Goddesses - My First Experience Making AI Art

in #hive-1224722 years ago

When @edprivat started to share his AI art here on Hive I felt drawn to try it out myself. I was fascinated and decided to start with Midjourney. This was a month ago. In the beginning, it was pretty frustrating, nothing turned out even close to what I had in mind. But I discovered that if you started to get something that was somehow interesting you could go from there. Eventually, the algorithm would work in your favor, haha. And I could definitely learn to write better prompts, I haven't gotten into the more technical part of writing prompts yet.

After some time, I started to enjoy it more and today I thought I would share what I have made so far. I haven't been very active, or I don't know, everything is relative, I just checked, and here is the result:

Lifetime Usage: 549 images (5.92 hours)

My idea was to create images of goddesses. So that's what I have been focusing on. I have also been playing around with more 'dreamy' landscapes etc, and some of them turned out in quite interesting ways. But in this post, I'll share some of the goddess images.

Namnlös design copy 3.png

The first issue I ran into was that the face kind of always turned out a bit distorted. I also got quite dramatic, dark, or sad results. My idea was more empowered, light, and happy goddesses haha. I don't know what changed but I somehow managed to move more in the direction I was looking for.

The first image below is one of my first attempts. She looks a bit sad but I liked it anyway. But, as you can see, something is not quite right with her nose. And she looks very young, almost like a child. That wasn't what I was looking for.

mamrita_wild_goddess_in_white_dress_golden_shimmer_blond_1b0a6288-ea59-4c67-a999-33aaf24f0773.png

Some alternations later I started to get some results I was rather happy with. Like @edprivat shared, I have so far found the 'remaster' function to be very helpful. The image below is created in that way. Even though I can't remember if that was the first result I got, probably not. I wanted long blond hair and a white dress so I think the algorithm got that right.

mamrita_wild_goddess_in_white_dress_golden_shimmer_blond_bb6b15dc-0891-4fda-92bd-802a11ae7c16.png

mamrita_wild_goddess_in_white_dress_golden_shimmer_blond_6b9c1479-07c7-47a6-b1c1-e97114610b14.png

mamrita_wild_goddess_in_white_dress_golden_shimmer_blond_ca001a7c-2c57-49bf-b77d-cc2917bfc379.png

Namnlös design copy 3.png

The process below is similar. I like the image somehow but not what I was looking for.

mamrita_wild_goddess_in_white_dress_golden_shimmer_blond_36499371-f88d-4547-9e91-ee21ace188c4.png

Below is a remastered version. This is actually my favorite so far. Her lips look a bit unreal maybe (and eyes, haha) but I like the expression.

mamrita_wild_goddess_in_white_dress_golden_shimmer_blond_9811f5a2-a2d2-4c2a-b4a9-750875fec830.png

The 2 below are also variations (if I remember correctly) of the same prompt. I actually found it quite interesting that she looks so 'dead', almost like a stuffed animal. But again, not what I was looking for.

mamrita_wild_goddess_in_white_dress_golden_shimmer_blond_a5fd87a3-1626-402e-b54e-4a4d30219264.png

The one below looks more alive, maybe, but a bit too conventional for my taste.

mamrita_wild_goddess_in_white_dress_golden_shimmer_blond_34eebe6d-5493-4e70-b773-8d485ec5deb4.png

Namnlös design copy 3.png

Below is the same process. The second image is the remastered version. (Here I was playing around with a more dramatic expression).

mamrita_goddess_of_the_underworld_rising_up_fierce_beautiful_fa_f480c491-876e-415e-ba4c-837b99d00f7b.png

mamrita_goddess_of_the_underworld_rising_up_fierce_beautiful_fa_32c0a333-b1cb-4ecf-a717-7f11b8b22ef8.png

Namnlös design copy 3.png

Moving on to more 'peaceful' results.

mamrita_beautiful_bright_goddess_returns_home_after_being_lost__f985dc58-8908-478b-93a1-ca224f7205e2.png

mamrita_beautiful_bright_goddess_returns_home_after_being_lost__22b1dfef-468b-45de-a42a-184d6e2d9546.png

One frustrating thing that often happens is that some detail kind of ruins the whole image. In my case, often a flower turns out in the 'wrong' place. In the picture below I'm not happy with 'the second moon' to the right. Of course, you can edit that with software if you like, but all images are unedited. Straight from Midjourney 🙂

mamrita_beautiful_bright_goddess_returns_home_after_being_lost__0661467f-6bcf-498e-863b-fbcd64c47d84.png

Namnlös design copy 3.png

I'll end this post with some pink goddesses.

mamrita_empowered_goddess_one_with_nature_white_dress_blond_hai_1b9d864d-eac1-44f6-b728-ffb61489d0f7.png

mamrita_empowered_goddess_one_with_nature_white_dress_blond_hai_7a57c4a1-5abd-4e56-ab90-033b0725410d.png

mamrita_empowered_goddess_one_with_nature_flowers_in_hair_white_f4561f70-1b95-4400-a7f1-03dc1ec9078f.png

I'm intrigued to learn and make more AI art. It's rather fascinating.

