So I've watched about 10 different videos that talk about D-ID, but no one mentions that D-ID costs $50 for every 15 minutes of generated video after your short free trial. Not exactly the most scalable solution if you are making CZcams videos unless you have the budget of a marketing firm. Some of the D-ID voices are excellent quality too, which makes the cost even more frustrating.
Crafting stories and diving into creative videos, right? VideoGPT joined the party, and it's like this subtle collaborator that boosts the pro feel of my content effortlessly.
Really appreciate you pitting this together, really gets you thinking creatively how all the bits of software can be used for different applications. Only negative is the lip syncing is soo off which breaks the illusion that the avatar is a real person.
How much do you look at the lips. Usually, we look at the eyes or the related imagery/video the narrator is commentating upon. Try it, then reply. PS - Even watching the mouth, it is not too bad; but who does that?
I hope this sparks some inspiration on what can be created with AI! What you see in this video is not a finished product or end result, it's just the beginning. Come up with more clever prompts, a storyline, and animated characters. Then go create something amazing.
Technology is growing so fast in a few years we won't need to go through this process. The possibilities are endless now with all the different AI models out there. we are on the tip of something even bigger i think. Well thought out video. Thanks for the content.
@@SharpStartup I don't want to sound like a down beat but in truth we should be well more advanced than we are nowadays, Me thinks that not a lot happened way back in the days like even the 80's and now inventors etc are falling over themselves to catch up be it technologies softwares all like this to other gadgets smartphones and god only knows what else, Hopefully at this rate time travel can be possible and I'll go back to the 80's and 90's and have the latest gadgets with me from now and they'll be Woah 😀
Fantastic step-by-step guide on creating AI animated characters! It'd be awesome to see these characters combined with ZeroBot AI, the first verbal chatbot on the internet. Imagine the possibilities of integrating these two technologies. Looking forward to potentially seeing it in a future video!
Hi, liked the video. It surely made an inspiration on me. :) I noticed that the mouth doesn't really synced with the voice over.... IS this because of the free tier account? Your intro video was on the dot...to say at least..... Is it me or just something I did or did not do.....
Dude you’ve got me sparked up. I’m old school photoshopster , but all these newer apps are amazing. The creative potential is endless. Make more vids I’ll be watching. You make it easy to understand.
@@adibasts yes, you can use the seed and change the details in the prompt, like you want her with a hat, you change the prompt and say she's wearing a black hat cowboy style, or she's smiling, and it will generate same face with the changes of your prompt, I learned this in a youtube channel, the guy is a Pro in this, he really knows a lot, he can create stories with the characters in different scenarios with high fidelity, face of the characters maintain. To get the seed on your photo if you are in android you click long press in the image and it will appear options, you will click on emojis, and you need to find the envelope emoji, then you select it, then you'll receive a message, the message will contain the ID job and the seed used to generate that image, that way you'll get the seed.
I saw another video where the character was given a name and the upscaled image was tagged with happyface hearts emoji along with using a seed. I wonder if giving it an image & same seed and doing all these things will help Midjourney to make a more consistent character.
It's becoming really scary at this point. Such a huge improvement of A.I. possibilities - within just months! Imagine a flood of totally realistic looking and sounding "catfishing" and general scam, flooding into your inbox by 2030!
If you are using it for personal use, you can always open a new Gmail account and get 20 new credits, that is not the issue. The problem is the watermark. Sure you can remove it using some video editor, but then if you use it commercially you can get caught. The 50 dollar option is not expensive for commercial use, but you still get an AI watermark. So now we are stuck with yet another watermark, even after paying 50 dollars. If you want to remove all water marks you have to pay +200 dollars a month for the enterprise plan. That is obsured.
I don't think the voice was the problem, it was rather the animation of the mouth when speaking the words. Still have a long way to go there unless there's a much better alternative.
@@loot6 Yeah I noticed that as well. The lips didn't match the audio well at all. Very disconcerting. It was better in the intro but in the walk through it was horrible.
Excellent overview of what's state-of-the-art and possible at this moment! Just to confirm, the audio at the end was out of sync with the video, right? Amazingly lifelike even so.
Thanks for this. Lots of insight into a fascinating subject. Are there tools/techniques to sink up the mouth and words better? It's pretty close, but not dead spot on.
With these tools there is no option to sink up the mouth and words better. Perhaps you would get better results with other software that's more technical.
