AI Seeps Into The 48 Hour Film Contest

Steve KSteve K Posts: 3,192
edited October 2023 in Carrara Discussion

The Houston 48 Hour Contest is this weekend, I'm doing a five minute animation.  In Carrara, of course.  I noticed a new rule this year about AI:

QUOTE

Using AI (Artificial Intelligence) in Your 48 Entry
You may use AI (Artificial Intelligence) in the creation of your film, subject to the following guidelines:

You may only use AI in ways that are legal in your jurisdiction.
You may not use AI to create any part of your entry prior to the Official Time Period.
You must have the rights to all work including work created with AI. In other words, any AI generated material that uses copyrighted material for which the filmmaker does not have a release is not permitted. (As an example, using a screenshot from the movie Pulp Fiction to create a new image is prohibited.)
Any footage created using AI is subject to both: copyright requirements and Rule D regarding footage (above). That is, not only must the filmmaker have rights to the footage, but any footage shot prior to the competition weekend, may only be used as outlined in Rule D.

END QUOTE

Rule D is the standard rule about all footage must be created during the contest, etc. 

Just FYI, I did not use AI.

Post edited by Steve K on

Comments

  • WendyLuvsCatzWendyLuvsCatz Posts: 37,858
    edited October 2023

    AI is a very broad medium

    PikaLabs, RunwayML, HeyGen, D-ID type animation is a quite different thing to the very random Deforum Stable Diffusion  videos I generally make, they use AI to animate images which can be your family photos, drawings etc as well as diffusion generated

    I think it could well be the future of CGI animation

    Post edited by WendyLuvsCatz on
  • DartanbeckDartanbeck Posts: 21,227

    It certainly is "A" future for CG animation.

    In watching "Light & Magic" on Disney + (excellent series, BTW) I've seen the sad 'fall of the miniature modelers' department due to the new (then) popularity and possibilities of CG.

    Nowadays sees an entire change-over, however, in how Both can make a much better movie. So the Miniature (and full size) Modeling department is back in full swing, once again.

  • Steve KSteve K Posts: 3,192

    WendyLuvsCatz said:

    AI is a very broad medium

    PikaLabs, RunwayML, HeyGen, D-ID type animation is a quite different thing to the very random Deforum Stable Diffusion  videos I generally make, they use AI to animate images which can be your family photos, drawings etc as well as diffusion generated

    I think it could well be the future of CGI animation

    I'm pretty lost here.  I'm just hoping for a good text-to-speech feature (to avoid voice actors), and it looks like the new version of my Vegas video editor has one.  Lots of voices, word emphasis options, expression levels (excited, whispering. etc.), etc.  

  • WendyLuvsCatzWendyLuvsCatz Posts: 37,858

    Steve K said:

    WendyLuvsCatz said:

    AI is a very broad medium

    PikaLabs, RunwayML, HeyGen, D-ID type animation is a quite different thing to the very random Deforum Stable Diffusion  videos I generally make, they use AI to animate images which can be your family photos, drawings etc as well as diffusion generated

    I think it could well be the future of CGI animation

    I'm pretty lost here.  I'm just hoping for a good text-to-speech feature (to avoid voice actors), and it looks like the new version of my Vegas video editor has one.  Lots of voices, word emphasis options, expression levels (excited, whispering. etc.), etc.  

    as I suggested to others ElevenLabs is quite good and has some free voice generation 

  • Steve KSteve K Posts: 3,192

    WendyLuvsCatz said:

    Steve K said:

    WendyLuvsCatz said:

    AI is a very broad medium

    PikaLabs, RunwayML, HeyGen, D-ID type animation is a quite different thing to the very random Deforum Stable Diffusion  videos I generally make, they use AI to animate images which can be your family photos, drawings etc as well as diffusion generated

    I think it could well be the future of CGI animation

    I'm pretty lost here.  I'm just hoping for a good text-to-speech feature (to avoid voice actors), and it looks like the new version of my Vegas video editor has one.  Lots of voices, word emphasis options, expression levels (excited, whispering. etc.), etc.  

    as I suggested to others ElevenLabs is quite good and has some free voice generation 

    OK, thanks.  Free is always good, but the new Vegas text-to-voice feature looks VERY feature filled.  The upgrade will cost me $150, but has many other new features also.  I'm a couple of verisons back, anyway ...

  • DartanbeckDartanbeck Posts: 21,227

    ...and if you use Vegas anyway, that's probably going to be a big Win. I imagine Magix has done a great job with it to include it in Vegas. Yet there's always that chance that it isn't that great - makes one stir a bit.

    Walt Sterdan, the creator of the Hrimfaxi series (Excellent!!!) has been using PlayHT after not being able to find all of the same voice actors. It has "Ultra-Realistic" voices options, which is what he uses. They're fantastic. I have Talkia, which is pretty cool but has that "I am an artificial voice" sound to it no matter how hard you try - especially if one is trying to actually use it for dialogue. I mean... they call them "Actors" for really good reasons!

