🎯 Key Takeaways for quick navigation: 00:00 *Leonardo AI Canvas 2.0 allows easy editing and manipulation of images, offering features like uploading, moving, erasing, and generating within windows.* 03:45 *Canvas 2.0 introduces new modes, including text to image, image to image, sketch to image, and more, providing creative options for generating and editing images.* 07:53 *The video explores using AI to add facial features like mustaches and glasses, demonstrating the flexibility and fun aspects of image manipulation.* 11:09 *The control net feature enables precise image manipulation by referencing the original image's layout while allowing artistic freedom by adjusting control net weight.* 13:41 *Depth to image uses depth of field effects for creative image changes, and pose to image replicates poses onto different characters, showcasing the versatility of Canvas 2.0.* 16:44 *The video concludes by summarizing the powerful techniques and features of the Leonardo Canvas editor for image manipulation and creativity.* Made with HARPA AI
Thank you so much! !This is the best video I found on Canvas... So many videos don't explain what each setting does or they do the crazy constant circling with their mouse curser that is so distracting. I am proficient in photoshop and illustrator so As usual... I try first get directions later... I was using this last night and if I had watched this video first lol I definitely would have saved myself so much time trying to figure it out. This video was so well done. Thank you! I will be subscribing to your channel You are an excellent teacher.
It is a great tool! One thing regarding the depth option in Controlnet. It is not directly related to depth of field in an image, it is analysing the actual depth in the photo. You can check out different sources about that. So it is a great way to use the layout of a scene as the basis for a new picture. Thanks for illustrating the new things in the new Canvas!
Thank you. I didn't mention seed as most people learning their way around won't need it. But essentially seed is a number which dictates the starting point of an image generation. So if you give Leonardo AI the exact same prompt twice, you will get different results. Apply the same seed, do the same prompt and the identical seed makes the images identical. you are then likely to get closer results with any changes you make if you use the exact same seed.
Edge to Image looks like it could have a lot of potential for making Textures for 3D projects and following the shape of the texture. I'm thinking Armor or Cyberpunk clothing pieces.
Great tutorial but... At 11:18, how did you manage to generate with ControlNet IN CANVAS MODE !? I CAN indeed do this in "AI Image Generation" mode ("normal" mode) but if I do this in Canvas mode, it does'nt work. It displays this error message : "controlnet requires an init image". I suppose it means I have to put an base image like I do when I'm in normal mode : I just click and drop my image in the left toolbar, at the bottom, where there is a "Image to Image" button and it is written : "Upload or drag and drop an image to use as input". But it seems that it is possible in normal mode only ! ... Or I'm missing something... 🤔
@@WadeMcMaster lexica is also good but the biggest drawback is it's models which have similar color grading and typical art format. Lexica is good for art based works not for realistic stuffs
Thanks for the tutorial! Clear and easy to follow. I used one image and used another as the pose to be mirrored to the first image. I have a free account and couldn't get what I want. Do you think Leo is the right tool to achieve this? Or I might have missed steps with the Canvas.
Thank you! I find Leo to be one of the better ones for sure. But someone's you do get weird results regardless. You could try playground ai too, you may get the result you're after there?
I believe the Midjourney is by far the easiest and best image quality for most things, but surreal I think it does best. However Leonardo AI has better resolution and can perform better in specific areas - the photoreal mode makes better images in many cases (not always), Leonardo AI has strengths in the themes of the models you can choose from, in many cases better than midjourney - but Midjourney just does almost everything well
Wow, Leonardo AI has done it again with their canvas feature updates! Can you also please cover the new photoreal feature in Leonardo AI? It looks like they have reached the same level of photorealistic quality as Midjourney. I'm wondering if I should ditch Midjourney and go all-in with Leonardo AI, since Midjourney's updates take forever.
Hi, your video is very helpful. But I would like to ask you: is in Leonardo a way to straight an image generated in any model of the Leonardo? thank you
Thank you so much for this video!! I had not tried Canvas because I had NO IDEA how it worked, but I've been having a blast playing with it now. Is there some way to make the bounding box larger though? It's great for adding small things to an image, but I can't figure out how to add large details.
Thank you! The only way is to up the resolution in the right bar, and it will change the box to fit that resolution. Otherwise, it's a case of shrinking things into place.
