algorhythms to auto-generate results for something, that should really be considered a time-saving tool as opposed to being the end-of-the-line finished result. There's always tweaking to do. Optimizing six to ten seconds of quality lip sync from sound file to figure can take a good amount of time. Mimic gets us close, but there's often extra stuff to add or tweak. Even as quick of a clip as a three second animation can reveal (especially after rendering) that a lot of additional motions are needed to get a natural looking result. Two and a half minutes will compound that enormously.

But then again, I don't know what you're doing. So if the Mimic translation turns out great for what you need, and you somehow have enough time to render such a monster... go for it!

argus1000

I try to have my scenes stay under 30 seconds. Especially the lip-sync scenes.It seems to me easier to handle.

Joseph_T

Thanks guys, I cut the track into 5 segments and am having a much easier time with it. Great movies....after my misserable attempts to do a video, I can really appreciate the work you put into those.

Dartanbeck

I love this mimic-in-Carrara test by Freelight Motion

 

Dartanbeck

Here's a nice tutorial by our very own SciFi Funk!

Part 1

Part 2

...and there's more. Here's his thread on the subject: * How to use Mimic Pro for Carrara effectively Tutorials - Sci Fi Funk

 

Joe Pingleton has done a good bit of experimenting with Mimic Pro in Carrara, logging his progress here, in this forum:

* Cararra and Mimic Pro Experiment

...and you may also be interested in * Mimic - DMC files for Genesis 2

WendyLuvsCatz

I have done a very long one

 

the bit with me talking is Carrara because the wav file was in fact too long for iClone!

but yes there is a limit, I actually thought it was 5 mins but I use 24fps

Dartanbeck

I agree with the 3d you, Wendy. So many people just need to revisit preschool and learn how to behave.

This was a great example, because I was just about to comment howe I prefer to keep my renders to less than ten seconds of animation. Most are less than six and many are closer to three, and even those get parts cut in the final product. However, that was really cool how your 3d you could just talk away without having the scene changing. I guess there are all sorts of different ways of doing things. 

I might have Mimic apply a short audio file slightly ove a second long, but the render might still go on for another second or two... or three or more. This is because in the end there'll be more dialog added in the mix - dialog whose actor's face is not in camera at the moment. It's important (in my belief from reading a lot of stuff from various places about this sort of thing) to not have all dialog being actually seen being spoken. But I'm into doing my thing... I need to remember that there are so many other sorts of projects going on that have nothing to do with what I'm studying! LOL

WendyLuvsCatz

prob longest one 

That Other Persona

Joseph_T said:

Thanks, it's just that I uploaded a 3 minute wav file and got an error, but it accepted 2:30.

You could go for half-half with some overlap and then combine the two movies outside of Carrara.

Many videos have an underlying audio but with a variety of camera shots.  Each camera shot gives you an opportunity to make a new, short clip.  Shorter clips are easier to manage, output, change, etc.  And it adds a little spice to the overall movie.

Dartanbeck

Oh absolutely. For example, I never use an audio channel from a render, except for aligning purposes in Sony Movie HD, where I mix the sound, dialog, music, overlay effects and transitioning to the multitude of clips that I use to cut the final product. I love that part! Still totally green at it? Yeah. But I love it! That's actually what all of this Carrara/Howler stuff is about for me.

Joseph_T

https://www.youtube.com/watch?v=aVlatSdjhGc

 

I could use your help. I did this as you suggested in 20 to 30 second segments, which worked out great. I'm using Sony Movie Studio and put fades or a page turn to connect the segments. This is my question...can you start the figure in the same postion that the previous segment ended with. I tried using NLA clips, but they revert back to the original position for each segment. What I want to do is start the next segment where the previous one ended. Can this be done with the Motion Tab?

 

Regards, Joseph.

3DAGE

When you create an NLA clip,. You have options to select a "loop offset" which is normally a bone (hip) which will be used to match the looping of the clip

If you missed creating that Loop offset,... don't worry.

You can Load the NLA clip "Data" back onto the timeline as Keyframes,. (NLA tab / Load clip data)

then you can create a new master clip, and specify a bone as the loop offset.

 

Persona Non Grata

.

mindsong

screamingbee.com and audio4fun.com both have voice 'morphers' that can re-tune your voice to that of other target styles (male to female or child, etc.) fairly well, especially if you 'help' by 'adjusting' your voice toward the target sound as best you can while recording (e.g. try to sound young if your target character is going to be morphed to a young voice.) screamingbee never goes on sale, but audio4fun's Voice Changer series will periodically do sales.

They will often have a slight 'synth' sound if you stretch too far from your starting voice, but otherwise work suprisingly well. Screaming Bee also has a slick plugin that will let you take a long voice recording (e.g. of a single person 'acting' all parts of a script, in one read), and let you select and assign sections of the recording to various preset characters that you can create, and it will render the final into a new file. Brilliant for animators, previz story-boarding, and books-on-tape type presentations.

I have the Carrara mimic plugin, but I still use the standalone mimic pro (can still be legit bought from toolfarm.com, ~$50, use your daz3d email and it will appear in your DAZ product library) because I know it and have a workflow... Its controls are pretty low-level and can be made to accomplish some pretty decent lip-sync. If the Carrara plugin counter-part has that same level of control, it should be pretty capable. Someday I'll play...

ETA: that deepmind link has some interesting technologies. It doesn't look like their TTS is ready for production, but perhaps soon, esp. with google's (is evil) backing.

cheers,

--ms