Відео

Oh yes. Leonardo does produce some realistic-looking images. Like you said, they're probably pushing people to move towards upgrading. Downside of Piclumen so far is just the inability to produce images that are up to 4k level. The results are impressive though

@@TalkingTechWithWilliams Agree, I'm sure they will update in the near future, since it's getting traction with people and they will probably go with paid plans too once they add more features .

AI to the world! 🔥

Thank you! I'm glad you found it useful🙂

There's an updated video that shows the image to image at work. Feel free to check it out here: ua-cam.com/video/dMNNPHwcejA/v-deo.html

Hey... Thank you for watching. I'm not sure I fully understand what you're asking though. Do you need me to make another video showing how to make images that aren't automatically zoomed? Please clarify and I'll love to help out :)

Great questions! Here are your answers: 1. The resolution for the 1080p generations is 1920x1080 2. No, Haiper hasn't made a provision where you can specify image sizes. That would probably come in next updates 3. For prompt adherence, Haiper does a pretty good job, as it doesn't have the "prompt enhance" feature used by other generators these days. 4. Confuse it?? (Not sure I understand what you mean here) 5. Yes it does, but I wasn't so impressed by the results. I suggest you give it a try, since these features are free to use at the moment.

@@TalkingTechWithWilliams Can it adhere to multi itemed prompts in the same that Dalle can? Thank you.

Oh for that, no it can't. At least at the moment though

You're welcome! I'm glad you found it helpful :)

Oh thank you! It's a great tool that I believe will always come in handy 😉

I'm glad it was helpful 🙂

It actually does. Click on "Image Creation", choose any preset order than Leonardo Phoenix, then click on the image upload icon just beside the prompt text area. You'll be able to upload your image and use the character reference. NOTE: Using the Character Reference option requires a premium account. I hope this helped :)

The video made was majorly a demo to showcase the possibilities that KREA can bring. However, if you want me to package something of that nature, I can surely do that :)

Oh it's nice to know other creative use cases of Krea 💯💯

You're right! Getting ideas from AI can't hurt 😉

That makes sense. I haven't tried that. Although, the fact that the tool still allows you to upscale and download an upscaled version makes it cool

Yes you can, but then here are a few things to note: 1. The duration of the video will determine how long the processing will take. 2. Check the settings section and select the "Graphic Card RAM-saving" mode in order to run it without making your PC lag

Hey... Thanks for the knowledge you've added to viewers. You're right. HEDRA does a good job when a clear front-facing photo has been uploaded. I decided to upload that photo in order to push the system to the limits and see how it handles situations like that lol.

@@TalkingTechWithWilliams 😂 You are very welcome my friend! God Bless You!

I'm glad you found this helpful! 🙂

Hey, I'm not sure I understand what you mean. Are you saying you use KREA's inbuilt image generator to generate images? I actually don't fully trust KREA's image generator to create consistent characters or even better ones than services like Leonardo AI at the moment. I'll advice you use Leonardo or Stable Diffusion or even DALL-E to generate your images, then import them to KREA

@@TalkingTechWithWilliams I'm talking about when I want to make a video and upload my personal image, KREA doesn't generate a video with me but with someone very similar to me

That's weird. Although, I believe the aim of the tool is to create a smooth transition between images. It also works best with more images, for that's when the storytelling becomes immersive.

Thank you! I'm glad you liked it :)

Yes sure! Although if you are to use it for your UA-cam channel, then I'll recommend having consistency in your character. That way, viewers will get used to seeing a character they know. I'll also recommend using a fictional character like the cartoon I used in order to avoid making it look like you're using deepfake videos on your channel Lol

@@TalkingTechWithWilliams Thank you , I like your channel bro.

You're right. To make a video as long as 4 minutes with ease, Morph Studio has made it a lot easier. I made a video on that too. Feel free to check it out here: ua-cam.com/video/1RyWYWFZl-I/v-deo.html

Your support means a lot to me, thank you for being a dedicated viewer!

Thank you! I'm glad you got value from this. For image generation, I would suggest you check out tools like the Bing AI image generator. It works really well, even though the dimensions are stuck in a square resolution. I have a video on it here: ua-cam.com/video/QGLQh3vXdC8/v-deo.htmlsi=CVnsSgNsxBol3Z3Y You can also check out Adobe Firefly. Let me know if you want to see a video on it too

Thanks for the valuable feedback. I guess I'd have to do better next time then 😊

This was my reaction at first too! However, the output images all have consistent facial appearances with slightly different outfits. That's Leonardo's main focus I believe

Not really any of them... but then again Sora isn't "real" yet as no one is able to use it.

@@noonesbiznass5389 True... we're still banking on what they've shown and told us. I think with the expectations they've created around Sora, OpenAI will be under so much pressure to deliver what they've promised

@@TalkingTechWithWilliams Yeh will be interesting to see... I'm suspicious because they are sure taking a long time keeping it out of people's hands :)

@@noonesbiznass5389 Lol or maybe they announced it too early

At the moment, no, but the desktop version is mind-blowing. I guess due to the processing power needed, having a mobile app that'll work as good would be difficult to achieve. NOT impossible though ;)

Right?! For a free app, I am impressed