I really appreciate your channel's detailed info. Your Analytical info opened up the hood a bit for non compu SCi or non AI specialists. I am an old school analog visual artist who is very curious about the all inner workings of these tools and trainings. How can the AI be trained on what foreshortening is if it does not know how to experience binocular vision? Or understand the distortions of seeing 3D space in 2D? This AI image generating landscape is so new to me and changing so fast. I realise your focus is on Mid Journey, but can you tell me if these image weight command works in Stable Diffusion? I really appreciated your inclusion of the old woman young wife 'silly Illusion Illustration" in another prompt video. I have been wondering how the AI can tell Which is duck and which is rabbit in the other famous gestalt, silly cartoon. I am guessing thresholds of categories, or class distinctions can be exploited by a weights slider?...In general the results really seem to be a lottery, so I am looking forward to seeing the integration of tools that allow you to fine tune the image via traditional design elements. Thanks. will be watching regularly.
Thanks Francois for reaching out! The AI world is moving fast! More training data added all the time. Midjourney version 4 knows so much more than version 3. Version 5 is right around the corner. Image prompts can be used in Stable Diffusion but the program works differently from MJ. You can train SD with a handful of images on a new concept - that's what the app Lensa does.I'm generally happy with MJ results, but I try to check in with the other AIs to see what's changing.
But you're not expressing yourself... typing a prompt is no different from typing a google search. You didn't create those images, a computer program created them using stolen data from artists. It wouldn't exist without the work of thousands of uncompensated artists, and MJ is making a ton of money off that. I hope the l@wsuit goes through. Notice how they don't even credit any artists in the video?
@@Abby-ug4xc for someone with such an opinion, you're really behind the times. You should use and understand the tools you're criticizing. Don't stand in the way of the democratization of art.
@@MakingPhoto Art has always been democratic. I grew up poor and art was my passion. There is very little barrier to picking up a pencil and notepad and tablets are more affordable every year. I have no use for image generators, even if they were ethical and not using stolen data. There's no creativity because you're not creating anything. You're prompting a program to generate something.
@@Abby-ug4xc No. You are telling anyone who can't draw that they can't express themselves. They do not use stolen data - don't believe everything you read! You simply don't understand how the GENERATIVE process works. Try it - then tell me you don't feel any sense of creativity. And even if you don't, it's not up to you to tell millions of other people that they can't feel the sense of creation. I am a classical musician and professional photographer. I can do art like others can't. But I'm not going to deny someone the experience of art because I don't like their tools.
Wow! And again Wow! you have really blown my mind. you have impressed me soooo much! You have a new subscriber to your channel. i ve never tried the img2img in Midjourney by far, but after your video you have opened my eyes to new possibilities. Thank you so much.
Wonderful. Thanks Janet. I like the tone of your voice and the immaculate delivery of the video. I just discovered Midjourney this week and I am very interested. I was struggling with the text prompting and you have save me with this video. Again, thanks. Would you like to give me a private training.
@@MakingPhoto Hi Jenn, thanks for your prompt reply. I visited the site and I would like to join the course, but I only have the Midjourney basic subscription ($10/month). Since I am new to the AI art, I would like perfect my skills before committing to an average tier subscription. Can I still join your Zoom session with my current Midjourney's subscription?
@@MinaLingo Mina. We can work within the Basic subscription. I can show you a lot of options before you try it for yourself. I have unlimited generations, so I don't mind.
@@MakingPhoto Hi Jenn. Sorry, my name is Del. I have checked the basic course and I realised that it was already started. When is the next course planned?
Very good. I realise I've been trying to use text when I could just be blending uploaded images. Time to have some fun. Thanks! :D "When you're showing Midjourney what you want, you're more likely to get it." Absolutely!
I love BlueWillow. BlueWillow is making some super artistic images, even in its current early stage. I believe that the speed of image generation is a little bit faster than MJ, and more significantly, BW is totally free.
Waiting for a day when I get Death star in expected style, but without mixing the shapes! It is blending the styles and subjects alike. Partial solution: have 1 image and text the rest.
The Death Star had no other text except "Death Star" and the photo. You can definitely make it better with more description. I was just blown away by how close it got on the first step.
I have been trying to use Mid Journey on a Mac book Pro and I see how to upload a photo with the plus sign. However Macs do not have a right click option. So how do you go about copying the URL via you are operating on a Mac book Pro?
