A few members asked me yesterday, how do I create the talking avatar in my videos on the new YouTube channel Real Stella Sky AI. I made a video to explain the process in 3 simple steps! I hope you like it! I already posted the prompts that I used in the video in the YouTube freebie in the classroom! Thank you for all the new subscribers! Yesterday, I had only 5, and now I have 40! That's amazing! Comment on this video. Which avatar do you like more? Subscribe to my channel. Then, comment here that you are subscribed. I will give you access to the Freebie. YouTube doesn't always show the names of the new subscribers, so please don't forget to comment here! If you want to create your own talking avatars, please use my affiliate links; this will help me bring more free tutorials for you! Here is the software you need: (my affiliate links) Dupdub: http://www.dupdub.com/?via=stella and Elevenlabs: elevenlabs.io/?from=partnersalazar873 Enjoy!

I finally got time to download and start using EJB from JUNE - yes, I'm behind. 😂 These were all done in Leonardo AI using either Phoenix or AlbedoBase XL (new Gen V2 mode). I also used the the "Graphic Design Pop Art" setting (AlbedoBase XL) for the woman on the balcony, BUT the real secret to the style is one of Stella's Geometric Presets images combined with one of the Earthy Tones images, both from EJB. I love how this one turned out! Some of the images use the prompts as-is, others I modified a little or a lot.

I've been using ChatGPT for many months now to create sets of color palettes with hex codes. Today I tried something different. I asked ChatGPT to describe this image. Then I asked for a color palette. It took several minutes, and it paused a few times, but ultimately provided an analysis of the character with RGB colors. I then asked for a conversion to hex colors. It was able to provide those. It runs several Python scripts, there's a bit of commentary from ChatGPT, and then it summarizes and provides the results. The entire analysis is way too long to post here, however, here is how I asked: First, I uploaded the character image below. I asked: "Hi, can you please describe this character?" After receiving the description, I replied: "Thanks! Is it possible to generate a basic color palette from this image?" (I'm always polite in my responses. Yeah, I know, it's weird 😆) In the first response, it returned all grays. The interesting thing to me is that it knew this was not the important parts of the image, and in the third image, you can see its response. Ultimately, it provided four colors, with hex codes. While this is really useful, I think it's a lot faster for me to just open the image in Milanote and extract the colors with the eyedropper tool. BUT - this is a fantastic Python script for automatic palette extraction, and it's still extremely useful. I think if it could run the analysis just a little bit faster, and I ask for the hex codes in the initial request, it will be faster than manually pulling the colors. In the last image, I added color swatches in the Milanote app, using the colors provided by ChatGPT. Hope this is useful or interesting to someone! Ultimately I use these images with color swatches for mood boards and reference when working on non-ai art like sculpting or painting, etc.

With some modification, of course, like changing the hair color and swapping out background elements. I really like the Nightcafe results here, the DreamShaper model is great at illustrative styles. Having a lot of fun with these! Here are two images from Nightcafe (DreamShaper XL Alpha 2 model), and two from Starry AI (StyleVision XL model).

Did you try the personalization feature already? I will show you my examples here. "Personalization (--p) Hey everyone! We're releasing an early test version of model personalization today. Here's how it works. What is Model Personalization? - Every time you write a prompt there's a lot that remains 'unspoken'. Our algorithms usually fill in the blank with their own 'preferences', which are really the combined biases and preferences of our community - But of course everyone is different! Model personalization learns what you like so that it's more likely to fill in the blanks with your tastes Model Personalization Requirements - Right now model personalization learns from votes in pair ranking and images that you like from the explore page - You need to have roughly 200 pair rankings / likes in order for the feature to work - You can see how many ratings you have on the pair rankings page, or by typing /info on Discord How do I use Model Personalization? - Just type --p after your prompt, or turn on personalization for all prompts using the settings button in the prompt bar - When you enable personalization you get a 'code' added after your prompts, you can share this code to so others can use the personalization effect being applied to that image - You can control the strength of the personalization effect by using --s 100 (0 is off and 1000 is maximum and 100 is default) PLEASE NOTE: The personalization is not a stable feature right now, it will change as you do more pair ratings (subtly) and we may be pushing out algorithm updates over the next few weeks. So just assume that it's a 'fun feature constantly in flux'. This is a totally new frontier of image synthesis. We hope we discover a lot in the process of doing this. Have fun!" (Source Midjourney news) I tried it with that prompt and got pretty glowing and sparkling results with personalization, which are really a part of my style!