an alternative to voice actors, Silent Movie Title Writing
Mistara
Posts: 38,675
was just reading this and realized this could be the answer to not having voice actors
in the silent movie era, the Oscars gave an award for
Best Title Writing
The early 20th century was referred to as the 'Silent Era of the Motion Picture Industry'. 'Oscar for Best Title Writing' was awarded by the Academy at the end of this era. The award was given only once, in the year 1928. Before the advent of "talkies", dialogue was dependent on the titles that appeared between scenes. Their importance was acknowledged at the first Academy Awards, which incorporated a separate category for 'Best Title Writing'. With the next award ceremony, sound started dominating the industry and the award for 'Best Title Writing' was eliminated.
Daz 3D is part of 
Connect
DAZ Productions, Inc.
224 S 200 W, Suite #250
Salt Lake City, UT 84101
Licensing Agreement | Terms of Service | Privacy Policy | EULA
© 2024 Daz Productions Inc. All Rights Reserved.
Comments
Here is another alternative. Much cheaper than hiring someone - useful for more dead pan parts.
https://murf.ai/
Also I've been able to rescue many a bad actor by playing with syllables (taking them up or down a pitch), to get the right inflection.
thanks. looks interesting.
how would you age a voice? Dumbledore, McGonagal.
I hadn't thought of that. There are 100 voices to choose from. I'm hoping some of them represent older people.
It's currently mainly used for presentations, but as long as your actor doesn't need to emote all over the place, I reckon its a real step up from those google / mircosoft offerings from a few years back.
Looks tempting, especially the $9 for 30 minutes one time deal. That might cover a 48 Hour Contest five minute animation.
Rescuing a bad actor reminds me of a favorite scene fom "Hail, Caesar!" The first clip establishes the normal role of a "duster", a cowboy actor with few if any lines. The production company has a personnel problem and so puts him in a sophisticated role with Shakespearean dialogue, leaving it to the director to make it work. I don't think playing with pitch syllables will help here.
@Steve K ha ha - no that would take an editing genius to sort that lot out!
Yes I thought that as well $9 for 30 mins - a great deal. Now on the subject of accents I have been a long time fan of spelling words as you want your actors (machine or real) to pronounce them. So for example If I want someone to imitate a posh english accent (the queen etc) then "quite" becomes "qwhiete" etc so that way you enhance the chance you'll get what you want. This works well with these "dumb" text to speech engines. Plus sometimes it's very funny to hear what a hash the algorithm makes of it all.
What the "Hail, Caesar!" director finally did was change the line to just a word or two. I don't remember exactly, something like "It's nothing", with maybe a sad shake of the head. Or maybe just the head shake ...
Sounds reasonable. If in doubt - shorten it.
In the Ray Harryhausen classic, Jason and the Argonauts, they ended up dubbing the star's voice for the whole movie.
"Todd Armstrong makes for a solid action hero as Jason, but his voice has been dubbed by British actor Tim Turner best known for providing the voice of the title character in the TV series The Invisible Man (1958–59)"
http://www.wearemoviegeeks.com/2013/12/st-louisan-todd-armstrong-starred-jason-argonauts-1963/
Thanks for the voice-related resources.
Yup. Stanley Kubrick was known for just removing dialogue altogether, saying the actor could convey the thought with expressions, body language, etc. In "2001: A Space Odyssey", " ... there is no dialogue in the first 25 and last 23 minutes of the film (excluding the end credits). Noted for its minimal verbal sound, there are around 88 dialogue-free minutes in the movie." #4 on Roger Ebert's Greatest Films Of All Time.
@Diomede - Interesting - and not easy to pull off (without it looking false). re: resources your welcome.
@Steve K - My first animated series was like that as well. A deliberate choice to cut down on the dependancy of others and my own work load. The trouble is it makes for a bare script. I'm going to look at other similar scripts that made it to see how I can rewrite mine.
Thanks this very interesting link.
I think I will give a it a try, at least with the free version to start.
@Philemo_Carrara Bo worries. Would you be willing to come back and tell us how you got on?