I haven't seen any sorry, there may be some on RUclips somewhere. If you can find an online service to remove the background (like remove.bg) from the photo and then upload it it should be easy enough.
i wish if pose to img could do like that for example giving my own img then pose to img chnage my image pose to something different instead of generting new one
i hope u would be able to understand what i wanna say -_- hope so@@WadeMcMaster plus also will you plz make a tutorial as a comparision of leaonardo AI and playground AI between, for example you sholud choose half body any human and make it OUTpainting and inpainting tries, for outpaint try to make his full body then with inpaint make him cyborg or muscular man >> plzzzzz highly req!!!
But how do you save? I do hit download but it only downloads it as a PNG file. Not JPEG, so I usually use the snip tool from my computer tools, snip it and save as JPEG but that seems like im doing it the wrong way.
It is akways getting from a realistic source into sonething anime or artistic. If you want to stay in a photorealistic cinematic style, what would you suggest to use as model? Or is Midjourney / Fulljourney a better option ?
Can you highlight the head and space to one side, and say "dog's head looking left" or "dog's head faceing left" or "dog's head side profile" or similar?
In Canva, just select the image ands there's an option called BG Remover - but it is premium. Leonardo has a background removal option you can find by opening up the generation in your feed / Library. If neither option suits, Adobe has a free background remover here: www.adobe.com/express/feature/image/remove-background That should cover just about all bases :)
I felt the same at first... But after spending all night playing with it, I found it easier than Mid Journey but harder than Dalle-2... But I got way better results with Leonardo. Especially with Canvas to help me fix images. Much easier than pulling them into photoshop or illustrator... I guess coming from Adobe software this seems so much easier to me... I like the help prompts but they could be a little more descriptive.. But still.. I am in awe of what they produced.
You could go into Canvas, upload a pic of you, upload the dogs head - then mask the surrounding area of the dogs head and try something like 'blend images together' - that may blend the dogs head onto the body effectively. I haven't tried it exactly so can't be sure, it can take some testing.
@@WadeMcMaster thanks. Yes, this exact thing you said is what I've been trying to do for weeks with Leonardo, but I don't see the blending image function anywhere. I only find the "adding" images to other images function which is totally different. Any idea where I can find the "blending images" function?
I've tried a lot of these and none will give me what I want. The time spent trying to get it to work would have been better spent on learning to draw. Or at least photobash.
🎯 Key Takeaways for quick navigation:
00:00 *Leonardo AI Canvas 2.0 allows easy editing and manipulation of images, offering features like uploading, moving, erasing, and generating within windows.*
03:45 *Canvas 2.0 introduces new modes, including text to image, image to image, sketch to image, and more, providing creative options for generating and editing images.*
07:53 *The video explores using AI to add facial features like mustaches and glasses, demonstrating the flexibility and fun aspects of image manipulation.*
11:09 *The control net feature enables precise image manipulation by referencing the original image's layout while allowing artistic freedom by adjusting control net weight.*
13:41 *Depth to image uses depth of field effects for creative image changes, and pose to image replicates poses onto different characters, showcasing the versatility of Canvas 2.0.*
16:44 *The video concludes by summarizing the powerful techniques and features of the Leonardo Canvas editor for image manipulation and creativity.*
Made with HARPA AI
Thank you so much! !This is the best video I found on Canvas... So many videos don't explain what each setting does or they do the crazy constant circling with their mouse curser that is so distracting.
I am proficient in photoshop and illustrator so As usual... I try first get directions later... I was using this last night and if I had watched this video first lol I definitely would have saved myself so much time trying to figure it out.
This video was so well done. Thank you! I will be subscribing to your channel
You are an excellent teacher.
Thank you so much for the kind words and the super thanks, it's a much appreciated gesture! I'm glad I could help :)
@@WadeMcMaster you absolutely did!!! Thank you!
It is a great tool! One thing regarding the depth option in Controlnet. It is not directly related to depth of field in an image, it is analysing the actual depth in the photo. You can check out different sources about that. So it is a great way to use the layout of a scene as the basis for a new picture.
Thanks for illustrating the new things in the new Canvas!
Thanks for the info, I didn't realize! So much to take in haha.
The best video yet
Thank you!
this is exactly the video ive been searching for.. thank u
Thank you, glad you like it!
Concise, clear. Thank you.
Glad it was helpful! Thank you!