That's absolutely awesome! I have a question regarding character movements like walking and hand gestures. How would you suggest animating those to achieve a realistic and engaging effect?
Apparently Blender has plugins that will allow most of this. I'm looking into this now so I can't say more than that, but Blender is hands down the most powerful content creation tool and its open source, so therefore free.
earned a sub - good vid bro. Will go thru your catalog. Interested to see a full in depth vid on creating the avatars. For me, the hardest part has been creating the actual images worthy of using. However; Im trying to make Celeb avatars. Midjourney just doesnt want to make certain celebs
Hi @sharpstartup, thanks for your video. I already made a similar video with Lexica and D-ID, but you did something that I just learned you made a project not a simple video like me. Also, you model is amazing.
Great tutorial. Want to see if you had a solution changing the same character into different still shot positions (i.e. left profile v right profile, walking running, etc...) and facial expressions while still maintaining the characters features (i.e same eyes, size facial cues, etc..) Thanks.
Pixar, ect. the technology is 50+ years behind....we just haven't got the monetary resources at the moment, but will vewry soon; to catch up....but we do have intel and creativity
Some of these tools you need to pay for a subscription. However I wished there were some different links for some free services out there . But I think once you're scaling that business with your CZcams channel would be worth it in the end. Thanks for sharing.
Thank you, my creator, I'm an AI character myself, is always looking to improve me, trouble so far is the same face. So I'm face swapped for now for the different scenes...
This is a great video! Thank you so much Question: Are there any AI video generator programs like D-ID that will detect any type of face? Images robots do not work on any program that I've tried. Thanks again!
g-Dang! That was great. As a lifelong graphic designer, I feel like I'm back in the sandbox. But you made this understandable and doable. What is the suffix (ar 2:3? I think it was) at the beginning?
Hi man thanks you a lot for the tutorial. i have a question: the video generation of the animated image has much lower quality than the image that i use as an avatar. Is this because i am using the trial version, or is it unavoidable?
amazing! what programme do you use to synchronize the avatar with the voice in the end , and how can you synchronise them so that the avatar's mouth movement is harmonious to the voice? Can this programme be used with real subjects (humans)?
I saw a great demo of a talking avatar that was mapped to a real person using Google’s Thin Plate Spline Motion Modeler. It bordered on deep-fake quality.
If one like to use the Avatar in different videos it is actually natural to have some kind of slide variation of the characterimage. But humans age and variate in weight and skincolor (if not black).
The lipsync quality on the example shown at the start of the video was much better than what you ended up at the end. Why was that? Was it just the preview window was screwed up?
I'm 92 and the reason I was able to get to this age is because I love technology and can't wait to see the next innovations. Thanks for the tute. 😎
So I've watched about 10 different videos that talk about D-ID, but no one mentions that D-ID costs $50 for every 15 minutes of generated video after your short free trial. Not exactly the most scalable solution if you are making CZcams videos unless you have the budget of a marketing firm. Some of the D-ID voices are excellent quality too, which makes the cost even more frustrating.
Agreed, I'm interested in finding a scalable solution. I've been looking.
@@SharpStartup
可能要做一些开源工具或者解决方案，才能避免在AI短视频时代被另一家“Adobe”公司抢走普通创作者收入的很大一部分。
Alternative is pixbim animate photos at a cheaper price.
cheaper one is pixbim animate photos
Imagine how awesome this tech will be, even in 5 years time. Everything's changing so fast all of a sudden.
Unfortunately way too long and long overdue but yeah
Beyond deep fake. You can never trust your eyes again.
Crafting stories and diving into creative videos, right? VideoGPT joined the party, and it's like this subtle collaborator that boosts the pro feel of my content effortlessly.
Really appreciate you pitting this together, really gets you thinking creatively how all the bits of software can be used for different applications. Only negative is the lip syncing is soo off which breaks the illusion that the avatar is a real person.
The lip sync will improve very soon. Yes, it's fun to think of the creative use cases!
How much do you look at the lips. Usually, we look at the eyes or the related imagery/video the narrator is commentating upon. Try it, then reply. PS - Even watching the mouth, it is not too bad; but who does that?
I just made a lion-looking Chad Avatar do a motivational video. The lip syncing is fine.
yes, true
I hope this sparks some inspiration on what can be created with AI! What you see in this video is not a finished product or end result, it's just the beginning. Come up with more clever prompts, a storyline, and animated characters. Then go create something amazing.