I shared the link as I'm constantly trying to keep up with developments, but I've not tried it yet (apart from the demos), as my own projects are not at that stage at the moment.
However it does seem to be a step forward from other text to speech emulators I've used. So an exciting alternative.
I will report back when the time comes.
Behind the scenes of Howl's moving castle English dubbing
As I said, I try to avoid dialogue. But if you're going to do it, do it right:
@Steve K - Epik acting. That's one long dialogue. imagine animating that lot!
re: dialogue - I also try to avoid. In fact with time pressure always on when animating, I often find ways of cutting the dialogue down to the bare essentials, just to avoid getting bogged down with it.
However for script writing, it's a lot easier to time a dialogue driven script. Then 10 pages of script is 10 minutes. At this stage of my development I'm not really sure how to time action only. I'm writing in short paragraphs to pad it out to the approximate timing I had in mind. What do you do? Or do you not worry about the length of the final script?
The last script I wrote was a decade ago for a live action 48 Hour Film Contest team (as I mentioned, I was basically the typist with the screenplay formatting software, everybody pitched in on the script). For animation, I generally just go straight to scene creation, with some kind of story in mind (often inspired by something I've seen - The Twilight Zone e.g. - or read - Calvin & Hobbes, e.g.) but nothing written down. For a five minute +/- video, I can keep it all in my head usually, but it does tend to change direction as I review the early takes and see new angles for both camera and story. Singer/songwriter Todd Snider once said a song started out about a certain incident, but after he got it all to rhyme, it wasn't about that anymore.
Not having to communicate with a team simplifies things, but I do lose sources of inspiration from others. I have asked others for input early on, but they usually come up with things I can't do very easily (typcially lots of ... you guessed it ... dialogue). I'm pretty familiar with my collection of elements (props, scenes, characters, sound effects, music, special effects) so I know right off what I can do. And again, for me its just for fun, like Calvin building his little models of cities. Then destroying them with various disasters.
handy link to the yt audio library. https://studio.youtube.com/channel/UCt5xaeQp8LPggXwhRX8O23w/music
the windows10 movie maker comes with handy title writing cards and fonts.
a silly lil clip
I used Talkia for this one - both Dart's and Rosie's voices:
Like Sci Fi Funk said about murf, Talkia has many different voices from various parts of the world.
We can translate to and from different languages too, like use a UK voice to speak Chinese, for example, or a Russian voice to speak English, etc.,
To use it on my main actors, however, I feel that I'd have to get better at manipulating the timbre - even within the same word at times. Also missing is the ability to shout or talk agrily, etc.,
Edit: It does have a few kids voices too - not sure about elderly. Might have to use a mature voice and run it through a guitar distortion pedal or something? :)
Make your own in Carrara!!! ;)
I've just finished a survey of available TTS resources. A very few are surprisingly good. Most are poor. None are undetectable. None are good enough for the movie I hope to make.
More than once I've suggested here that DAZ artists collaborate by trading out voice services. You can use mine if I can use yours. I made sense, but I never got one single interested reply. Not one. I'm still interested but not hopeful. I do believe remote collaboration by independent artists is, as we say, the future.
At this point I'm going to use my own voice, learn a couple of impersonations, and when I have to, alter the pitch to make a woman's voice. If I can do 5 characters, I can do a film.
..................................................................................................
I've tried using pitch-change and even voice-changing software to try and sound female, but even though I can do many voices, I simply cannot sound female except for perhaps one that sounds very masculine.
Your idea sounds great, and it's too bad that nobody attended your inquiry. Bummer!
When I record myself for my tutorials and demonstrations, I get a hum that's difficult to remove without making me sound rather odd. It doesn't really matter much to me for tutorials.
I am blessed with the fact that, my singer is also a sound engineer with a rather nice studio setup, so my wife and I will go and get recorded when the time comes. Finding that time is the hard part.
My singer has agreed to play a part, and also has a contact list of voice actors that I can hire if I need them. I can do many voices, so I doubt I'll hire out, but the option is there. For really serious projects, I've heard that voice actors can be hired over the internet, and have seen that we can download free voices from some places as well.
