First We are in agreement that their
documentation for Anilip is abysmal
(and I am being charitable here)
Tell me, with 2, after you animate the lipsync, just what buttons do you press to get the track onto the timeline?
Second I personally do not load text versions of my dialogue just the .wav audio file generated by my AI voice service and the base lipsinc is generated
Now the reason you are not seeing new keys in the Daz Studio timeline is because all of the animated parameters of a Daz figure are not visible by default so you need to access parameter types pulldown and check every box so it reads “TROAH”
(see red circled area of my attached pic) Right click open in new tab for full size
You should then see the viseme keyframes of the lipsinc.
I have over 40 years in the electronics manufacturing business, and some of that was a tech writer. I would get fired for doing what they’ve done. The software itself is a Beta version, not suitable for public release. The QA people wouldn’t let this out the door.
TROAH! Wolf I wouldn’t live long enough to find that out! HA! Thank you!
My voice recorder will arrive tomorrow. Then I’ll get a chance to see if I can trick Mimic with it. If yes, I prefer Mimic. If not (growling in advance), I know Anilip will take an audio file as input. Thank you Wolf.
Well Wolf, I figured out how to use a recorded file to lipsync with Mimic Live. I’m pretty pleased with the result. See what you think. https://www.youtube.com/watch?v=Zn_dtSfql3k
Actually, no. The Sound Blaster X3 preamp allows the mic input source to be specified
as “What You Hear.” This audio was scraped off of a YT broadcast in real time. I was amazed at how well it worked. Next I’m going to try MicMonster with it.
Indeed there are many TTS option available now
both free and paid. I used a paid service for my animated fan film based on the ghost in the shell franchise over a year ago
But these days I mostly use one of many free options online.
I notice the head and eye movements as your characters speak. Did you put that in yourself? Mimic adds that. I find it makes them less stiff. IRL the voice is so much more fluid, more dynamic, than any service I’ve seen.
The TTS options are a little better now than they were a few years ago. Still, I haven’t found one that lets me put the emotion in the voice. I was hoping SSML would change that, but it hasn’t done much. I’ve noticed many services want to charge me just for LISTENING TO IT rather than for actually downloading it, much less actually using it. I stay away from those. MicMonster does give me “unlimited characters,” as they call it, so I’ll get started with them. Most would charge you for each breath you take if they could. I will have to keep looking.
I notice the head and eye movements as your characters speak. Did you put that in yourself? Mimic adds that.
Yes I always layer on a manual face puppet pass in Iclone
to randomize the facial performance
I find that with long speeches the Auto head movement
of mimic is a bit too linear and begins to look
pre-programmed and a bit “robotic” for my taste.
MicMonster does give me “unlimited characters,” as they call it, so I’ll get started with them. Most would charge you for each breath you take if they could. I will have to keep looking.
For me it’s a matter practical logistic as well as economics.
I made my 50 minute “Ghost in the shell” fan film with Iclone in 74 days working completely alone .
And I was at the last few months of an annual AI voice service plan I had purchased for $50 USD earlier that year.
If I were commissioned animated some big animated production with a budget today, I would hire voice actors.
But for my personal projects for youtube it makes no sense
so I will just use the free ones.
At one point I suggested, in a forum, that artists form a voiceover coop - I give an hour for an hour. It seemed like an obvious idea. Nobody was interested. Nobody. If Grok could talk, he would be perfect at voiceover.
It seemed like an obvious idea. Nobody was interested.
To be completely honest I could see alot of potential problems with a Co-Op.
such as people being unwilling to voice some dialogue that conflict with their personal/political world view etc
If Grok could talk, he would be perfect at voiceover.
ChatGPT can talk with it’s “read aloud” feature.
it has several built in voices.
you just tell chatGPT to repeat the following text
and past your dialogue.
no audio downloads but easily handled by screen recording the playback recitation with OBS or similar and extracting the audio with your video editor.
Every Character in this animated series ,that I am doing for my young grandsons, is a FREE chatGPT voice reading my script out loud
BTW, Whatever you do NEVER update Daz studio beyond version 4.23 if you still want to use any version of mimic.
I’ve never been able to see through the eyes of other people.
I’m going to examine 11labs today, having “spoken” with Grok about it just now. Best voice quality, lowest price, easily avoidable auto-billing, worst UI available. Ever tried it?
Actually it seems I still have 10,000 free “Credits” left over at Eleven Labs but IIRC the voices are excellent although I agree about the poor interface design.
I studied Jacob Nielsen and Steven Krug for usability guidelines. They made so much sense, still do to me. I expected everyone would read the book. Instead, everyone forgot how to read. Usability audits don’t take place either. We are in a world of “press every button and see what happens.” Eventually that works for some people, although it does take longer
The voice is the last brick in the wall. Sort of. I feel I am near the answer. The script will be next. It’s always last with me.
IMO, if you put something together, say, a minute or two of speaking characters in a scene, you’re not just an artist - you’re a director. Casting, script, wardrobe, lighting, cameras, sets, you do it all.
Agreed. I was able to salvage much from my croaked machine, including Mimic. Never “clean up” your Downloads Dir. Time is moving faster than we think.
I can live with the ElevenLabs interface and it has been updated over time.
What ElevenLabs offers, as do other services, is cloning a voice. I cloned my wife’s voice and I was quite impressed. I could even have her speak Dutch, even though she doesn’t know it.
With voice actors, especially when you work remote, you have to deal with different recording environments. I’ve listened to demo reels and I was appalled by the low quality of some. And why can’t anybody just speak normally?
In my case also, the dialog often changes when I watch a scene for the first time. I think by myself, “that doesn’t sound right”, and then I alter the dialog. That would be difficult with voice actors.
For instances where I want a certain strong emotion that I can’t get otherwise I use a program called Revoice Pro that lets you impress the intonation of one voice on another.
ElevenLabs has a good reputation for voice quality, and (nowdays) as professional a UI as will be found. I’m signed up as a “free” user, and if I need more characters, it’s affordable. Their attitude is good - there’s no time limit on evaluating the service before paying up. Of course I regret the passing of traditional desktop software, but we got what we got.
You’re right, working with voice actors would have its disadvantages. It’s easy to forget that.
I asked Grok if he could speak. He said yes he could, but only for a pile of money. Too bad. I wonder what Grok would be capable of.