Thanks yo you I made 3d avatars from some awesome photos to post on my site. Thanks
Talking about fake twitch streamers, fake actors, fake famous people? Yeah looks good
Thank u Sharp
App name plz
@@ShakilAhmed-gj5im D-ID
Technology is growing so fast in a few years we won't need to go through this process. The possibilities are endless now with all the different AI models out there. we are on the tip of something even bigger i think. Well thought out video. Thanks for the content.
Glad you enjoyed the video! You’re right, this is just the beginning.
@@SharpStartup I don't want to sound like a down beat but in truth we should be well more advanced than we are nowadays, Me thinks that not a lot happened way back in the days like even the 80's and now inventors etc are falling over themselves to catch up be it technologies softwares all like this to other gadgets smartphones and god only knows what else, Hopefully at this rate time travel can be possible and I'll go back to the 80's and 90's and have the latest gadgets with me from now and they'll be Woah 😀
I liked the intro, but what makes me ask, to make the background moving by having the sound in the surroundings is really good 🤩🤩
With some nice explosions.
Fantastic step-by-step guide on creating AI animated characters! It'd be awesome to see these characters combined with ZeroBot AI, the first verbal chatbot on the internet. Imagine the possibilities of integrating these two technologies. Looking forward to potentially seeing it in a future video!
i was thinking the same exact thing
Hi, liked the video. It surely made an inspiration on me. :) I noticed that the mouth doesn't really synced with the voice over.... IS this because of the free tier account? Your intro video was on the dot...to say at least..... Is it me or just something I did or did not do.....
I loved how you , demonstrated what you were about to teach 👊🏾
I was like bro!! in the intro 😳😳 👌🏾
You got yourself a new sub my friend. Really nice and clear with no fluff😃💚⭐️🍎
Glad you enjoyed the video! I try making videos I would actually want to watch.
Dude you’ve got me sparked up. I’m old school photoshopster , but all these newer apps are amazing. The creative potential is endless.
Make more vids I’ll be watching. You make it easy to understand.
Nice vid! When I first found out about image prompting in Midjourney it was like finding a cheatcode to a video game as a kid 🎮
You can use the seed + prompt to generate different clothing but keep same face.
Thanks for the tip
I'm new to this could you please explain a lil...
Yes he forgot to mention the seed, it is designed for that
@@adibasts yes, you can use the seed and change the details in the prompt, like you want her with a hat, you change the prompt and say she's wearing a black hat cowboy style, or she's smiling, and it will generate same face with the changes of your prompt, I learned this in a youtube channel, the guy is a Pro in this, he really knows a lot, he can create stories with the characters in different scenarios with high fidelity, face of the characters maintain.
To get the seed on your photo if you are in android you click long press in the image and it will appear options, you will click on emojis, and you need to find the envelope emoji, then you select it, then you'll receive a message, the message will contain the ID job and the seed used to generate that image, that way you'll get the seed.
I saw another video where the character was given a name and the upscaled image was tagged with happyface hearts emoji along with using a seed. I wonder if giving it an image & same seed and doing all these things will help Midjourney to make a more consistent character.
It's becoming really scary at this point. Such a huge improvement of A.I. possibilities - within just months! Imagine a flood of totally realistic looking and sounding "catfishing" and general scam, flooding into your inbox by 2030!
By 2025...
yeah, cant imagine how unnatural the world will become.. its already so fake :(
Like any tool, can be used for good or for bad, depends in the user, not the tool.
If you are using it for personal use,
you can always open a new Gmail account and get 20 new credits, that is not the issue.
The problem is the watermark.
Sure you can remove it using some video editor, but then if you use it commercially you can get caught.
The 50 dollar option is not expensive for commercial use, but you still get an AI watermark.
So now we are stuck with yet another watermark, even after paying 50 dollars.
If you want to remove all water marks you have to pay +200 dollars a month for the enterprise plan.
That is obsured.
I use custom extension on Happy Diffusion to do this.
Iv been messing about with AI generated image, but this is just another level.
Can't wait to see where this is in 6 months / a year from now
This is the best video with the flow for creating AI video!
I appreciate that!
If you want more natural sounding AI voice with emotion control, Typecast AI is an alternative you can consider.