Is it possible to add a parameter to a variation? I.e: instead of just clicking the V3 button say: add water to V3 or generate V3 variations with water in the background
This is a whole big thing! It is POSSIBLE to put text on an image, but it's difficult. I have a whole video on this ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-_6q5FQIRzAs.html
is there a special order in which you need to put the url of the pictures if I wanted to blend my face with some character to make sure it is my face on the characters and not vice versa?
No, but you can try adding --iw 2 to a --v 5 prompt. IW is image weights. This might help. Also, start prompt by describing the important features of the images. What you want the AI to make sure and see.
Hi, I have recently started watching your content and have to say its awesome. I tried blending my own pic with an AI generated one like the horse with girl example but it did not work.
Thank you for the video. Wondering how do I get a colorful image when I have just a pencil sketch? My cousin draw a policeman when he was a kid. I still have his picture, and I wanted to generate colorful cartoons with the style of his policeman. However, the AI keeps rendering new colorless pencil drawings no matter what prompts I add. I'm using Bluewillow though, because that's still free. I'm thinking now if it's worth paying for Midjourney for a month just to get this idea implemented and if it's possible at all?
You should be able to get an image based on the sketch. Add the sketch as an image prompt and prompt for colorful image. If you want a free image generator, try SeaArt. I have another video on this AI. There should be a slider where you can increase the image weight.
@@MakingPhoto I have tried adding words "colorful colored cartoon policeman". The best result I got was a colored face and colored drawing lines but the background and the uniform remained completely white. I guess, I'd need to add more hints like "blue uniform" and also background.
@@camelCased Yes, you're on the right track. Use the sketch as the image prompt. Call attention to the details you like about the sketch and add missing information.
It may be the face thing I talked about. MJ doesn't like covering the face with anything. Try more than one photo of you and just the background and see if that works. Then try adding prompts for the terminator part.
This is amazing thank you! i'm very new to AI and wonder if you can advise on how I can get my creations to fit within the boundaries - so often I like a creation but part of it is lopped of which is frustrating. Many thank! :)
Carol, thanks for tuning in! Yes, this sometimes happens. Not as much as it used to. Try adding "wide angle" to your prompt. There are some other tricks like "full body shot" for portraits.
Hi Jenn, Is this feature (Img2Img) only for Midjourney Subscribers? I was able to upload . However, the right mouse to copy the image URL is Disabled. Nothing shows up on my screen. I've created my own server with Midjourney /imagine working with no hiccups. Any suggestions?
Excellent tutorial that made made me sub! How do you tranfarevface from one image to another unchanged? So transfare face from img 1 to a knights body in img 2? (I hope question makes sense)
Not all my avatars worked every time. Are you starting with a short description of yourself from the images? The only thing I can think of is to change out the images. Start with the strongest image. Then add another and another. Make sure the images are a good likeness of you.
This is a great video, just wanted to ask, to get best results should the image that you want to use from the URL be the same aspect ratio as the output image or does it not matter to much?
Thanks! I haven't noticed much difference. Version 4 aspect ratios are so limited anyway. Let's check this when the aspect ratios get fixed (or version 5 comes out!)
Odd. I tried Img prompting in V4 yesterday and got an error saying that v4 requires a text prompt as well. Maybe that was just for a single image, not two
for the AI selfies, I ran into the issue where it merges 'your' photo with the reference photo, instead of a straight swap. Did you have any similar problems?
Mmm I’m trying to create a photo of myself next to a celebrity. I uploaded my photo and the celebrity photo but MJ Blended them into one character. Is there a way to let MJ create a photo of me next to the celebrity, Rather than blending?
Mmm I’m trying to create a photo of myself next to a celebrity. I uploaded my photo and the celebrity photo but MJ Blended them into one character. Is there a way to let MJ create a photo of me next to the celebrity, Rather than blending?
"it's great to transfer styles blocked by the AI" thanks I puked. The lack or respect for artists, like they're just products to shop from like a grocery store, is just depressing.
This is a ridiculous statement. Have you never been inspired by existing art in your own work? MJ takes a reference and creates something unique for you, just like your own brain may be inclined to do when seeing something inspiring.
Naturally no artists were credited for having their work used in this video. 🤕 I wonder how many people watching this are unaware that many artists (including, explicitly, Greg Rutkowski, the artist of the fire-breathing dragon) do NOT consent to having their work used in this way?
Make a choice. You can be an artist and produce art or you can use AI and be a thief and a phony. AI is not a creative tool. AI replaces you. You are no longer required.
If you spent your life creating artwork only for it to be done by a machine faster and better, there are 2 things you can do. Shut up about it. Find a new job.