and thanks for the tip! I haven't tried image prompting yet but now I'm excited haha

@@FutureTechPilot Update from watching your latest video on the new “Style reference” prompt. It’s even more of a wow! with a mix of image prompt at the beginning and -sref as used in your video. 🤩

can't wait to try @@pinupkulture2231

Enjoy!!

Just trying my best! Thanks for helping me experiment haha

I have too much on my plate to facilitate that kind of interaction right now, but I'll keep it in mind for the future! thanks for the suggestion

😂

Actually though 🤣

Actually though 🤣

Actually though 🤣

and thanks a lot for taking the time to comment!

Enjoy!

And thanks for sharing your thoughts! It's hard to keep up with all these updates haha

I miss them too lol but someone complained about it recently.. I'll see if I can bring it back!

@@FutureTechPilot I will love that brown. And on Alpha websites inpaint buttons on there. Re-edit it the picture now on niji V6 and V6. I will love to do a video on that too. I just did live talking about it. What it brings to midjourney alpha sites

Cheers!

hahaha there are way too many rabbit holes to explore

@@FutureTechPilot I’ve taken this entire day off from Midjourney. Even though my sub renewed today. I made the misstep of reading the feedback channels - that was too much noise and I got derailed. Better now but I’m not going to open it until tomorrow. We’ll see.

I'm still working on them! (The technology keeps changing and it's difficult to keep everything up to date) They're almost ready to film, and then once that is completed, it won't take long for it to be ready online. The beginner course will be available first, and then I'll start on the Advanced course.

I love the prompt understanding - it's actually hilarious 😂

Yah did you try any new tricks? or just a description worked?

@@FutureTechPilot Just by using variations , and pic in prompt too. Will not get it right in all pics but it's a lot more consistent than before. In 5.2 or Niji 5 there was a pic of a medieval character that I was using as prompt for clothing and it wouldn't get right in any options and now it gets close enough often! Was getting disappointed with MJ but this has gotten by hopes back up.

@@FutureTechPilot Oh, btw, I forgot to say that --style raw works better (along with the --niji 6)for the consistency I talked about.

yessss

Midjourney parameters are so fascinating 😂

😂 literally used 4 hours last night just exploring for this video

Competition is a great thing!

💯

haha yah that will be fun

😂 I like it too actually

I'm not familiar enough with Leonardo! Especially cause Niji is an algorithm specifically trained on Anime

Hmm ... that's definitely what it was trained on, so it would be tough. But you can try negative prompting! So you would say something like '--no anime, japan animation' at the end of your prompt - try to describe the look you don't want and the bot might be able to avoid it!

@@FutureTechPilot I will try that. Thank you!

​​@@FutureTechPilotUpdate: I've been experimenting, but it's usually a hit or miss when trying to create a non Anime style, for instance a western animation style even with --style raw. After a few variations, it inevitably transitions back to Anime. It even dramatically changes the features of my characters in spite of the cref image, de-aging them from adults to underaged kids with big eyes and questionable clothing. It's too bad, because I find Niji to be a lot better than the standard V6 when it comes to the illustrative style, character sheets, dynamic movement and expressions, etc. Hopefully we get something like Niji for western animation someday. Something that allows us to create something in the style of Marvel or DC. Anyway, thanks for the video!

@@kayskreed Yah let's hope the tools continue to improve!

@@FutureTechPilot Update: Wow. I actually managed to pull it off, i.e. creating a non-Anime i.e. western graphic novel or comic style in niji 6. So, I'm still testing it out, but I had to start by creating my character in V6 (not Niji 6) first, until I was able to get the style I wanted. V6 is good for closeups, especially photorealistic style, and since many western comics are drawn in that style it was the place to start. It doesn't typically give me whole body character sheets since it's more portrait focused, but that's okay. I then took that reference image I made in V6 in the western animation style and brought it to Niji 6, but here's the important part: I had to strengthen the image weight (--iw) parameter to higher than +1 and bring down the character weight (--cw) to zero (since the cref is a portrait), otherwise the proportions are all wrong since the head will be huge, the character bug-eyed, and the body miniscule in relation to the head. CW zero only focuses on the face/hair to my understanding, which is all you need, since you can redescribe the outfit in the text prompt, and with --cw 0 it will adjust the body proportions accordingly and realistically. Then I just added my prompt description with the scene and character in mind, with the cref image link I made in standard V6 at the end and voila! Been struggling for a month on this, I was on the verge of giving up. I'm ecstatic! Give it a shot! I hope this helps anyone else struggling with creating non-Anime art in niji 6.😄

oh yah baud

That's interesting! I know you didn't explicitly ask for advice - but what I would do is put together a quick reference image in photoshop or something. Literally take a picture of a snowmobile and place it on a dirt background, then use that as an image prompt + some words and use different image weights like --iw 2 or 3

Thanks, I will try an image prompt. I wish I could show you the differences. HAHA! Thanks again for your channel.😀😀😀😀@@FutureTechPilot

🤝

What do ya mean?

It's very difficult to get traditional animation such as the hand-drawn style of Looney Tunes and Disney in regular Midjourney. It's very much hit-or-miss unlike Niji, which nails anime perfectly every time.@@FutureTechPilot

@@astralislux305 It will be very easy soon

Do you say that because of the speed of AI progress? Or do you know something specific? If so, please share! @@ancientchaos1337

What do ya mean?

@@FutureTechPilot like a comics books

@@willianserafim1 best to do that in photoshop!

🤝