Timothy Nice’s Post

View profile for Timothy Nice, graphic

Sr. Director @ LightRiver | Product Designer & App Creator | Embracing AI, FlutterFlow, and NoCode Tools | Offering Tech Tips, Tools and Growth Resources

Watch out Midjourney! I'm a huge midjourney fan, (Started on V2). I hated the discord interface initially but got so used to it I find that I still use it and all the --ar 16:9 --r 10 over the web interface. I also love that I can seamlessly move between mobile and web with discord and have the same experience, something I didn't realize I even cared about when it comes to Gen AI. Now that I'm also apart of the Leonardo Creator Program, and have access to all of the features without limits, I find myself focused on the things that Midjourney can't do. They can both - Both generate great images (I have different preferences for each depending on the type of image, maybe that should be a post...) - Both upscale - Both have the ability to use images to influence new generations - Both have lots of folks creating cool things to inspire you and give you ideas BUT - Leo has some really neat things. - (Beta) Universal Upscale - This is more akin to what I've seen from Magnific as far as quality and detail, it's CRAZY and decently controllable. I sorta can't get enough. Upscaling both Leo and MJ images. Lots and lots of upscaling... (I'll post more at some point) - Edge to Image (uses the lines / edges of an image for guidance rather than all aspects. - Elements (I used to think these were dumb, but if one hits with what you're looking for, it's quite nice and you can get a really consistent look quickly) even a very slight value can give some nice flavor. Sorta a more intentional way of adding some random or weird in MJ. This Process: LEFT: This was someone else anime image gen that I started with. (Thank you random Leo user thats only done one thing: Bxjsjsjsiwiw) CENTER: Used Edge to Image and reused the same prompt from the original but tweaked it to be more realistic. Really nice results with this IMO. RIGHT: Upscaled, a few times, with some photoshop and making out some of the weirdness I didn't want in the upscale, then upscaled again, etc. It can get a little harsh, contrasted and even an occasional seam line when you really want to clarify something and let the AI take control, but easy photoshop to tone back the parts you don't want. Anyway, the full final image at huge rez here: https://lnkd.in/gyrdiB2i Original Image here: (Warning, it's very... Anime) https://lnkd.in/gSxbeD8V #genai #aiart Midjourney Leonardo.Ai

  • No alternative text description for this image
Sumil Dutta

Workplace Strategy, Data Analytics and Design | AI enthusiast

2mo

Agree. As a beginner in field of digital art, I found Leonardo interface much more intuitive and easy to navigate, and was able to generate some interesting results in a short amount of time

Dan Alexander

Business Development Manager for Grunwald Ltd

2mo

I find myself using Leo more and more as you can adds loads of detail with more control. Also taking concepts from Midjourney as a base image and use Edge to edge/Upscale for great results.

Like
Reply

Yes, Leonardo.ai is quite interesting and works well with photos and digital art images. Still I'm struggling to come up with anything meaningful when I'm trying to work from my scanned old pencil sketches in the style of European comics/illustrations.

Like
Reply
Christopher Sicurella

MONOxID | AI trainer | YOU.com Ambassador

2mo

😊 And it has a nsfw ON/OFF switch. Sorry, not sorry MJ. Love my Leonardo.

Like
Reply
Richard Johnston

Graphic/Digital Designer & 1st Class Graduate | Still learning!

2mo

I agree, the Universal Upscaler is brilliant!

See more comments

To view or add a comment, sign in

Explore topics