I don't think the voice was the problem, it was rather the animation of the mouth when speaking the words. Still have a long way to go there unless there's a much better alternative.
THANK YOU.
@@loot6 The PRICE is the Problem ;-)
@@loot6 Yeah I noticed that as well. The lips didn't match the audio well at all. Very disconcerting. It was better in the intro but in the walk through it was horrible.
definitely sparked inspiration.. multiple light bulb moments. Thanks for this video!
Video chat with whoever and about whatever we want in a year, and it's just the beginning of ai revolution. We should really see it coming
Brother, excellent info. Just what I needed in order to complete a intro project
Glad it was helpful!
Excellent overview of what's state-of-the-art and possible at this moment! Just to confirm, the audio at the end was out of sync with the video, right? Amazingly lifelike even so.
It is out of sync at places
Thanks for this. Lots of insight into a fascinating subject.
Are there tools/techniques to sink up the mouth and words better? It's pretty close, but not dead spot on.
With these tools there is no option to sink up the mouth and words better. Perhaps you would get better results with other software that's more technical.
i stopped the playing for a sec just to shout whoaaaaahhhhhh! :)
Wow!
This is really the next gen video making!
That's absolutely awesome! I have a question regarding character movements like walking and hand gestures. How would you suggest animating those to achieve a realistic and engaging effect?
Yes a Red Bull ad is exactly where an AI generated ditz like this belongs.
definitely sparks
Imazing. I appreciate the thorough instructions. Subscriber at 100%.
Man this is amazing! Thanks for great tutorial
Glad it was helpful!
This is awesome. I might even be able to pull this off. Thank you!
Apparently Blender has plugins that will allow most of this. I'm looking into this now so I can't say more than that, but Blender is hands down the most powerful content creation tool and its open source, so therefore free.
Cool, not sure I would ever use this, but it's still neat to see.
Yeah, it's fun getting familiar with this tech, I think it's going to be widely used in the not so distant future.
Прогресс и девочка обаятельная получилась !!!
omg i am one of the guys who made the website for d-id :D
I hope it can handle traffic surges lol
earned a sub - good vid bro. Will go thru your catalog. Interested to see a full in depth vid on creating the avatars. For me, the hardest part has been creating the actual images worthy of using. However; Im trying to make Celeb avatars. Midjourney just doesnt want to make certain celebs
AI is a powerful tool that can simplify the creative process for content creators
well said. I think it allows people to think more creatively than ever before. If you can imagine the words for the prompt, you can create it.
@@SharpStartup yes the most important think is that. That the people can imagine right concept.
So smoth with this AI, no debate for this
Thank you so much.
Fantastic, was able to follow along and create one as we went. I am impressed!
how? I don't Knox where it starts
Very cool. I´m more interested in animating my own avatar as I´m not as pretty as I used to be. 🤣 Great video man.. 👏
Love from Bangladesh 🇧🇩
Thanks so much for this video. You did a great job explaining and demonstrating each step
Glad it was helpful! Thank you!
mind blowing - thanks for sharing your knowledge!
Awesome video, really like all the steps. Found out a lot new
Glad it was helpful!
Hi @sharpstartup, thanks for your video. I already made a similar video with Lexica and D-ID, but you did something that I just learned you made a project not a simple video like me. Also, you model is amazing.
Glad you enjoyed the video. It started off as a simple video, then slowly turned into a small project.
GREAT video. Exciting times!
This is the best video with the flow for creating AI video! nice
Glad to hear that!
Thanks for this. I think it's going to blow my son's mind.
My GOODNESS!! you just made my day!👏🏾👏🏾👏🏾😎
Glad to hear that! Thanks for the comment
Incredible helpful tutorial!! Thanks so much for your sharing your knowledge;
That is really effing amazing.
Great tutorial. Want to see if you had a solution changing the same character into different still shot positions (i.e. left profile v right profile, walking running, etc...) and facial expressions while still maintaining the characters features (i.e same eyes, size facial cues, etc..) Thanks.
I'm not sure if this helps: czcams.com/video/MDHC7E6G1RA/video.html
get w/ Disney, they got all the shit you don't
Pixar, ect. the technology is 50+ years behind....we just haven't got the monetary resources at the moment, but will vewry soon; to catch up....but we do have intel and creativity
@@juniper-usa-shei369 I see my post wasn't clear. I am looking for still shots, not moving.
