Fascinating how word order and syntax mobility alters images. Mind-blowing! Can't wait for the time writers will write stories as prompts and see the various visual outcomes -- still and video! Not far away. A new art form is in the works.
I agree, it's amazing how little a change can make such a difference and as you said, the further that extends into more complex media, the 'butterfly effect' could be quite noticeable! It's an interesting time...
Nice work as always, Wade. If you are still interested of viewer recommendations, here is a list of prompts that I’ve really been having fun with Alien rococo Neonpunk Industrial hallucination Gothic constructivism Peter Saville Design Psychedelic steampunk By Virgil Finlay
I agree with the use of hallucination word. I've been using or adding "flawless hallucinations" and it works great for imaginative artworks. You can add that to any prompt and it gives a whole new twist to the images.
This is an experiment I have been doing: I describe the item using my real artwork or photo, use Grammarly to fix it, and then modify the wording using a free AI rephrase tool. The outcomes are astounding. I have observed that synonymy produces visually appealing results that differ. Please tell me that this is not a Rum talk.😅
I really enjoy your content. Very well structured and great quality overall. As well as very informative. Thank you. Interesting to see how those power words lead to a more photography style, great to know.
Very interesting video. Thank you! Can you advise me on how to create portrait type pics of people and keeping their clothes the EXACT same from one pose to the next? I always end up with different clothes. Thx so much!
That's a tough one, I can only imagine they have a specific prompt or image prompt they add to each image to keep it consistent. Also vary region might be the tool that is used to fix minor differences after a figure is generated. In which case a highly descriptive prompt used specifically in vary region could be a good way to nail that clothing the same. I'd need to work on that to know for sure though
Wade, you have a way about you that brings back why some (most) of us are doing this...to have fun . lol. Im having a good time and thanks for the help in doing that with more efficiency. Have you tried using 5.2s describe on V6 yet? Or using Chat GPT to translate to another languge with better syntax and fluency. Google translate has a tendency to translate phrases word by word as opposed to have a native speaker my rearrange the whole paragraph completely since languages can be so different. Thanks mate!
That's the name.of the game isn't it, have fun :) I haven't tried those things but I was considering another one of these videos. So I'll add what you said to the notes :)
In my experience it seems as though MJ to a degree keeps in mind previous prompts you have input. With your sunset knight, if u began with the simple prompt you wd not get the same result. It is ALWAYS learning . At least that's what seems to be happening, I can't see 'under the hood' to be certain, but it makes sense, It is an AI
Fascinating how word order and syntax mobility alters images. Mind-blowing! Can't wait for the time writers will write stories as prompts and see the various visual outcomes -- still and video! Not far away. A new art form is in the works.
I agree, it's amazing how little a change can make such a difference and as you said, the further that extends into more complex media, the 'butterfly effect' could be quite noticeable! It's an interesting time...
this is the type of video i like to see
Awesome! Glad to hear it, thank you :) I'll try to make more.
Nice work as always, Wade.
If you are still interested of viewer recommendations, here is a list of prompts that I’ve really been having fun with
Alien rococo
Neonpunk
Industrial hallucination
Gothic constructivism
Peter Saville Design
Psychedelic steampunk
By Virgil Finlay
I absolutely am, and these look mint! Thank you very much mate, I'll put these on the next video :)
I agree with the use of hallucination word. I've been using or adding "flawless hallucinations" and it works great for imaginative artworks. You can add that to any prompt and it gives a whole new twist to the images.
Sounds awesome :) I will give it a try for the next viewer styles video! Thank you
This is an experiment I have been doing: I describe the item using my real artwork or photo, use Grammarly to fix it, and then modify the wording using a free AI rephrase tool. The outcomes are astounding. I have observed that synonymy produces visually appealing results that differ. Please tell me that this is not a Rum talk.😅
That's cool! Ill try it
I really enjoy your content. Very well structured and great quality overall. As well as very informative. Thank you.
Interesting to see how those power words lead to a more photography style, great to know.
Yeah it was interesting to see! I learned a few things making this one
Great finds! Thanks!
Awesome comparisons, all great to know bro
Thank you mate! A few surprised me.
Very interesting video. Thank you! Can you advise me on how to create portrait type pics of people and keeping their clothes the EXACT same from one pose to the next? I always end up with different clothes. Thx so much!
That's a tough one, I can only imagine they have a specific prompt or image prompt they add to each image to keep it consistent. Also vary region might be the tool that is used to fix minor differences after a figure is generated. In which case a highly descriptive prompt used specifically in vary region could be a good way to nail that clothing the same. I'd need to work on that to know for sure though
Wade, you have a way about you that brings back why some (most) of us are doing this...to have fun . lol. Im having a good time and thanks for the help in doing that with more efficiency. Have you tried using 5.2s describe on V6 yet? Or using Chat GPT to translate to another languge with better syntax and fluency. Google translate has a tendency to translate phrases word by word as opposed to have a native speaker my rearrange the whole paragraph completely since languages can be so different. Thanks mate!
That's the name.of the game isn't it, have fun :) I haven't tried those things but I was considering another one of these videos. So I'll add what you said to the notes :)
In my experience it seems as though MJ to a degree keeps in mind previous prompts you have input. With your sunset knight, if u began with the simple prompt you wd not get the same result. It is ALWAYS learning . At least that's what seems to be happening, I can't see 'under the hood' to be certain, but it makes sense, It is an AI
That could be true, I know they are aiming to add that as a primary feature in the future but there could be a degree of it happening now
Very good video thanx
Thank you!
I like the experiments. But 4K seems to be confusing for me.
Thank you! Well essentially 4K just directs the image toward being a photo (in the instance anyway) but doesn't improve quality
👋
🤘