Can someone please explain why in Canvas Editor in "Img2Img" with Alchemy enabled (any version, any settings), after generating an image when "Accept" is pressed, the generated image scales up outside of the canvas area and crops by it (with HD on), or scales down (with HD off). That makes no sense a makes the whole function useless. Leonardo support team showed itself complitly useless.
Thanks for the awesome video. Was so trying to figure out how to edit an image created in Leonardo that kept cutting off the arm. I could not figure it out how to get the right arm included in the image we created. Than I found this great video. It really helped. Now I just need to find a video to figure out best way to optimize Leonardo AI images for print formats so it does not pixelate when on a 4x6 banner, etc. It downloads as jpg only. Keep up the great work on videos.
Thank you! If it helps, I have some videos on the upscaling for print. They are primarily Midjourney focused, but cover an external tool for upscaling. I've been using Topaz Gigapixel effectively for that: Upscaling from Midjourney & using Topaz Gigapixel ruclips.net/video/ak5lvbGtYCQ/видео.html An older video on upscaling & print: ruclips.net/video/v6dnHXzIJPo/видео.html I do recommend using the universal uspcaler or one of the native upscalers within Leonardo AI and then using this tool to maximise the quality. I hope that helps!
Sometimes it can be difficult. I would try prompting 'hair' on the top part of the head maybe? Even select deeper and redo a good portion of the top as just the tiny bit at the top often adds stuff to it. Its tough sometimes but hopefully that helps
@WadeMcMaster Hey, do you know if it is possible to talk to Leonardo, while using canvas, to point out certain element of that image you are working on. For example: ( masking the target area ), "draw a magical staff that uses the same motif as the one seen on the belt of the character". I know by default AI tries to make coherent artistic choices, but can we ask it to pay special attention to some part of the image? I can do the testing... but it is hard to tell if (in case in works) it is pure random luck...
Man that's a good question. I don't know if you can, but I get the impression it's not quite there yet. But I would need to ask or experiment I think...
It works good and is fun with personal photos. However it gave a photo of my son at 14 weird fingers but a perfectly perfect wrist watch 😂 it enlarged the area around him perfectly. We were hiking on a trip
@@Echoes317 There is nowhere to add negative prompts in canvas unless I add them directly to the prompt field but it doesn't matter anyway as Canvas in leonardo has quit working correctly at all for me anyway. I can't change or modify anything like pixel size, number of generations etc. so I quit using it. Pity as it was really useful when it worked.
it sucks, i tried to fix a hand with too many fingers and as a result I got something that looked even worse. I move back to photoshop; even that doesn't give me the best results but it still works better then this.
I do think Photoshop's generative is the bets ai image editing app out there. They all have their drawbacks unfortunately but hopefully they'll improve in time :)
You can definitely do this using Canvas on Leonardo AI. This video ( ruclips.net/video/aw10ykp5n5A/видео.htmlsi=oUtIm1fRkf4F0R1a ) covers some cavnas features along with ControlNet, which woul dbe a great help for what you're trying to achieve.
i only have 4 options on the left, cant even upload my own image! any tips? thanks!
Not sure why mate, are you in the Canvas Editor specifically? Not realtime Canvas?
very cool and update tech.
Can't believe I am the first one ❤❤❤
By the way love to see video😊
Your $1000 check is coming ;) Just kidding, but nice :)
Thank you so much!! For your nice work!🔥🔥
Thank you! Much appreciated :)
Can someone please explain why in Canvas Editor in "Img2Img" with Alchemy enabled (any version, any settings), after generating an image when "Accept" is pressed, the generated image scales up outside of the canvas area and crops by it (with HD on), or scales down (with HD off). That makes no sense a makes the whole function useless.
Leonardo support team showed itself complitly useless.
I'm not sure why that is happening sorry mate. Maybe it's a temporary glitch or try another browser. It shouldn't be happening
I subscribed... just did my first painting following your tutorial.. Thanks!
That's awesome! I love hearing that kind of news :) I hope you have fun making more!
Good lesson. Thanks.
Thank you :)
Thanks for the awesome video. Was so trying to figure out how to edit an image created in Leonardo that kept cutting off the arm. I could not figure it out how to get the right arm included in the image we created. Than I found this great video. It really helped. Now I just need to find a video to figure out best way to optimize Leonardo AI images for print formats so it does not pixelate when on a 4x6 banner, etc. It downloads as jpg only. Keep up the great work on videos.