Namnlös design copy 3.png

Thanks for reading 🌸

Love and blessings to you all 💚

Mamrita Coaching.png

Sort:  

They are very magical and beautiful, very inspiring.

Thank you for your kind words. Glad you liked them 🙂

You are welcome☺️!

Ethereal and beautiful. AI art is something I want to explore

Thank you 🌿 Yes it's a bit addictive once you get into it 🙂

Great work. If you've got a powerful computer, you can make this stuff on your own computer using Stable Diffusion, and it isn't all that hard. Saves the money on the MidJourney subscription!

Oh, thank you for bringing this to my attention. I'll look it up for sure 🙂

I wrote a tutorial in the community a little while ago, it might be a bit out of date, but it still relevant. Feel free to check it out.

Thank you 🙂

No problem, I hope to see more of your artwork soon!

Thank you. I just read one of your posts and one thing that I have discovered using Midjorney is that so many words are banned. I have already found that a bit limiting but I'll try to work around it somehow, to come up with more creative words I guess.

With the local model of Stable Diffusion, you can be a lot more artistic. Some people have even trained their own models with different content than the core data.

With MidJourney, you can of course use a thesaurus to get around particular words, but they make it clear in their terms of service that they do not like that.

If for example, you wanted a historical, athenian statue or bust; you'd have a hard time getting that represented in MidJourney in the same way in which such statues are on display in public squares throughout Europe.

Thank you for this valuable information. I have a few things to ponder now 🙂

If for example, you wanted a historical, athenian statue or bust; you'd have a hard time getting that represented in MidJourney in the same way in which such statues are on display in public squares throughout Europe.

Yes, that I understand from my experience so far.

Such beautiful work! I haven't really started with this yet, but I want to. I realized that Midjourney uses Discord but I don't know how to proceed, what's your advice there? Thank you for the inspiration, the images are truly gorgeous and there's a deeper revelation about reality and life, I think, in the way the algorithm starts working in the artist's favor after a while.

Thank you for your kind words, I'm glad you found this post inspiring 🙂

You can read more about how to get started with Midjourney here:

https://midjourney.gitbook.io/docs/

You'll need a subscription though so it's not free (you can create 25 images for free if I remember correctly but that won't take you very far, after that you need to subscribe).

And there are of course other options, Midjourney was just the one I chose to start out with.

Looking forward to seeing some of your work soon 🙂

Hi there, amaIng art... And needless to say very beautifuly generated... Ofcourse you have some part in that creativity!.

Tell me is midjourney the program, or well one of the programs you fancy art makers use for these AI generated arts🤓

Anyways added to my notifications I would love to see more of this! Perhaps even try this myself!

Thank you, I'm glad you liked it 🙂

Well, Midjourney just happened to be the program I started out with. I haven't explored other options (yet) so I can't really say if it's the 'best'.

Thank you for the encouragement 🌿

Well honestly all of those generated art looks amazing!! To say the least..

Thank you 🙂

It's SUPER fascinating!! It would be nice to add commands like "change eyes to green" or "take away two legs from six legged kangaroo'? You are right about the algorithm though.

Your goddesses are lovely. Funny adding "beautiful" as a prompt word sometimes, it's so subjective. So many questions. Funny all my subjects are white - we don't say "white girl' but that's what it inevitably throws up.

riverflows_girl_with_long_green_hair_foraging_in_forest_for_mus_ea40f9bb-f6b0-4363-99ea-fe606455fbcd.png

riverflows__post_apocalyptic_Australian_landscape_with_a_girl_r_12827e20-85c0-4763-b53d-497495b9bd73.png

Yes, the algorithm can't really cope with input like that (yet).

And yes I know, I do use 'beautiful' as a prompt word quite a lot.

I should post more soon 🙂

Wow I really love it! You already have your own style, I wouldn't be able to create the ones you did! Excellent work 👍😁 I look forward to see more!!!! Thanks for the shout out 😌

Thanks @edprivat 🙂 That's very kind and encouraging. I'm hoping to see more of your art soon 😊