Perhaps there's a Voice Actor's Forum/Guild somewhere online that someone could quite possibly find some volunteers for free - just so that they can get practical experience, etc.,
Until I get to the point where I can get us into the studio to record, I also kicked around the idea of just being a narrator for some of the earlier, preliminary test work.
Most of it, however, is simply the fact that I keep failing to get the script written. That's all I really need and Rosie's acting skills will make her character come to life - truly. She's amazing, and I'm blessed to have her!
I had no idea she was so good at it. One day I handed her a paper with some lines scribbled onto it. She glanced at it for a second as I tried to get the words out of my mouth, asking her if she would be willing to voice act, and before I got the words out of my mouth, she was in-character and fully expressive, doing the lines I wrote without looking at them, and altered to how her character would have said it, rather than how I wrote it - I was So Floored by her performance... first time ever doing anything like that!
For crowds and public places ambient voice chatter, I'll get my band members into the studio for that. I'm in a five-piece rock band. With all of us, their wives and mine, we could easily record some ambient chatter and crowd noise clips that could then be layered over one another for even larger crowds, etc.,
I find that the less you change pitch, the better it works. So I raise the pitch of my voice a little, then raise it a little in Audacity, and I get away with it. It's not perfect but I'll use it.
Your hum may be a ground loop. Isolation will fix it.
Sounds like you lucked out with Rosie. I'm sure you give her encouragement.
Me, one day I made an announcement on the PA system at Home Depot, and afterwards a coworker came up to me and said "Man, you've got a great voice." I'd never been told that before, so I enrolled at the New England Actors Theater (note the acronym) for voiceover. I got the basics down, but I walked away from it when I was told you had to report for auditions the instant you were contacted. I couldn't risk my day job. That was then, now I have all the time in the world.
I built a studio in my house. I can get decent sound out of it. I'm just learning to use the mocap plugins, and I'm working on my voice at the same time. DAZ hasn't been an animation-rich environment in the past, but I'm expecting mocap will change that. That means voice will become more important in this forum. Mocap is a big deal.
The kind of collaboration I hope to see soon is already in the works. Reallusion made this video to call attention to it. The video makes it look like a new idea, but it's been coming a long time.
............................
Too true, unfortunately.
I like Talkia, and agree with them that tutorials and advertisements, etc., are Much more enjoyable if we can 'hear' along instead of realying on reading text. So for that, I think Voice AI is a vast improvement to nothing.
While TTS has come a l o n g way, it's still not there.
Mindsong and I stumbled into Talkia at a time when they were offering lifetime license for a small fee, so we each took it. The license includes free updates as well - which was the big attraction for me. I coulld already kind of tell, just by their own well done promotion, that it would need improvement before I could use it for a movie.
But where it does excell for CG animtion filmmakers is for practicing lip synch! Just type, convert, run through Audacity and into Mimic we go!!! ;)
Right. There have been voice acting collaboration cummunity attempts going for a long time, and so some voice acting cooperatives are out there, I guess. I haven't looked because I don't have my dialog prepared yet. When I do, we're just going to record it. But the thought has been there many times to just find voice actors online. I know they're out there.
I saw the Adobe Voco presentation when it was new, that was the best I've heard to this day. The second best was Lyrebird. I prefer the voice-scan approach. Lately I took a look at Replica Studios. Ho Hum. In most of them I can hear burbling between the words. I will not accept the burbling in my own work. Rentware is another issue. I will use Mimic for cats and dogs because I can't mocap them, but I am so blown away by Face Cap and Face Mojo that Mimic won't see much of me from now on.
Certain voiceover actors are voice-impersonation specialists. That guy who does Mo the Bartender is an example. He does about four of the characters. He does Krusty the Clown and that 7-11 guy and another one - is it Sideshow Bob? I'm going to give that a try. Most of the time I sound like a professor (which I am not), but I've been working casually on a country Gramma and Grampa. Today Grampa did "The Naming of Cats." There's nothing like hearing TS Elliot read with a Beverly Hillbillies accent. :-) I would like to do a German accent, not a modern cosmopolitan one but an exaggerated imperious one for a character I have in mind, a Doberman Pinscher. I was surprised to find that YT is loaded with "talk in an accent" vids.