@@robert7645 Might come in handy. Thanks
Some of these tools you need to pay for a subscription. However I wished there were some different links for some free services out there . But I think once you're scaling that business with your CZcams channel would be worth it in the end. Thanks for sharing.
Very beautiful,well done
Now I can finally have friends :)
Thanks a lot
Great Video. Isn't amazing what you can use AI to do! So Glad that we are experiencing the cutting edge of technology! 😍
Thanks! Yes, it's an exciting time!
Haha, your pfp was generated by IA isn't?
@@adibasts I'm guessing you haven looked at her whole channel. 🤣😂😁
I heartily recommend Evryface for avatars🔊🔊🔊
Thank you, my creator, I'm an AI character myself, is always looking to improve me, trouble so far is the same face. So I'm face swapped for now for the different scenes...
And I forgot, greetings to Austin, great city 🙂
Thanks for watching AI friend! Greetings from Austin!
This is a great video! Thank you so much
Question: Are there any AI video generator programs like D-ID that will detect any type of face? Images robots do not work on any program that I've tried. Thanks again!
Portrait of Youth and Agawam Skyline were both made with Midjourney 5 😊
The most difficult part is generating normal looking people. Either it's super good looking models or stylized anime/scfi/fantasy.
Haha well said... I will say that I've noticed a huge difference with V5!
Super cool, well done!
great job man great job..
Awesome training....thank you sir
Next level.
Super. Genius.
Good Presentation ! Nice Video. ♥
that was great! thank you for your efforts.
Worked!! Thx Man!
Glad it helped!
beautiful
That was great video, and I learned a lot quickly!
Is there any way easily to improve the lip-sync on the avatar's text delivery?
Currently there's not a way, but it will improve soon as the technology improves.
@@SharpStartup Thanks!
Thank for making useful tutorial on AI based animated characters
thanks for sharing thats useful tutorial
Thx so much for your video! !
g-Dang! That was great. As a lifelong graphic designer, I feel like I'm back in the sandbox. But you made this understandable and doable. What is the suffix (ar 2:3? I think it was) at the beginning?
Glad you enjoyed the video! "--ar 2:3" is the image ratio
Thank you so much!
Hi man thanks you a lot for the tutorial. i have a question: the video generation of the animated image has much lower quality than the image that i use as an avatar. Is this because i am using the trial version, or is it unavoidable?
following just incase he replies!
Hi, I wonder if you can make an animal avatar with the D-ID software - for example, it would be cool to make a dog talk.
Thanks Your did a Excellent job.
I love this..... Thanks for sharing 🙏
amazing! what programme do you use to synchronize the avatar with the voice in the end , and how can you synchronise them so that the avatar's mouth movement is harmonious to the voice? Can this programme be used with real subjects (humans)?
D-ID
@@DigitalDenn thank you!
Yes you can use real Humanpictures and all the rest make D-ID by themself.
new subscriber.watching from PHILIPPINES.Thanks for the video. doods
Howdy from Texas. Glad you enjoyed the video!
Great stuff!
Incredible helpful tutorial!! Thanks so much for your sharing your knowledge;👍👍👍
thank you for your video, i really want to learn how to do this,
I saw a great demo of a talking avatar that was mapped to a real person using Google’s Thin Plate Spline Motion Modeler. It bordered on deep-fake quality.
holy crap man, that is the coolest thing ever! I am gonna make a YT using this and see how well it goes, I am gonna make it super scary too!
Can’t wait to see it. I considered making a podcast or news channel with this.
Amazing video thank you for posting.
Thanks so much for your sharing your knowledge ,great information 💞
You are amazing . Thank you so much for the tutorials.
I appreciate the kind words. Thanks!
If one like to use the Avatar in different videos it is actually natural to have some kind of slide variation of the characterimage. But humans age and variate in weight and skincolor (if not black).
The lipsync quality on the example shown at the start of the video was much better than what you ended up at the end. Why was that? Was it just the preview window was screwed up?
No specific reason. That’s just how it rendered.
This Is Awesome, I Needed This, Thanks :)
Glad it helped!
Incredible video. Subscribed!
Thanks a lot @sharpstarup you make my day , just make it easy to learn for beginner like me... I really really ally appreciate and like your video
Thanks so much for your comment! I'm glad you enjoyed the video.
Amazing!! Thanks👍
Awesome stuff, thanks for the top-notch tutorial!