Thank you! If it helps, I have some videos on the upscaling for print. They are primarily Midjourney focused, but cover an external tool for upscaling. I've been using Topaz Gigapixel effectively for that:
Upscaling from Midjourney & using Topaz Gigapixel ruclips.net/video/ak5lvbGtYCQ/видео.html
An older video on upscaling & print: ruclips.net/video/v6dnHXzIJPo/видео.html
I do recommend using the universal uspcaler or one of the native upscalers within Leonardo AI and then using this tool to maximise the quality.
I hope that helps!
Thank You! It was driving me nuts. All my images are coming out incomplete. Tops of heads missing etc... How do I fix it?
Sometimes it can be difficult. I would try prompting 'hair' on the top part of the head maybe? Even select deeper and redo a good portion of the top as just the tiny bit at the top often adds stuff to it. Its tough sometimes but hopefully that helps
@@WadeMcMaster Thanks
Use the prompt "merge images" or "merge image" for that
Thank you!
@WadeMcMaster Hey, do you know if it is possible to talk to Leonardo, while using canvas, to point out certain element of that image you are working on. For example: ( masking the target area ), "draw a magical staff that uses the same motif as the one seen on the belt of the character".
I know by default AI tries to make coherent artistic choices, but can we ask it to pay special attention to some part of the image?
I can do the testing... but it is hard to tell if (in case in works) it is pure random luck...
Man that's a good question. I don't know if you can, but I get the impression it's not quite there yet. But I would need to ask or experiment I think...
It works good and is fun with personal photos. However it gave a photo of my son at 14 weird fingers but a perfectly perfect wrist watch 😂 it enlarged the area around him perfectly. We were hiking on a trip
Yeah its great for scenery. Getting human features can be a challenge haha
Use the negative prompts to limit the deformities
@@Echoes317 There is nowhere to add negative prompts in canvas unless I add them directly to the prompt field but it doesn't matter anyway as Canvas in leonardo has quit working correctly at all for me anyway. I can't change or modify anything like pixel size, number of generations etc. so I quit using it. Pity as it was really useful when it worked.
@@BlackCat_2 Right on the bttom left hand corner is a button for negative prompts. Wish you the best of luck
Hi, how do I remove text from images in canvas ai ?? Thanks for the video 🙏
You could highlight the area, use the negative prompt to type in 'letter, numbers, words'
I want to edit an image to have a complete different painting style. how can i achieve this?
I would say experimenting with the control net options with an image prompt would be best
I can't find a way to save my modification version to update my seed matrix , why ?
Do you mean save the image from canvas into your Leonardo ai library? I don't believe you can't the moment sorry
@@WadeMcMaster where is the save button on canvas ? i find no where, i am a green hand, i just found a download image button to device
You can't at this stage, only download unfortunately
@@WadeMcMaster thanks
I think for joining the image you do not need to erase the area so that the a.i will have some sort of a guide.
Yeah I think your right mate.
it sucks, i tried to fix a hand with too many fingers and as a result I got something that looked even worse. I move back to photoshop; even that doesn't give me the best results but it still works better then this.
I do think Photoshop's generative is the bets ai image editing app out there. They all have their drawbacks unfortunately but hopefully they'll improve in time :)
i cant play with it on mobile
That's a strong reason to use BlueWIllow or MJ - discord I think makes it easier for mobile. But I know Dall E has a decent app interface.
@@WadeMcMaster will... ummm... your true but this is the only free one
Yeah thats true :(
I couldn't just do a black and white old photo in a modern color treatment
Photoshop may be better for that, they have a colorize neural filter for old photos
oh brother thank you ..... I just needed to change the face to an otherwise kille image .....
I'm glad you found it useful mate! Thank you :)
any body knows how to change my selfie into sthng cool with ai?
You can definitely do this using Canvas on Leonardo AI. This video ( ruclips.net/video/aw10ykp5n5A/видео.htmlsi=oUtIm1fRkf4F0R1a ) covers some cavnas features along with ControlNet, which woul dbe a great help for what you're trying to achieve.