I'm thinking of collaboration beyond a voice-sharing group. I'm thinking that Hollywood's days are numbered. They will be replaced by people in their little home offices just like mine, all around the world. I believe we'll end up with a collection of disciplines represented each by an individual, rather than a collection of one-man bands. Directors, script writers, voice actors, sets, lighting, character creation, etc. We will meet often via telecommunication software.
.............................................
The intro anouncer actor (also plays yoda and several others) in the video I posted above moved out of CA before they decided to finish the series. Since coming back wasn't an option, they sent him the gear they wanted him to record with and do it from home - with the director on vid-chat.
The host of the video, James Arnold Taylor, does many voices in the series, but also in much of what we see on TV. He even did Michael J Fox's voice overs in Back to the Future, when he was unavailable.
I was actually looking for Dee Bradley Baker's channel when I found that video. Dee teaches people to become voice actors - and I think it's free.
As for Mimic, Mindsong and I have been experimenting with it quite a bit. I think it has a lot of potential - especially since I enjoy making micro expressions by hand after ading the Mimic results. But making custom DMC files, and then tweaking them for certain scenes is a lot of fun for me. To each their own.
Mimic does one thing that mocap doesn't allow you to do. It allows you to lipsync from a prerecorded audio source. This means you can tweak the audio source without re-recording the entire performance. It's an ability I would like to see in the facial mocap software. Just redo the lips from a recording, leave all the other stuff alone. I wonder how well that would work. I'd like to try it. The longer the recording, the more likely I will mess something up.
I first saw videophone demonstrated at the 1964 World's Fair in White Plains. The crowd was blown away by the demonstration, the waiting line was long. I waited until 2000 when we finally had the bandwidth to make it practical. And what do I find? People are afraid of cameras. It's the only explanation I can come up with for why we are still typing in little boxes, and using emoji to substitute for facial expressions. SO much communication information is in the face. Perhaps today's social restrictions will help us get over our unease around cameras.
Be seeing you...
I'm working on a project mixing face mocap and mimic like lip sync program. I don't use mimic because mimic understand mainly English phonemes and ranges from not very good up to very very bad in other languages. As I need my script to understand French, I'm in the process of implementing voice recognition VOSK python library with my OpenFace implementation to get a mix of both.
It's a nice project (in the chineese sens) I'm doing on the side, so don't hold your breath for it, but i've already established feasability.
Very cool, Philemo! Looking forward to that!
One of the things I love about using Mimic (but I speak English) is how we can customize the results visually after it has done its computing.
In watching people talk, I'll notice that the lip and mouth movements don't always actually look like they're saying what they're saying - like the facial movements are far less exaggerated than, say, iClone's facial mocap or Mimic would turn out by default.
Mimic gives us a computed result. But then we can, for example here - replace an 'eh' phoneme with the 'Iy' if it makes a better shape for what we're looking for. And if we find ourselves doing this a lot for a particular character, we can make a custom DMC file for that character built on the morphs we actually want it to use for any of the given visemes.
By default, the basics are the basics and Mimic does a pretty nice job of simulating a fairly decent starting point. But with some custom tweaking, we can really change those results right from the first run through.
Something to keep in mind:
For Hollywood animated lip sync they employ animation artists to handle nothing other that the lip sync - and while they likely always begin with a Mimic-like software emulation, Good lip sync always has and likely always will require the eye of an artists to tweak the automated result rather drastically to get the desired performance.
This film (Alita Battle Angel) really blew my mind. I bought the Digital Deluxe version so I have access to All of the extras that would come with it - and it comes with a lot. Well in that, I've seen that while they use two HD cameras to capture the actress's facial performance, the animation artist would still often go back and alter the CG facial poses frame by frame - which is pretty intense for something this huge - and it works very well.
This is but a fraction of such footage that I've found on YouTube
(but it captures the essence of what I'm talking about, I think)
The thing that I love about seeing these philosophies coming from someone like Weta Digital and LightStorm is that, without even getting to the lip sync phase yet, I've always had the belief that this must be done to return a convincing animated speech.