Thank-you! You explained this beautifully! I'm a new subscriber!
Awesome! Thank you, much appreciated :)
Thanks for another comprehensive overview. ✅
Thank you as always mate! I always appreciate the support :)
seeeeed??
very detailed and professional, ty
Thank you. I didn't mention seed as most people learning their way around won't need it. But essentially seed is a number which dictates the starting point of an image generation.
So if you give Leonardo AI the exact same prompt twice, you will get different results. Apply the same seed, do the same prompt and the identical seed makes the images identical. you are then likely to get closer results with any changes you make if you use the exact same seed.
@@WadeMcMaster thank you very much, it will help a lot.
great tutorial, thank you!
No problem, thank you!
Lots to take in. Just Subscribed to your channel. I will check out your other video. Thanks 4 sharing! Cheers!👀👍😁
Awesome, thank you so much!
This is a great tutorial!! I’m new to this application and I need to learn the basics, this does just that. Thank you very much. Subscribed!!!❤❤
Thank you for watching the video and the sub! Much appreciated :)
Great explanation, you make a great content!
Thank you! Much appreciated mate :)
Edge to Image looks like it could have a lot of potential for making Textures for 3D projects and following the shape of the texture. I'm thinking Armor or Cyberpunk clothing pieces.
Yeah thats true! Man there are so many great uses for this stuff. It's amazing,
Great tutorial but... At 11:18, how did you manage to generate with ControlNet IN CANVAS MODE !? I CAN indeed do this in "AI Image Generation" mode ("normal" mode) but if I do this in Canvas mode, it does'nt work. It displays this error message : "controlnet requires an init image". I suppose it means I have to put an base image like I do when I'm in normal mode : I just click and drop my image in the left toolbar, at the bottom, where there is a "Image to Image" button and it is written : "Upload or drag and drop an image to use as input". But it seems that it is possible in normal mode only ! ... Or I'm missing something... 🤔
Im not sure why thats happening, as long as your generation window sits over the image it should work. It could be a temporary bug too
Try creating jewellery or 3d objects or vehicles or perspective sort of designs/artworks
I haven't yet! You get good results?
@@WadeMcMaster I have used lexica and playground, they give excellent results with proper prompts.
I've used Playground too, it's an awesome platform - have been meaning to try Lexica, but havent yet!
@@WadeMcMaster lexica is also good but the biggest drawback is it's models which have similar color grading and typical art format. Lexica is good for art based works not for realistic stuffs
Thanks for the tutorial! Clear and easy to follow. I used one image and used another as the pose to be mirrored to the first image. I have a free account and couldn't get what I want. Do you think Leo is the right tool to achieve this? Or I might have missed steps with the Canvas.
Thank you! I find Leo to be one of the better ones for sure. But someone's you do get weird results regardless. You could try playground ai too, you may get the result you're after there?
very useful video
Thanks for the video. Which one do you think understand and create better surreal art images, Midjourney, or Leonardo?
I believe the Midjourney is by far the easiest and best image quality for most things, but surreal I think it does best. However Leonardo AI has better resolution and can perform better in specific areas - the photoreal mode makes better images in many cases (not always), Leonardo AI has strengths in the themes of the models you can choose from, in many cases better than midjourney - but Midjourney just does almost everything well
The quality is much improved if Alchemy option is activated. A huge difference in quality.
Agreed, 100%
Wow, Leonardo AI has done it again with their canvas feature updates! Can you also please cover the new photoreal feature in Leonardo AI? It looks like they have reached the same level of photorealistic quality as Midjourney. I'm wondering if I should ditch Midjourney and go all-in with Leonardo AI, since Midjourney's updates take forever.
I'll check it out! Thank you for the suggestion
@@WadeMcMaster your welcome, looking forward to the upcoming review on photoreal,, thank you so much..
No worries mate :)
Hi, your video is very helpful. But I would like to ask you: is in Leonardo a way to straight an image generated in any model of the Leonardo? thank you
Do you mean straighten an image? To fix a poor angle?
Thank you so much for this video!! I had not tried Canvas because I had NO IDEA how it worked, but I've been having a blast playing with it now. Is there some way to make the bounding box larger though? It's great for adding small things to an image, but I can't figure out how to add large details.