    Here's a little discussion we have going on: Mimic: Who owns what?

    Hrimfaxi Adventures Episodes 1 and 2 (ends at Ep 2) use real voice actors and a genius combination of Mimic Pro results followed up with puppeteering to add other facial movements

    Starship Hrimfaxi episodes use the Ultra-Realistic PlayHT voices and 2D animation software

    It's all fantastic. If I made money doing this stuff, I'd certainly invest in PlayHT's Ulra-Realism!

  • Steve KSteve K Posts: 3,192

    Dartanbeck said:

    ...and if you use Vegas anyway, that's probably going to be a big Win. I imagine Magix has done a great job with it to include it in Vegas. Yet there's always that chance that it isn't that great - makes one stir a bit.

    I do use Vegas, having switched from Magix' Movie Edit Pro some years ago, mostly to get the Particle Illusion plugin.  I've gotten used to Vegas and just finished a 48 Hour video with it, pretty much second nature by now.  Here is a video on the new text to speech feature:

  • UnifiedBrainUnifiedBrain Posts: 3,588
    edited October 2023

    Steve K said:

    Dartanbeck said:

    ...and if you use Vegas anyway, that's probably going to be a big Win. I imagine Magix has done a great job with it to include it in Vegas. Yet there's always that chance that it isn't that great - makes one stir a bit.

    I do use Vegas, having switched from Magix' Movie Edit Pro some years ago, mostly to get the Particle Illusion plugin.  I've gotten used to Vegas and just finished a 48 Hour video with it, pretty much second nature by now.  Here is a video on the new text to speech feature:

     

    Vegas is apparently using the same text-to-speech engine as Microsoft does in their Clipchamp product.  With Clipchamp, you can use it for free.  There are three caveats.  One, you need to use either Edge or Chrome as a browser.  Two, it only outputs as an MP4, so you may need to strip the video component (which is not hard).  Three, some of the voices do not have the full range of expressions.

    I've used it, and the voices are very good.  It is slightly easier for me to get the right vocal inflection with Clipchamp, as opposed to Elevenlabs (which is also very good).  And unlike Elevenlabs, Clipchamp has no monthly character limits.  It does have a duration limit for each clip.

    As a 3D artist, I am against using AI in general in the creative process.  But as a newbie animator, I don't put "voice creation" in the same category as animated character and scene creation.  I think it is safe to say that no animator expects to create a voice.  Instead, they USE voices.  So in this regard, I don't feel bad about using AI, especially if the people's voices used all were paid for the use of their voices.  It is similar to musicians (violins, trumpets, etc) being paid to create VST samples for use in a DAW.

    Screenshot of the Clipchamp engine:

    Clipchamp window.JPG
    879 x 877 - 43K
    Post edited by UnifiedBrain on
  • UnifiedBrainUnifiedBrain Posts: 3,588

    Dartanbeck said:

    ...and if you use Vegas anyway, that's probably going to be a big Win. I imagine Magix has done a great job with it to include it in Vegas. Yet there's always that chance that it isn't that great - makes one stir a bit.

    Walt Sterdan, the creator of the Hrimfaxi series (Excellent!!!) has been using PlayHT after not being able to find all of the same voice actors. It has "Ultra-Realistic" voices options, which is what he uses. They're fantastic. I have Talkia, which is pretty cool but has that "I am an artificial voice" sound to it no matter how hard you try - especially if one is trying to actually use it for dialogue. I mean... they call them "Actors" for really good reasons!

    Here's a little discussion we have going on: Mimic: Who owns what?

    Hrimfaxi Adventures Episodes 1 and 2 (ends at Ep 2) use real voice actors and a genius combination of Mimic Pro results followed up with puppeteering to add other facial movements

    Starship Hrimfaxi episodes use the Ultra-Realistic PlayHT voices and 2D animation software

    It's all fantastic. If I made money doing this stuff, I'd certainly invest in PlayHT's Ulra-Realism!

     

    Dart, thanks for posting this!

    Walt, congratulations on your efforts!  The videos are well-paced, which is uncommon from newly-minted animators.  For some reason, I liked the 2D animation style more than the 3D.  And the voices in the 2D sounded relatively good.

    Dart, it has been interesting evesdropping on your recent conversations.  Seems like I have good timing in recently getting more serious about animation. :)

  • WendyLuvsCatzWendyLuvsCatz Posts: 37,858

    the threat is real though

    one of my subscriptions 

     

    one of my own AI videos made in an hour using online tools and a video editor 

    my own music ipad Garageband

  • Steve KSteve K Posts: 3,192

    UnifiedBrain said:

    Vegas is apparently using the same text-to-speech engine as Microsoft does in their Clipchamp product.  With Clipchamp, you can use it for free.  There are three caveats.  One, you need to use either Edge or Chrome as a browser.  Two, it only outputs as an MP4, so you may need to strip the video component (which is not hard).  Three, some of the voices do not have the full range of expressions.