Thank you! The only way is to up the resolution in the right bar, and it will change the box to fit that resolution. Otherwise, it's a case of shrinking things into place.
Hey Wade, thanks for the vid. I need a little help. There is no "pose to Image" on Leonardo anymore, at least not that I could find. What happened??
Not sure, I have it here still. Sorry I have no idea
there is canvas training on how to add another background to the original photo
I haven't seen any sorry, there may be some on RUclips somewhere. If you can find an online service to remove the background (like remove.bg) from the photo and then upload it it should be easy enough.
i wish if pose to img could do like that for example giving my own img then pose to img chnage my image pose to something different instead of generting new one
That would be cool. I imagine in time a feature like that will appear.
i hope u would be able to understand what i wanna say -_- hope so@@WadeMcMaster
plus also will you plz make a tutorial as a comparision of leaonardo AI and playground AI between, for example you sholud choose half body any human and make it OUTpainting and inpainting tries, for outpaint try to make his full body then with inpaint make him cyborg or muscular man >> plzzzzz highly req!!!
Is it possible to add a picture with a picture like saying at the dog to a photo of yourself
You can upload two images, erase around them and then generate a blend between them.
But how do you save? I do hit download but it only downloads it as a PNG file. Not JPEG, so I usually use the snip tool from my computer tools, snip it and save as JPEG but that seems like im doing it the wrong way.
It's only PNG unfortunately. There are some online converters you can use to convert PNG to jpg easily.
It is akways getting from a realistic source into sonething anime or artistic.
If you want to stay in a photorealistic cinematic style, what would you suggest to use as model? Or is Midjourney / Fulljourney a better option ?
I prefer midjourney for that personally, but I know some people do get good realistic results with leo, I need to dig deeper with it
@@WadeMcMaster right….. have you tried Fulljouney?
I haven't heard of it until you mentioning it now, I'll take a look :)
@@WadeMcMaster Thank me later 😉😎
How can I change the position of the dog head (turn to left or right)? Does anyone knows
Can you highlight the head and space to one side, and say "dog's head looking left" or "dog's head faceing left" or "dog's head side profile" or similar?
nice
Thank you!
How to remove background in canva?
In Canva, just select the image ands there's an option called BG Remover - but it is premium. Leonardo has a background removal option you can find by opening up the generation in your feed / Library.
If neither option suits, Adobe has a free background remover here: www.adobe.com/express/feature/image/remove-background
That should cover just about all bases :)
difficult. proposal: ai prompt to type what a user want to do, connected then to automated system actions. interface is messy. needs a simplification
I hear ya, it's hard to have a tonne of options and keep it simple. In time hopefully it gets easier
I felt the same at first... But after spending all night playing with it, I found it easier than Mid Journey but harder than Dalle-2... But I got way better results with Leonardo. Especially with Canvas to help me fix images. Much easier than pulling them into photoshop or illustrator... I guess coming from Adobe software this seems so much easier to me...
I like the help prompts but they could be a little more descriptive.. But still.. I am in awe of what they produced.
Nothing happens! Is this for paid users? What you have shown here is not working. My canvas is always the same after prompt and generation.
It may be not working properly right now (which happens). I have used this on the free plan many times.
Nice. But the most important task for me is, "how do I put my pet's face on a human body" with Leonardo Ai free version? Thanks.
You could go into Canvas, upload a pic of you, upload the dogs head - then mask the surrounding area of the dogs head and try something like 'blend images together' - that may blend the dogs head onto the body effectively. I haven't tried it exactly so can't be sure, it can take some testing.
@@WadeMcMaster thanks. Yes, this exact thing you said is what I've been trying to do for weeks with Leonardo, but I don't see the blending image function anywhere. I only find the "adding" images to other images function which is totally different. Any idea where I can find the "blending images" function?
I don't think there is a blending images function sorry mate (I havent seen one)
優質
Thank you!
too much using token, not really worthed.
in this case firefly still on top by large margin even for the price.
That's fair mate, everyone has a preference :)
The problem with leonardo is that it is too expensive, and not free
I have a separate free account which works well. You jsut have limitations on certain features and the amount you can generate daily.
I've tried a lot of these and none will give me what I want. The time spent trying to get it to work would have been better spent on learning to draw. Or at least photobash.
I love AI, but photo bashing in Photoshop is still one of my favourite things to do :) in time this technology will improve