    Many thanks for all that.  Especially since I've learned that the Vegas text-to-voice feature is only available with their subscription products, not my paid for version even with the $150 upgrade.  I avoid dialogue anyway in short animations, and even in the 48 Hour contest, the required line of dialogue can be shown as text on a sign or whatever (which I just did in the Houston contest).  I did have some dialogue, all of it from my Digital Juice collection of human short phrases (e,g, "Quiet on the set").  And I do have a basic text to voice feature in Magix' "Music Maker", just three voices, but enough for my minor needs.  So the freebies you folks mentioned are now of great interest to me, and I have no problems with those three caveats.

  • DartanbeckDartanbeck Posts: 21,227

    Vegas ROCKS! I love how it can teach me how to do something I don't know how to do right on the interface, even if I've moved stuff around. It's pretty freaking powerful too, and I have the cheaper version, not Pro. 

    Many AI engines will use the same voice catalogs, but I don't think that always means that they give the same results. 

    @Wendy - I know it's here and that it's freaking Awesome, but rather than see it as a Threat, I welcome it as another media. Everywhere I look I see stuff that blows away my home attempts, but that wil  Never stop me - or even scare me. I watch it and go Whoa... that's Cool! And then I go back to what I'm doing.

     

    But you know me. I'm a bit... odd.

  • WendyLuvsCatzWendyLuvsCatz Posts: 37,858
    edited October 2023

    https://www.youtube.com/live/D9GpKINVomo?si=SPT7wUoBkbKiMbEH

    artists certainly do see it as a threat, this is painful for me to watch as I am awestruck by Fennah and his 3D modelling and drawing skills

    I imagine he has the same take on us using 3D models and assets

    it's why even though I use AI personally I would never use it professionally 

    it's a toy to me, but like any kid I want to play with the big shiney toys and not have the grownups lock them away 

    but

    on topic

    I don't really agree with it being used in this 48hour film contest

    the voices well I guess the same argument could be used, one should be paying voice actors

    or have the talent to do voices 

    I have neither money or talent devil

    yeah, cannot really give you a take on what to do

    Post edited by WendyLuvsCatz on
  • DartanbeckDartanbeck Posts: 21,227

    I love your talent.

    I think that it's fine to allow it. It still takes creative filmmaking skills to get anywhere. Let the media be media and let the audience decide. 

    I'll always win the day as long as I keep liking what I make ;)

  • WendyLuvsCatzWendyLuvsCatz Posts: 37,858
    edited October 2023

    I watched a bit more of Fennah's livestream, he is indeed disgusted by those who buy 3D assets to create "art" too as I suspected 

    yeah I am happy he is so talented but I am not going to live a miserable life just because I am not

    and I am not going to just pick up a pencil either unless it's to stab my talentless content using heart cheeky

    I still admire his work especially his Blender animations  as I do many I don't necessarily share other views with

    Post edited by WendyLuvsCatz on
  • DartanbeckDartanbeck Posts: 21,227

    So... let me get this straight. He uses AI to create art, yet he doesn't like artists who utilize a design team? 

    Whatever! LOL

  • WendyLuvsCatzWendyLuvsCatz Posts: 37,858
    edited October 2023

    Dartanbeck said:

    So... let me get this straight. He uses AI to create art, yet he doesn't like artists who utilize a design team? 

    Whatever! LOL

    no, he models in Blender from a cube like our Stezza does in Carrara and draws but he did a 3 hour livestream criticising Shadiversity a medieval sword and cosplay YouTuber for using AI on his art 

    I originally linked because he was discussing the huge job losses he and others in the digital art community were experiencing due to AI

    I was sympathetic but he needlessly trashed Shad and right now I am rather disgusted with him.

    Post edited by WendyLuvsCatz on
  • Steve KSteve K Posts: 3,192

    WendyLuvsCatz said:

    I watched a bit more of Fennah's livestream, he is indeed disgusted by those who buy 3D assets to create "art" too as I suspected 

    That seems pretty harsh.  I use purchased 3D elements to make short animations for my own amusement, and have a lot of fun with it.  "Art" seems like a pretty lofty term.

    And just to clarify, I started this forum just to show what the 48Hour folks think about AI.  The text-to-voice discussion was a tangent.  I don't think of that in quite the same way as general AI, although I see the similarity. indecision

  • If you want to comment on someone's views from a video it would be ebst to do so in the comments section, ideally have watched at least the sections on which you wish to comment.

  • StezzaStezza Posts: 7,811

    WendyLuvsCatz said:

    Dartanbeck said:

    So... let me get this straight. He uses AI to create art, yet he doesn't like artists who utilize a design team? 

    Whatever! LOL

    no, he models in Blender from a cube like our Stezza does in Carrara

    didn't take long for my style to be copied lol laughyes 

Sign In or Register to comment.